-----Original Message-----
From: Shawn K. Hall [mailto:vb@...]
Sent: Friday, March 23, 2001 9:12 PM
To: VBPros@yahoogroups. com
Subject: [VBPros] 'long' math

Hello All,

I've had this idea for quite a while about how to perform operations
on very large integers. Numbers that are beyond the scope of your
imagination in some cases. A few days ago someone pointed me to a
resource that 'works' for smaller 'very large integers' - but it
needs some work.

I tried mulitplying a number into the thousand digit range and it
'died' - to put it politely. And it isn't capable of some of the
features I need to implement. Anyone have any ideas on how I can add
extensive mathematical features using "very" large numbers to a
project?

Things I'll need to be able to do 'long math' including roots,
powers and stuff with numbers in the *billions* of *digits* range.
Is it even possible? How? I don't care if I have to resort to C++ -
as long as I can expose it to a VB object model.

Thanks,

Shawn
http://ReliableAnswers.com




To unsubscribe from this group, send an email to:
VBPros-unsubscribe@yahoogroups.com



Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.

-----Original Message-----
From: DonEMitchell [mailto:donemitchell@...]
Sent: Friday, March 23, 2001 7:58 AM
To: VBPros@yahoogroups.com
Subject: RE: [VBPros] XML Parser Project Proposal

Shucks, Will.

 

Here was my original question...

 

> and want to know if we could all work together to polish up a stateful DOM using the Q-XML stateless XML parser? 

 

DOM = stateful, a document object model

 

My original request then was to break statelessness.  I'm working on a desktop app with small info-sets, not on a scalable component...statefulness is just what I'm after.  The Object Model of state simplifiies my life, and makes my projects feasible from a one-man factory with sore arms.  While holding state, the object holding it actually becomes the applicaiton object model. Therefore, I plan on the trainers that are to be using my app to literally create my application as need be by editing an XML file.  I've worked on this for a long time.  They can literally code into my application with an XML editor.  These people will be hospital staff hand picked by our staff to learn how to edit XML to create state, an extensible application object model.  They will be able to go one step further, and that is to define behaviors, and supply mapping to bind behaviors to the respective state models.  All from XML, and all depending upon state.  DOM means DOM.  And I literally mean Document, as in medical documents, voice enabled ones that carry grammar definitions in their state, and also bind to voice-recognition events from SAPI wrappers in their state objects.

 

Yes, I wanted to display my DOM for public preview.  One simple elegant node object finding tags with an elegant parser.

 

The QXML module is to my project a real nice way to locate tag elements for constructing an object model.  I'm only needing to replace my tag-finder class, and QXML works OK-fine.  I'm well aware of its useage of longs to mark and manipulate tags by postions, and that is my initial excitement, as that is the way my tag finder works.  Not as elegantly, though. I'm excited that you turned us on, Will, to such a polished Deterministic parser.

 

I'm eager to follow your stateless progress, Will.  Please keep us informed.  Sorry I'm not able to spread both directions.  No doubt statelessness may generate more subscriber interest.

 

Project request closed.

 

By the way, should I open the list up to non-moderated membership, so anyone can join without my approval?  I wonder if moderation will stop some from joining if they find us the yahoo groups listing.  Opinions?

 

Cheers,

DonE

 

P.S.  If I sound a little perturbed, I'm sorry...I just found out that my boss sold a contract for deliverable software, $500,000.00 per year license to the hospital, done deal, signed, my deliverable software, with a 90 day window for production! Not a good day to tell me I need to look deeper.  And I truly don't know what it is they want us to do.  I only know we will be receiving an HL7 Message Record, and we'll populate it with dictation that relies on unknown data-merges from unkown data locations. I don't know if I'll be receiving files or IP data streams. 90 days/half a million bucks per annum. My vapor ware. It may be a good day for me to call a head hunter.

 

I'll have to delete this message from the archives before more subscribers arrive. It reeks of bad attitude.  But it's off my chest. Thanks for listening fellows. 

-----Original Message-----
From: Rawls, William [mailto:wrawls@...]
Sent: Friday, March 23, 2001 3:09 PM
To: 'VBPros@yahoogroups.com'
Subject: RE: [VBPros] XML Parser Project Proposal

 

I would offer that perhaps everyone needs to take a much longer and closer look at QXML. It already exhibites all the qualities discussed in this document. The simplicity of the code allows it to automatically adjust. It's concept of a "Node" is a Long, an index into the string. Functions allow the exploration of siblings, children, attributes, text, xml, and XPath.

 

After working with QXML intimately for a few weeks, I have not found anything it cannot do. I was thinking we would add functions to QXML to extend the usefulness while maintaining statelessness. We could break the stateless nature of the class, which I've thought of doing, but it's statelessness offers many scalability advantages.

 

I would ask everyone to explore QXML with large and small XML files with an example that employs every function exposed by QXML. I think you'll find it's performance is very acceptable.

 

I have thought of recoding QXML in VC++ as a Thread-neutral stateless class to allow for higher degrees of scalability as a server side component.

 

 

Example of some capabilities (including XPath) in QXML:

 

Dim XMLDoc As String
Dim X As QXML.Parse
Set X = New QXML.Parse

 

XMLDoc = "<X>" & _
     "<Y D=1 E=""&gt;Something&lt;"" F=""William M. Rawls"">" & _

     "<Z><A>1</A><B>2</B><C>3</C></Z><Z><A>1</A><B>2</B>" & _
     "<C>3</C></Z><Z><A>1</A><B>2</B><C>3</C></Z></Y></X>"

 

Debug.Print X.nodeXMLPath(XMLDoc, "X\Y\Z")
Debug.Print X.nodeTextPath(XMLDoc, "X\Y\Z\B")
Debug.Print X.xmlAttribute(X.xmlNode("Y", XMLDoc), "F", XMLDoc)

 

Set X = Nothing

-----Original Message-----
From: DonEMitchell [mailto:donemitchell@...]
Sent: Thursday, March 22, 2001 11:35 PM
To: VBPros@yahoogroups.com
Subject: RE: [VBPros] XML Parser Project Proposal

Good Morning,

 ----Original Message-----
From: Tore [mailto:tbostrup@...]
Sent: Friday, March 23, 2001 12:56 AM
To: VBPros@yahoogroups.com
Subject: Re: [VBPros] XML Parser Project Proposal

That's more in my kind of language :->.

I did understand from the beginning of your original post that you were
talking about creating another DOM parser, but after that, you lost me
somewhat.

I've seen the MSXML DOM in action on a 3.6 Mb XML file trying to be loaded
into IE5.  The process was killed after 5-10 minutes and about 120 Mb of
memory consumed.

I have two thoughts.  First, there is little need to maintain all the
information about each node in memory - especially if each node carries a
significant amount of information.  Parsing the file, only critical
information could be "cached", along with a pointer to the node in the file.
This way, the additional information would only be loaded when used - kind
of a Just-In-Time DOM implementation. 

 

I agree there, Tore, regarding just-in-time caching, but if I understand you correctly, it means that any changes written back to the file would require re-parsing the file past the point where the changes were made.  I think this addresses the "reason to be" of the "node", which is to represent the structure in an XML file as an object instance within a programming process space.

 

Second, I'm not sure I like the use of a collection indexed by the full path
of each node, thereby representing a linked structure in a flat media.  Why
not just maintain node links, where each node has a collection of children.
This should allow for more flexibility and take all the trouble out of
moving the branch of nodes (just modify or add link to the root of the
subtree, and everything else follows).

 Yes, regarding linked structure, node links, this is what a node is. A node contains node children.  The linked tree would stay in memory until destroyed.  Having an additional separate collection enables the use of getNode(XPath) which is I suppose a convenience to the alternative of coding out a path of...

 

     Node("ChildName").Node("GrandchildName").Node("GreatGrandchildName...

 

An example XPath I used in Vox2Data:

Private Const cXPathVox2DataSplit = "\UserProfile\Applications\Vox2Data\Metrics\frmVox2Data\Split"

 

The above constant would represent a lot of typing using explicit Node.NodeChild qualified property access. 

 

And perhaps a more important use of a collection of references to nodes is that access to any particular node is accomplished with a string vs. hard-coding into the node tree.  Strings can be modified at runtime whereas hard-coding is hard-coding.

 

My thoughts are that a project to create an XML parser should attempt to be XML 1.0 compliant, and I think what I've discussed is in line with that in this email message.  However, full-compliance may be a bit much.  Also, I think that extra functionality could be added to a partially-compliant node that would make it attractive also, like a just-in-time loading of nodes, which I did with a version I made.  I loaded Romeo and Juliet in XML format, a quarter-megabyte file, only into the top node, all children nodes were non-existent.  The children nodes were instanced on-demand.  I did this technique with an attribute-of-convention I called "LoadChildren".  That technique wouldn't work, however, for string-based node-references (getNode(XPath)) without some more look-ahead-instancing based on the string XPath, but that too could be do-able me thinks.

 

What is the modeling info you don't think you can code in a structured set
of name, content and attributes?

I suppose anything could be modeled with node content alone, not even using attributes.  To avoid bogging down a modeling design, I envisioned more convenience qualities per node, namely, Semantics (keywords used in commands) and Pragmatics (linking semantics to syntactic nodes per algorithmic function), to complete Webster's Dictionary definition of Semiotics as one-node representing a Semiotic quanta.  With these added qualities of my convention, what I didn't explain is that a completely de-coupled command interpreter could manipulate a node-tree to become the silicon-slave, the beast-of-burden, the work-horse of information society.  By adding extra qualities of convention, the convention would enliven information mobility and threshold-based-raising of algorithms. Those things are all best achieved when and if a basic XML node is created, though, for sure, huh? :)

 

Cheers,

DonE.

 

I haven't done much with XML (read a little, helped out a collegue with a
sample, and taken advantage of persisting a recordset as XML in order to
create a filtered clone), so I'm all for doing something with it.  So far,
the aforementioned is the only real use I have had (or seen) for it.

I like excitable, but I reserve the right to ask stupid questions when I get
lost... :->

Good night,
Tore.


----- Original Message -----
From: "DonEMitchell" <donemitchell@...>
To: <VBPros@yahoogroups.com>
Sent: Thursday, March 22, 2001 1:02 PM
Subject: [VBPros] XML Parser Project Proposal


> Ah... well sure, Tore.
>
> Like the MS DOM, or Document Object Model, I was envisioning a project
> whereby we document and fabricate a few simple functions that glue
together
> to form a node.  The node will self-replicate more child nodes as it
> discovers child XML elements.  This tree of linked nodes that forms is
then
> a memory instance of linked nodes that model the structure of the XML
string
> given to the top node.
>
> The nifty thing about the node is to build a collection as the nodes are
> constructed holding references to all of the nodes in the tree structure,
> keyed by path within the node-tree.  The collection is a flat structure,
an
> flat-ordinal list of all of the nodes discovered from the top XML element
> down into the structure.  This allows a fellow to programmatically access
> any value contained in the XML structure by knowing it's path.  Having a
> structure of linked nodes also allows for the reverse creation of an XML
> string form the tree of linked nodes.  That creates a component that can
> load structured XML, allow programmatic modification via path-qualified
> access, and reconstruction of an XML string from the modified node tree.
>
> How'd I do this time?
>
> Here's another link with more explanations into another companies
high-speed
> parser...
> http://www-106.ibm.com/developerworks/xml/library/x-elexml/index.html?
This
> link came in this morning from dev-xml@yahoogroups.com.  These guys are
> talking about millisecond access to structured XML.  Looks mondo.
>
> The QXML module is essentially a group of VB string search procedures that
> provides a number of  ways to discover the position of the XML elements
> starting from the top XML string.  As the QXML searches discover the
> starting and ending positions of the XML element tags, these positions are
> used to lift the names/values/attributes from the XML and instance classes
> to hold them.  The class is the node.  The node has parent properties that
> is linked to the next-higher node that contains the discovered node.  A
> recursive technique from the top XML element is able to build the entire
> tree by adding child nodes to other nodes as discovered.  It's really
quite
> a simple process, unless one tries to implement complete XML 1.0
compliance,
> which gets more tricky.  Microsoft's implementation is big and slow.
>
> Lot's of uses for a DOM abound.  A document engine can be implemented
> straight up with the addition of a template-per-node (a pattern) and some
> string substitution. The patterns and values in the context of the
node-tree
> can be sent to an XSL processor for any of today's XML/XSLT bells and
> whistles (but much faster, as the node stores it's template, and a search
> for a pattern-match is bypassed).  A structured-extensible persistence
> mechanism can be built from the DOM more easily.  I suppose the focus of
my
> proposed project would be the construction of an XML parser with little
more
> than a test harness for starts.
>
> After I lapsed into my 'dreamy stuff' in my discourse, I was referring
about
> ways that a node could be extended to contain more modeling information
than
> the simple XML qualities of name, content and attributes.
>
> I'm not sure where you stand with XML, Tore, but the extra stuff I was
> referring to is beyond XML and came from my two years of attempting to
punch
> through on a way to model meaning from a spoken text string in a running
> context.
>
> Sorry to have 'lapsed' on ya :)  There's a word I found when looking up
the
> definition of semantic. The word is Semiotics, which is the science of
signs
> that represent meaning.  That science is composed of three branches, a)
> Syntactics, b) Semantics, and c) Pragmatics.  This is from Webster's
> dictionary. While being immersed in a project with semantic lists of
command
> phrases that needed to link or map to an execution entry point
> (voice-automation) I felt drawn uncontrollably to this area.  So if I get
> off on a lark from time to time pay no thought :)  Uhmmm.  If you were
> wondering what my dreamy stuff 'does', it would primarily be a discerning
> interpreter that could invoke algorithms within context of a running
dialog
> of commands.  I guess I was really out there without explaining my techno
> babble.  I'm excitable :)
>
> Cheers,
> DonE

To unsubscribe from this group, send an email to:
VBPros-unsubscribe@yahoogroups.com



Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.

To unsubscribe from this group, send an email to:
VBPros-unsubscribe@yahoogroups.com



Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.

To unsubscribe from this group, send an email to:
VBPros-unsubscribe@yahoogroups.com



Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.

I think we're on the same sheet.  I was just sharing an observation with
regards to the stateful vs. stateless discussion, that the "stateless" was
not something that really contributes to solving an application function of
purpose.

And yes, with a "stateless" machine, you need the interactor to supply state
information.

Regards,
Tore.


----- Original Message -----
From: "DonEMitchell" <donemitchell@...>
To: <VBPros@yahoogroups.com>
Sent: Saturday, March 24, 2001 11:53 PM
Subject: RE: [VBPros] Fun with Trees - Water cooler geek speak


> Full circle...
>   -----Original Message-----
>   From: Tore [mailto:tbostrup@...]
>   Sent: Saturday, March 24, 2001 11:31 PM
>   To: VBPros@yahoogroups.com
>   Subject: Re: [VBPros] Fun with Trees - Water cooler geek speak
>
>
>   "Statelessness" is an artifact of computer technology - a requirement
for
>   scalability of applications running on a webfarm.  It is not an artifact
> of
>   actual function to be resolved.
>
>   As a matter of fact - the statelessness is dealt with in various ways to
>   create stateful solutions with a stateless machine...
>
>   Well put!
>
>   To create a stateful DOM with the stateless Q-XML module.  That was my
> original project proposal.  Are we on the same sheet of music yet?
>
>   Perhaps better would be to have a solution-tree' where various
techniques
> of holding state were implemented  from various tiers of a program.
>
>   Regards,
>   Tore.
>
>   Do you understand what I mean by 'holding state'?  For instance, how can
a
> programmer do a stateless query for a particular node if the name, Id,
> whatever, is not held on memory somewhere?
>
>   Here is an example of a completely stateless artifact...
>
>   msgbox
>
eFromPath(
>           inputbox( "What's the NodePath"), _
>           inputbox("Type in your entire XMLDocument ")
>
>
>   In this example, the state is held only in the user's mind, and he types it
in when prompted.
>
>   I have
been considering un-persisted data to be data-of-state when loaded into
memory.  Is this improper use of the term?  Call it solution-state?
Project-State? In-memory-State?
>
>   I feel that we have been confusing terms and techniques through the
history of this discussion.  And I'm guilty of trying to explain a new
paradigm without proper introduction of terms, I'm sure.
>
>   Comments anyone/everyone?
>
>   Cheers,
>   DonE
>
>
>

Maybe you have found some insight that I am missing.  However, in the end,
the result has to be a replacement.  You replace a piece of information with
another representation - otherwise it wouldn't represent the same
information.  Whether the representation is in form of block substitutions
or formulistic expressions, it still boils down to the least unit that can
be represented, and the number of distinct representations.  If the
representations are not distinct, you have lossy compression, which may be
OK for some images, but not for true information - the simplest example of
which is text.

If you are thinking of using a single or preset number of functions, the
LARGEST value is NOT what determines how many distinct pieces of information
you can represent.  That is governed by how many distinct values you can
represent with that function, given the input it will receive.

For instance:

1 * n can represent just as many values as 2 * n, although 2*n is larger if
n > 0.

n! can also just represent n distinct values.

If you are thinking of generating formulas dynamically based on how to
represent an answer of varying patterns and lengths, I guess there is a
possibility that you can find an interesting approach.

And please don't let me stop you... :->  ... or as another one of my
favorite quotes:  "Don't let what you cannot do interfere with what you can
do!"

Regards,
Tore.


----- Original Message -----
From: "Shawn K. Hall" <vb@...>
To: <VBPros@yahoogroups.com>
Sent: Sunday, March 25, 2001 5:04 AM
Subject: RE: [VBPros] 'long' math


> Hi Tore,
>
> > Sorry to disappoint you Shawn, but whoever told you 'it
> > is impossible to have a file smaller than the least
> > unique value in the compression table' (referring to
> > the maximum compression ratio) was correct.
>
> > Just think about it.  If you have "a few characters" the
> > number of different values are limited to the largest
> > number that can be represented by that number of bits,
> > i.e. 2^n.  So with 1 "character" (byte), you can represent
> > 2^8 = 256 DIFFERENT values.  Regardless of what they are.
> > Multiplying a value by itself doesn't help - you still
> > only get the same NUMBER of values, although the largest
> > value will be greater.
>
> It's the *largest* value that's important. The rest are
> inconsequential, as it were. When someone tells you to find out the
> value of "50!" would you say that the resulting value could not be
> larger than the length of the initial elements? Of course not. That
> would be insane. 50! is some 65 digits after it's computed. There
> are, of course, other means of representing the resulting value than
> just the value itself. That's one of the reasons we have mathematics
> in the first place. To overlook such a powerful tool in light of
> compression is outrageous. I simply can't accept that just because a
> number is 'big' that it cannot be reduced and represented by it's
> formulaic factors.
>
>
> > The whole compression issue is based on using the minimal
> > length (in bits) representation of a common sequence of
> > bits.  The more common, the shorter the representation.
>
> If you're using a replaceable scheme of compression (the most common
> type). The inherent problem with replaceable compression schemes is
> that they rely on values to be repetitive, and they look at 'blocks'
> or segments of data and build a dictionary out of those
> representations. That's fine for *that* system, but would be useless
> in the GAB system I've got up there (I call it "Gig' A Byte" - GAB -
> pronounced Gabby). The assertion you propose is that it is a
> replaceable compression system, and it's not. I full well agree with
> you that in terms of a replaceable compression system it would be
> impossible to reduce the value beyond the smallest block. But in a
> formulaic compression system that would not be the case.
>
>
> > But whatever approach you use, there is a theoretical
> > limit to the compression.  Images compress much better
> > than text because continuity can be taken into
> > consideration.  The next big breakthrough in text
> > compression (if any) may attack linguistic redundancy,
> > which can be as high as 75% in the English language.
>
> What you are talking about here is just another form of replacement
> compression. GAB would enable you to ignore the content of the
> document and deconstruct it based on its numerical representation.
> It's a big number. Prolly very big. But it can be reduced. Otherwise
> there'd be no need to have mathematics in the first place.
>
> The example the document described (I don't have a link handy and I
> don't remember what site it was on, though I think it was regarding
> GZip) told of how 'ludicrous' it would be to assume that you could
> represent more 'data' with smaller representations because if you
> tried you would end up with less "representations" than "resultant
> values". Following the logic of the example, in order to represent
> the number three, you could only use smaller numbers: so you could
> only use a one or a two. Fine - but then how would you represent a
> one or a two? If you used the two to represent three then you'd have
> a one there to represent itself and the two, which would be
> impossible to interpret. I agree with the sentiment, if not the
> conclusion. Yes, assuming that the values were intended to be
> reduced to absolutes - you would find it impossible to reduce values
> past a certain scheme. But to throw an arbitrary shell around the
> thing as though the limitation was in the values themselves is
> wrong.
>
> The problem is within the interpretation. In order to interpret
> compress or deflate or any of the other 'replacement' schemes you
> have to build a data dictionary to consult once the data has been
> compressed. The theory is that the data dictionary makes the
> compression system absolute, as though that alone made it
> transportable.
>
> Mathematics is absolute. There is NOTHING on this earth any more
> absolute than math. By ignoring the effectiveness and universality
> of mathematics by NOT using it as the compression scheme they've
> made a major faux pas. You can render a formula as powerful as you
> want using common characters, simple ASCII. And the resulting values
> would blow your mind. x^n, x!, pi - just a couple values that could
> quickly increase the significance of a number. And if that
> significance could be harnessed by a mathematical system *capable*
> of tracking every digit from the least significant to the most - you
> could reasonably represent random data via a significantly smaller
> formula. Consideration would have to be made to accomplish tweaking
> the data for specific value changes. Perhaps a system to stretch a
> value several times across the sum toggling digits or providing for
> those minute changes that would affect the difference between a
> butterfly wings and a chihuahua (btw, that last statement made
> perfect sense before I saw it in the composition window).
>
> Anyway, my point in all my rambling is only that simply because
> representing something with a smaller value is the goal, does not
> mean that it has to be a replaceable system. The entire intent of
> this method would be to avoid a replaceable system, in fact. Using
> the mathematics that everyone is already familiar with to accomplish
> even greater compression.
>
> Regards,
>
> Shawn K. Hall
> http://ReliableAnswers.com
>
>
>
>
> To unsubscribe from this group, send an email to:
> VBPros-unsubscribe@yahoogroups.com
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>

-----Original Message-----
From: Tore [mailto:tbostrup@...]
Sent: Saturday, March 24, 2001 11:31 PM
To: VBPros@yahoogroups.com
Subject: Re: [VBPros] Fun with Trees - Water cooler geek speak

"Statelessness" is an artifact of computer technology - a requirement for
scalability of applications running on a webfarm.  It is not an artifact of
actual function to be resolved.

As a matter of fact - the statelessness is dealt with in various ways to
create stateful solutions with a stateless machine... 

 

Well put!

 

To create a stateful DOM with the stateless Q-XML module.  That was my original project proposal.  Are we on the same sheet of music yet?
 

Perhaps better would be to have a solution-tree' where various techniques of holding state were implemented  from various tiers of a program.

 

Regards,
Tore. 

 

Do you understand what I mean by 'holding state'?  For instance, how can a programmer do a stateless query for a particular node if the name, Id, whatever, is not held on memory somewhere?

 

Here is an example of a completely stateless artifact...

 

msgbox getNodeFromPath(

        inputbox( "What's the NodePath"), _

        inputbox("Type in your entire XMLDocument ")

 

 

In this example, the state is held only in the user's mind, and he types it in when prompted.

 

I have been considering un-persisted data to be data-of-state when loaded into memory.  Is this improper use of the term?  Call it solution-state? Project-State? In-memory-State?

 

I feel that we have been confusing terms and techniques through the history of this discussion.  And I'm guilty of trying to explain a new paradigm without proper introduction of terms, I'm sure.

 

Comments anyone/everyone?

 

Cheers,

DonE 

Hi Tore,

> Sorry to disappoint you Shawn, but whoever told you 'it
> is impossible to have a file smaller than the least
> unique value in the compression table' (referring to
> the maximum compression ratio) was correct.

> Just think about it.  If you have "a few characters" the
> number of different values are limited to the largest
> number that can be represented by that number of bits,
> i.e. 2^n.  So with 1 "character" (byte), you can represent
> 2^8 = 256 DIFFERENT values.  Regardless of what they are.
> Multiplying a value by itself doesn't help - you still
> only get the same NUMBER of values, although the largest
> value will be greater.

It's the *largest* value that's important. The rest are
inconsequential, as it were. When someone tells you to find out the
value of "50!" would you say that the resulting value could not be
larger than the length of the initial elements? Of course not. That
would be insane. 50! is some 65 digits after it's computed. There
are, of course, other means of representing the resulting value than
just the value itself. That's one of the reasons we have mathematics
in the first place. To overlook such a powerful tool in light of
compression is outrageous. I simply can't accept that just because a
number is 'big' that it cannot be reduced and represented by it's
formulaic factors.


> The whole compression issue is based on using the minimal
> length (in bits) representation of a common sequence of
> bits.  The more common, the shorter the representation.

If you're using a replaceable scheme of compression (the most common
type). The inherent problem with replaceable compression schemes is
that they rely on values to be repetitive, and they look at 'blocks'
or segments of data and build a dictionary out of those
representations. That's fine for *that* system, but would be useless
in the GAB system I've got up there (I call it "Gig' A Byte" - GAB -
pronounced Gabby). The assertion you propose is that it is a
replaceable compression system, and it's not. I full well agree with
you that in terms of a replaceable compression system it would be
impossible to reduce the value beyond the smallest block. But in a
formulaic compression system that would not be the case.


> But whatever approach you use, there is a theoretical
> limit to the compression.  Images compress much better
> than text because continuity can be taken into
> consideration.  The next big breakthrough in text
> compression (if any) may attack linguistic redundancy,
> which can be as high as 75% in the English language.

What you are talking about here is just another form of replacement
compression. GAB would enable you to ignore the content of the
document and deconstruct it based on its numerical representation.
It's a big number. Prolly very big. But it can be reduced. Otherwise
there'd be no need to have mathematics in the first place.

The example the document described (I don't have a link handy and I
don't remember what site it was on, though I think it was regarding
GZip) told of how 'ludicrous' it would be to assume that you could
represent more 'data' with smaller representations because if you
tried you would end up with less "representations" than "resultant
values". Following the logic of the example, in order to represent
the number three, you could only use smaller numbers: so you could
only use a one or a two. Fine - but then how would you represent a
one or a two? If you used the two to represent three then you'd have
a one there to represent itself and the two, which would be
impossible to interpret. I agree with the sentiment, if not the
conclusion. Yes, assuming that the values were intended to be
reduced to absolutes - you would find it impossible to reduce values
past a certain scheme. But to throw an arbitrary shell around the
thing as though the limitation was in the values themselves is
wrong.

The problem is within the interpretation. In order to interpret
compress or deflate or any of the other 'replacement' schemes you
have to build a data dictionary to consult once the data has been
compressed. The theory is that the data dictionary makes the
compression system absolute, as though that alone made it
transportable.

Mathematics is absolute. There is NOTHING on this earth any more
absolute than math. By ignoring the effectiveness and universality
of mathematics by NOT using it as the compression scheme they've
made a major faux pas. You can render a formula as powerful as you
want using common characters, simple ASCII. And the resulting values
would blow your mind. x^n, x!, pi - just a couple values that could
quickly increase the significance of a number. And if that
significance could be harnessed by a mathematical system *capable*
of tracking every digit from the least significant to the most - you
could reasonably represent random data via a significantly smaller
formula. Consideration would have to be made to accomplish tweaking
the data for specific value changes. Perhaps a system to stretch a
value several times across the sum toggling digits or providing for
those minute changes that would affect the difference between a
butterfly wings and a chihuahua (btw, that last statement made
perfect sense before I saw it in the composition window).

Anyway, my point in all my rambling is only that simply because
representing something with a smaller value is the goal, does not
mean that it has to be a replaceable system. The entire intent of
this method would be to avoid a replaceable system, in fact. Using
the mathematics that everyone is already familiar with to accomplish
even greater compression.

Regards,

Shawn K. Hall
http://ReliableAnswers.com

> Still, do you mean *billion* of digits?

Billions of digits.

Imagine, for example a single file of one meg. That file, converted
to a number would be 'large'. Each byte represents a max value of
256 alone. Do the binary toggle dealy thing and you quickly get into
the decimal digits to represent the enormity of size for the file.
1024 k = 1048576 b. Do the math on the byte values and you can get a
rather large summary number. Each byte effectively doubles the
number of potential values. Granted, it's much smaller effectively
than a decimal system where each character would represent an
increase of ten fold, but still pretty large.

With only 32 bits (one long integer) you get into values of 2
billion or so. How big of a number can you achieve with 8388608 bits
(the number of bits in a meg). It only doubles the summary value
each time but that's still gonna be on heck of a number.

Regards,

Shawn K. Hall
http://ReliableAnswers.com

I wasn't really thinking about a multiuser type of issue.  However, this is
a classic locking issue - well known to databases.  If you open the file for
read-only, you won't make any changes, and anybody can look at it
simultaneously.  If you open it for update, you can take two approaches.
Either, you block others from opening it, or you allow concurrent read
access.  Concurrent reads will not see your changes until you commit.  At
that point, the concurrent reads will have to be notified* of changes
applied, and in worst case, they would have to re-parse the file.  Of
course, if the files are served up by a server process, the server would
hold the physical pointers, and the clients would use a virtual pointer.  If
it is to be a multi-user accessed updateable mechanism - it would have to go
through a server (arbitrator).

Active notification would send a message to the user, passive notification
could in effect take place when a node is not found where it is expected, or
the date/time stamp of the file could be verified at each access.

Regards,
Tore.

----- Original Message -----
From: "Shawn K. Hall" <vb@...>
To: <VBPros@yahoogroups.com>
Sent: Saturday, March 24, 2001 4:20 PM
Subject: RE: [VBPros] XML Parser Project Proposal


> Hello all,
>
> > It depends.  You could do "dynamic" caching where you
> > kept all updates in memory.  When saving, you'd either
> > load all the info into memory, or use a temp file with
> > a copy of the content as the source, and write back to
> > the original. Or you could establish the temp file when
> > opening or parsing the file, and use it to save
> > updates.  Since the data is found through a pointer
> > in memory, you can save updates to the end of the file.
> > The structure kept in memory would represent the
> > "proper" order, and the temp file would be somewhat
> > jumbled.
>
>
> I had the exact same thoughts for preserving 'speed' for this
> project when I first read the discussion between Will & Don. Now I
> don't think it'd work that well.
>
> What would happen if you opened another request to read the file?
> Since the chanes were not written yet they would be missed, and if a
> minor change was made to that then saved the other process would be
> essentially screwed.
>
> I guess you could force it into a single instancing class interface
> for each file, but that would risk many-reads-many-writes within the
> same memory simultaneously. That too, would be a 'bad' thing.
>
> I'll have to think about this for a while. I'm really at a loss for
> how you could preserve speed and functionality at the same time in a
> multi-instancing environment.
>
> Regards,
>
> Shawn K. Hall
> http://ReliableAnswers.com
>
>
>
>
> To unsubscribe from this group, send an email to:
> VBPros-unsubscribe@yahoogroups.com
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>

Sorry to disappoint you Shawn, but whoever told you 'it is impossible to
have a file smaller than the least unique value in the compression table'
(referring to the maximum compression ratio) was correct.

Just think about it.  If you have "a few characters" the number of different
values are limited to the largest number that can be represented by that
number of bits, i.e. 2^n.  So with 1 "character" (byte), you can represent
2^8 = 256 DIFFERENT values.  Regardless of what they are.  Multiplying a
value by itself doesn't help - you still only get the same NUMBER of values,
although the largest value will be greater.

The whole compression issue is based on using the minimal length (in bits)
representation of a common sequence of bits.  The more common, the shorter
the representation.  You can choose the length of sequences you wish to
compress - they could even be variable length.  The most common sequence can
be designated by a single bit.  The second most common sequence can be
replaced by two bits.  The reason you can't have two one-bit values, is that
one of them will be used to signify end-of-value, since values will be
variable length.  So if we say 0 is the "end-of'value", the most common
sequence will be represented by a 0 by itself.  The third most common value
would be represented by three bits, etc.

This approach can be modified by using clusters of bits instead of single
bits - the frequency distribution of sequences will determine the best
encoding cluster length.  In a two-bit cluster, let's say 00 indicates the
end-of value, and ia also used for the most common sequence.  The three next
most common sequences can be represented in 4 characters (01 00, 10 00, 11
00), etc.  With three bit clusters, you'd have 000 as the end-of-value and
representation of the most common the 7 next most values represented in 6
bits.

Next, we can use variable length clusters, specifying only the unique
end-of-value combination, for instance 00.  That allows us to use any bit
string ending with a 100 and not having two consecutive 0's in it.

Finally, you may modify the technique to use not a single end-of-string, but
multiple, depending on the sequence represented.  I wouldn't be quite
qualified to describe this, other than you's have certain patterns that
would be recognized as terminating patterns, based on the preceding pattern.
This may not always result in additional savings.

Another approach altogether could use a positional method.  By breaking the
original string into blocks, and using a value + bitarray representation of
the value's occurrance, you could get very dense representation of high
frequency sequences, but very inefficient representation of low frequency
values.

Additional linguistic rules of combinations may be taken into consideration
to improve compression.

But whatever approach you use, there is a theoretical limit to the
compression.  Images compress much better than text because continuity can
be taken into consideration.  The next big breakthrough in text compression
(if any) may attack linguistic redundancy, which can be as high as 75% in
the English language.

Regards and HTH,
Tore.

----- Original Message -----
From: "Shawn K. Hall" <vb@...>
To: <VBPros@yahoogroups.com>
Sent: Saturday, March 24, 2001 3:22 PM
Subject: RE: [VBPros] 'long' math


> Thanks Don,
>
> I'll look into assembly routines to accomplish it.
>
> > Do you really mean *billions* of digits?  You
> > may need a few terabyte of memory (well, at
> > least several hundred computer cycles per digit
> > for a serialized computation process) to store
> > the intermediate stuff.  Lots of memory and/or
> > time. Sounds like a task for a quantum computer.
>
> There's gotta be a better way. The task is designed (in theory) for
> any computer that can run the mathematical formulae, ideally any
> computer on the web.
>
> Here's the deal:
> I read through the specs and algorithms for compress, deflate, and a
> few others and realized that they were wasting a lot of space by
> abandoning the 'known' compression tables after so many registers.
> Apparently (now that I hear it'll take a terabyte of memory <g>),
> that's why.
>
> The assertions made in the dissertations accompanying these specs
> states very plainly that 'it is impossible to have a file smaller
> than the least unique value in the compression table' (os something
> like that).
>
> I disagree. This logic is flawed in that it assumes that the only
> representation of data is binary. If you represent a file of x bytes
> as a number - it'll be a big number - but it'll be dissectable into
> representative elements.
>
> The way I see it, tell me I'm wrong here, if you could reverse
> engineer an equation to recreate the value from a formula then you
> could essentially store literally gigabytes or even terabytes in
> only a few characters. Using a simple ascii notation you could
> reduce the file sizes to something *very* manageable (compared to
> terabytes) with only a few keystrokes. A 10 digit number to it's own
> power could easily recreate a file in the terabytes range.
>
> There's been a 'problem' with having patented compression schemes
> (like the one used in gif) which are really just a waste of time. If
> I can accomplish even a proof of concept demonstration of this
> method (1) bandwidth wouldn't matter anymore, (2) disk space
> wouldn't matter anymore, (3) information could easily be exchanged
> *without* a computer (imagine spys passing notes on paper again!),
> (4) scalable removeable media could literally fit on a pinhead, and
> (5) I'll be rich. <g>
>
> This all assumes that it's possible to do it though. I guess there
> are ways around performing strict mathematics on the values and
> appending values or OR/XOR'ing values could be done to 'fake' long
> math. The thing is that I know it's possible. I can see the
> equations in my head and I've got the numbers working with only a 16
> bit processor up there - but putting it into the code-pane is a
> little more daunting.
>
> > Ken Young
>
> Thanks, I checked out the site but there was no email address
> listed. I'll just call him next week. I've got some time on my hands
> now and would really like to be able to at least get started on this
> theory of mine.
>
> Is anyone interested in helping me with this?
> Am I overlooking something that everyone that went to school to
> learn computers already knows (I never had the opportunity to learn
> this stuff outside of 'life experience'). This is the first time
> I've brought this concept to light outside of my 'Ideas' database -
> so please don't share it with anyone. I don't think I'm greedy, per
> se, but I'd really like to at least have my name on the concept if
> it survives.
>
> Regards,
>
> Shawn
> http://ReliableAnswers.com
>
>
>
>
> To unsubscribe from this group, send an email to:
> VBPros-unsubscribe@yahoogroups.com
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>

"Statelessness" is an artifact of computer technology - a requirement for
scalability of applications running on a webfarm.  It is not an artifact of
actual function to be resolved.

As a matter of fact - the statelessness is dealt with in various ways to
create stateful solutions with a stateless machine...

Regards,
Tore.

----- Original Message -----
From: "DonEMitchell" <donemitchell@...>
To: "VBPros@yahoogroups. com" <VBPros@yahoogroups.com>
Sent: Saturday, March 24, 2001 3:38 AM
Subject: [VBPros] Fun with Trees - Water cooler geek speak


> Hi Pros,
>
> I feel a bit more professional today.  I've come to terms with my boss
> inside my own head.  I'd like to have all the money in a pile that has
been
> wasted, and roast him over the bon fire it could make.
>
> But he's a really nice guy, that's just the misdirected anxt that occurs
> when one is a mushroom, kept in the dark and fed B.S., huh? Or so I'll
> purport.
>
> I removed all the over-excited messages that built up to my manic release.
>
> Say, Will, may I use the file folder you started named XML where your QXML
> wrapper is? Or had you thought of using that for your personal library.
Let
> me know.  I'm attempting courtesy. I'd like to upload to the files section
a
> few test XML files I have gathered, like some of Shakespeare.  It should
> fall under an XML category of file folder, so I ask.
>
> Tore, Shawn, did you know you can create folders and upload to the VBPros
> files section?  I also turned on unmoderated membership.  If that is not a
> comfortable thing, like with your brain-child discussion, Shawn, then I
can
> shut it down to approved-membership.
>
> I invited the Seattle VBDeveloper's Assoc. president to join us, and/or
> forward the invite to the officers/membership. (So far no response.)
>
> And also, a new VBPros bylaw, anything that has the word 'Water cooler' in
> the subject line will be deleted from the archives after a while.  This
way,
> we can communicate without puttin' a heap of stuff in the archives that
may
> not be pertinent months from now.
>
> Additionally, another by-law, we can reply to a message in the appropriate
> thread to keep a thread running, thereby making the 'thread' view a nice
> tool for adding context to the message archive.
>
> Oh!  Fun with Trees!
>
> Me thinks I'm going to work on an example project for the files section.
I
> will be attempting to exemplify how I manipulate state, and how
> state-behaviors can be used in a decoupled manor (oops -- event binding)
> from state.  And I will be using a treeView because it's a well-understood
> example of a state-machine, my kind of state, anyway.  Each treeView node
> crosses two dimensions.  There is the visual tree-based nodes and child
> nodes which exemplify cardinality (a.k.a., relationships) and there is the
> underlying key table which is a flat array, thereby requiring all node
keys
> to be unique.  I will use the path to a node as it's unique key, thereby
> simulating what my DOM (or the MS DOM with NodeList) can do.  Then, I'll
get
> around eventually to presenting a node that will discover persisted state
> (XML) which is parsed to an instanced state-model.  From there, perhaps
and
> only if there is group understanding of where my particular and personal
> project is going I will open up to a group project.
>
> Have you'uns noticed I don't know how to communicate technically?  That's
> another symptom from being a one-man coder factory for the better part of
a
> decade.
>
> But 'fun with trees' will use the WalkThrough and the IXMLNotify interface
> to build a node-tree in a TreeView control.  Consider it my
teething-project
> :)  I'd like to have my mark left somewhere, even tooth marks :)  (I
reserve
> the right to change my plans without a moments notice:)
>
> I'm in no shape personally to collaborate.  I'm more in the flame-my-boss
> mode right now, so this will be a nice anchor for my hot-air balloon.
>
> Your a good man, Will Rawls.  Say, Will, interesting QXML treatment of
your
> additional methods to find a node from a Path by separating the node names
> of an sPath argument and drilling down to the last name in a loop.  I want
> you to know that I AM NOT an ASP man.  Perhaps at times an ASS man.  Can
you
> read the apology in that?
>
> So 'fun with trees' may be the first release in a series of example
projects
> that exemplifies that if there is no state in a project, then the state
must
> be held in the user's mind, therefore, that violates the new paradigm of
> User Interfaces that are Inductive (a Microsoft term, IUI, a.k.a., the new
> MS Money interface that doesn't rely on the user to understand what the
> interface is suppose to do before he uses the interface.  If you guys will
> bear with me, I too have a soap-box, and God help me to learn how to
preach
> without damnation. <a moment of silence while all my regrets are pondered>
>
> By no-state-in-a-project, I'm meaning that somewhere in the creation of a
> project state MUST be UNDERSTOOD in the creators mind -- UNLESS the
project
> itself is a tool to allow for the authorship of a new state, in which case
> STATE is CREATED by the user.  And anything we as people create ourselves,
> we as creators inherently and by definition UNDERSTAND.
>
> Hey, Will!  I'm sounding Slice-And-Dicey!, huh?
>
> State is the essence of meaning -- when and if it is conveyed in an
> understood context.  Smoke that in your pipe.  More better said, a storage
> of ordinal items as leaves held in a cardinal relationship as a tree is a
> tree of knowledge, or an essence (model if you will) of meaning.
>
> Just because one separates a UI from a business tier does not mean that
> state can be lost, state is merely stretched aross the round-trips needed
to
> marshall the parameters from the client to the server.  The return result
> may or may not change the state held in the client or held in the server.
So
> we see that the art of state-discovery/state-creation is intertwined
within
> the art of programming.
>
> Take a web page that is very stateless, all parameters are supplied to
some
> CGI or nifty wrapper around a stateless QXML parser, well Gee!  Where does
> the information, uhmm, parameters come from?  Well, if they weren't hard
> coded into the web page, they came form user input. Or I suppose they
could
> have came from some algorithm the user turned loose on the world, in which
> case, the user still understood when to turn loose the algorithm, and that
> knowing what to do or when to invoke an action is the understanding of
> state, a state that describes when behavior is appropriate.
>
> Hey guys, where's all this state-stuff coming from?  Exactly and
> specifically from a study of requirements into a Linguistic User Interface
> (LUI).  Start talking to the Enterprise like Janeway, and write down each
> parameter that is embedded into the command language, the command-speak,
and
> you will come up with a list of things that each fit somewhere into a
> knowledge tree of things understood by the user that are dereferenced from
> memory as we speak.  Yep, we use symbols (words) to represent things we
hold
> in memory that conjure feelings which are used to reckon judgment to
adjust
> behavior.  The symbols dereference to a spot in a tree of relationships.
>
> I'm not talking out of the norm.  A philosopher penned in the 60's that we
> can only understand another's message when we have shared the same
struggle.
> Through effort we build a base of understanding of similar efforts.  Each
> behavior is based on a judgment...oh, wait...too far off topic....back to
> the trees...
>
> I declare that if my boss doesn't throw me a curve ball that I'll submit a
> simple example that uses my DOM (a ubiquitous node) to hold the states
> discovered in context (a.k.a. the walkthrough, which travels states on a
> known course, children before content (the XML 1.0 serialization
technique)
> thereby, knowing the course, the stored state of the tree has meaning.  I
> will need to re-write my DOM first to use the QXML parser.
>
> Tada!  I explained myself a wee-bit.
>
> Your coder in the bush, way up in the corner of the continent ferry close
to
> Seattle,
>
> DonE
>
>
>
> To unsubscribe from this group, send an email to:
> VBPros-unsubscribe@yahoogroups.com
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>

-----Original Message-----
From: Shawn K. Hall [mailto:vb@...]
Sent: Saturday, March 24, 2001 12:22 PM
To: VBPros@yahoogroups.com
Subject: RE: [VBPros] 'long' math

Thanks Don,

I'll look into assembly routines to accomplish it.

> Do you really mean *billions* of digits?  You
> may need a few terabyte of memory (well, at
> least several hundred computer cycles per digit
> for a serialized computation process) to store
> the intermediate stuff.  Lots of memory and/or
> time. Sounds like a task for a quantum computer.

There's gotta be a better way. The task is designed (in theory) for
any computer that can run the mathematical formulae, ideally any
computer on the web.

Here's the deal:
I read through the specs and algorithms for compress, deflate, and a
few others and realized that they were wasting a lot of space by
abandoning the 'known' compression tables after so many registers.
Apparently (now that I hear it'll take a terabyte of memory <g>),
that's why.

The assertions made in the dissertations accompanying these specs
states very plainly that 'it is impossible to have a file smaller
than the least unique value in the compression table' (os something
like that).

I disagree. This logic is flawed in that it assumes that the only
representation of data is binary. If you represent a file of x bytes
as a number - it'll be a big number - but it'll be dissectable into
representative elements.

The way I see it, tell me I'm wrong here, if you could reverse
engineer an equation to recreate the value from a formula then you
could essentially store literally gigabytes or even terabytes in
only a few characters. Using a simple ascii notation you could
reduce the file sizes to something *very* manageable (compared to
terabytes) with only a few keystrokes. A 10 digit number to it's own
power could easily recreate a file in the terabytes range.

There's been a 'problem' with having patented compression schemes
(like the one used in gif) which are really just a waste of time. If
I can accomplish even a proof of concept demonstration of this
method (1) bandwidth wouldn't matter anymore, (2) disk space
wouldn't matter anymore, (3) information could easily be exchanged
*without* a computer (imagine spys passing notes on paper again!),
(4) scalable removeable media could literally fit on a pinhead, and
(5) I'll be rich. <g>

This all assumes that it's possible to do it though. I guess there
are ways around performing strict mathematics on the values and
appending values or OR/XOR'ing values could be done to 'fake' long
math. The thing is that I know it's possible. I can see the
equations in my head and I've got the numbers working with only a 16
bit processor up there - but putting it into the code-pane is a
little more daunting.

> Ken Young

Thanks, I checked out the site but there was no email address
listed. I'll just call him next week. I've got some time on my hands
now and would really like to be able to at least get started on this
theory of mine.

Is anyone interested in helping me with this?
Am I overlooking something that everyone that went to school to
learn computers already knows (I never had the opportunity to learn
this stuff outside of 'life experience'). This is the first time
I've brought this concept to light outside of my 'Ideas' database -
so please don't share it with anyone. I don't think I'm greedy, per
se, but I'd really like to at least have my name on the concept if
it survives.

Regards,

Shawn
http://ReliableAnswers.com




To unsubscribe from this group, send an email to:
VBPros-unsubscribe@yahoogroups.com



Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.

Hello all,

> It depends.  You could do "dynamic" caching where you
> kept all updates in memory.  When saving, you'd either
> load all the info into memory, or use a temp file with
> a copy of the content as the source, and write back to
> the original. Or you could establish the temp file when
> opening or parsing the file, and use it to save
> updates.  Since the data is found through a pointer
> in memory, you can save updates to the end of the file.
> The structure kept in memory would represent the
> "proper" order, and the temp file would be somewhat
> jumbled.


I had the exact same thoughts for preserving 'speed' for this
project when I first read the discussion between Will & Don. Now I
don't think it'd work that well.

What would happen if you opened another request to read the file?
Since the chanes were not written yet they would be missed, and if a
minor change was made to that then saved the other process would be
essentially screwed.

I guess you could force it into a single instancing class interface
for each file, but that would risk many-reads-many-writes within the
same memory simultaneously. That too, would be a 'bad' thing.

I'll have to think about this for a while. I'm really at a loss for
how you could preserve speed and functionality at the same time in a
multi-instancing environment.

Regards,

Shawn K. Hall
http://ReliableAnswers.com

Thanks Don,

I'll look into assembly routines to accomplish it.

> Do you really mean *billions* of digits?  You
> may need a few terabyte of memory (well, at
> least several hundred computer cycles per digit
> for a serialized computation process) to store
> the intermediate stuff.  Lots of memory and/or
> time. Sounds like a task for a quantum computer.

There's gotta be a better way. The task is designed (in theory) for
any computer that can run the mathematical formulae, ideally any
computer on the web.

Here's the deal:
I read through the specs and algorithms for compress, deflate, and a
few others and realized that they were wasting a lot of space by
abandoning the 'known' compression tables after so many registers.
Apparently (now that I hear it'll take a terabyte of memory <g>),
that's why.

The assertions made in the dissertations accompanying these specs
states very plainly that 'it is impossible to have a file smaller
than the least unique value in the compression table' (os something
like that).

I disagree. This logic is flawed in that it assumes that the only
representation of data is binary. If you represent a file of x bytes
as a number - it'll be a big number - but it'll be dissectable into
representative elements.

The way I see it, tell me I'm wrong here, if you could reverse
engineer an equation to recreate the value from a formula then you
could essentially store literally gigabytes or even terabytes in
only a few characters. Using a simple ascii notation you could
reduce the file sizes to something *very* manageable (compared to
terabytes) with only a few keystrokes. A 10 digit number to it's own
power could easily recreate a file in the terabytes range.

There's been a 'problem' with having patented compression schemes
(like the one used in gif) which are really just a waste of time. If
I can accomplish even a proof of concept demonstration of this
method (1) bandwidth wouldn't matter anymore, (2) disk space
wouldn't matter anymore, (3) information could easily be exchanged
*without* a computer (imagine spys passing notes on paper again!),
(4) scalable removeable media could literally fit on a pinhead, and
(5) I'll be rich. <g>

This all assumes that it's possible to do it though. I guess there
are ways around performing strict mathematics on the values and
appending values or OR/XOR'ing values could be done to 'fake' long
math. The thing is that I know it's possible. I can see the
equations in my head and I've got the numbers working with only a 16
bit processor up there - but putting it into the code-pane is a
little more daunting.

> Ken Young

Thanks, I checked out the site but there was no email address
listed. I'll just call him next week. I've got some time on my hands
now and would really like to be able to at least get started on this
theory of mine.

Is anyone interested in helping me with this?
Am I overlooking something that everyone that went to school to
learn computers already knows (I never had the opportunity to learn
this stuff outside of 'life experience'). This is the first time
I've brought this concept to light outside of my 'Ideas' database -
so please don't share it with anyone. I don't think I'm greedy, per
se, but I'd really like to at least have my name on the concept if
it survives.

Regards,

Shawn
http://ReliableAnswers.com

-----Original Message-----
From: Shawn K. Hall [mailto:vb@...]
Sent: Friday, March 23, 2001 6:43 PM
To: VBPros@yahoogroups.com
Subject: RE: [VBPros] Broad spectrum 

 

<snip> 

I've been rolling a project idea around in my head for a few years
now and maybe this is the group to bring it up in. Will send under
seperate cover.

Regards,

Shawn K. Hall
http://ReliableAnswers.com 

 

I hope you do share your ideas with us!  I'd like to provide whatever perspective I can.  I have watched your replies at the Karland list for a long time, and have yet to be unimpressed.  So, I'm eager to hear about your brain-child.

 

Cheers, (more so today)

 

 

-----Original Message-----
From: Shawn K. Hall [mailto:vb@...]
Sent: Friday, March 23, 2001 7:12 PM
To: VBPros@yahoogroups. com
Subject: [VBPros] 'long' math

Hello All,

I've had this idea for quite a while about how to perform operations
on very large integers. Numbers that are beyond the scope of your
imagination in some cases. A few days ago someone pointed me to a
resource that 'works' for smaller 'very large integers' - but it
needs some work.

I tried mulitplying a number into the thousand digit range and it
'died' - to put it politely. And it isn't capable of some of the
features I need to implement. Anyone have any ideas on how I can add
extensive mathematical features using "very" large numbers to a
project?

Things I'll need to be able to do 'long math' including roots,
powers and stuff with numbers in the *billions* of *digits* range.
Is it even possible? How? I don't care if I have to resort to C++ -
as long as I can expose it to a VB object model.

Thanks,

Shawn
http://ReliableAnswers.com




To unsubscribe from this group, send an email to:
VBPros-unsubscribe@yahoogroups.com



Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.

Here is a link from ontologyStream@... by Dr. Paul S. Prueitt that was really fresh...

http://www.bcngroup.org/area3/pprueitt/private/KM_files/frame.htm

The title page...

Sense Making   Environments:
Some preliminary judgments

Dr. Paul S. Prueitt

presented at

 e-Gov KM Conference, Washington D.C.

April 13, 2000

Will/Don,

What is the easiest way for me to get started playing with (and getting some
positive effect from) SND, and how much time will it take before I know if
it fits the way I work?

Tore.


----- Original Message -----
From: "Rawls, William" <wrawls@...>
To: <VBPros@yahoogroups.com>
Sent: Friday, March 23, 2001 6:11 PM
Subject: RE: [VBPros] Broad spectrum


> Yeah. "Sandy Shell edition" which is actually Slice and Dice 1.8.154
> You might need to install the latest Slice and Dice before installing
Shell
> Edition.
>
> Shell edition runs inside VB as an add-in (as always), but will also run
> outside VB as a tray icon. This is very useful for processing clipboard
> operations from other programs like notepad, interdev, VC++, etc.
>
>
> -----Original Message-----
> From: DonEMitchell [mailto:donemitchell@...]
> Sent: Friday, March 23, 2001 12:26 AM
> To: VBPros@yahoogroups.com
> Subject: RE: [VBPros] Broad spectrum
>
>
> Hi Will,
>
> I would love to take you up on voice-automating VB through SliceAndDice.
> But later.  I'm at near maximum bandwidth on the amount of interests I can
> allow before a meltdown.
>
> I've spent a significant amount of time with XML nodes, selfishly enough,
> because I'm at that exact place in my work-a-day for pay existence.  I'm
> looking forward to some SND reprieve, soon.
>
> Are there any new releases since SND 1.8.11 from the web?
>
> Cheers,
> DonE
>
> -----Original Message-----
> From: Rawls, William [mailto:wrawls@...]
> Sent: Thursday, March 22, 2001 12:35 PM
> To: 'VBPros@yahoogroups.com'
> Subject: RE: [VBPros] Broad spectrum
>
>
>  <snip>
>
> Voice enabling VB is as easy as Voice enabling Slice and Dice. Slice and
> Dice hides the details and indeed the need to work with the VB IDE. It
> exposes the VB IDE to you if you want, but is directly intended to hide
the
> complexities from you. All the Sandals (Slice and Dice External DLLs, aka
> add-ins) demonstrate how to manipulate Slice and Dice with simple DLLs and
> forms.
>
> If you give me some working voice enabling code in DLL format I'll show
you
> how easy it can be.
>
> This is chaotic. Cooperation that understands that what I have and what
you
> have work together to create more than either of us have and more than the
> simple sum.
>
>  <snip>
>
>
> Yahoo! Groups Sponsor
>
>
<http://rd.yahoo.com/M=162801.1342140.2934632.1280039/D=egroupmail/S=1700115
>
364:N/A=599120/*http://www.knowledgestorm.com/jump_white.html?c=Yahoo&n=eLer
> t_ComputersInternet_ProgrammingLang_WhiteGridTime&t=ad> Click Here to Find
> Software Faster
> Click Here to Find Software Faster
>
>
<http://us.adserver.yahoo.com/l?M=162801.1342140.2934632.1280039/D=egroupmai
> l/S=1700115364:N/A=599120/rand=122086071>
>
> To unsubscribe from this group, send an email to:
> VBPros-unsubscribe@yahoogroups.com
>
>
>
> Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service
> <http://docs.yahoo.com/info/terms/> .
>
>
>

I don't know whether this discussion is still of interest, but...

----- Original Message -----
From: "DonEMitchell" <donemitchell@...>
To: <VBPros@yahoogroups.com>
Sent: Friday, March 23, 2001 12:34 AM
Subject: RE: [VBPros] XML Parser Project Proposal


> Good Morning,
>    ----Original Message-----
>   From: Tore [mailto:tbostrup@...]
>   Sent: Friday, March 23, 2001 12:56 AM
>   To: VBPros@yahoogroups.com
>   Subject: Re: [VBPros] XML Parser Project Proposal

<snip>

>>   I have two thoughts.  First, there is little need to maintain all the
>>   information about each node in memory - especially if each node carries
a
>>   significant amount of information.  Parsing the file, only critical
>>   information could be "cached", along with a pointer to the node in the
>> file.
>>   This way, the additional information would only be loaded when used -
kind
>>   of a Just-In-Time DOM implementation.
>
>   I agree there, Tore, regarding just-in-time caching, but if I understand
> you correctly, it means that any changes written back to the file would
> require re-parsing the file past the point where the changes were made.  I
> think this addresses the "reason to be" of the "node", which is to
represent
> the structure in an XML file as an object instance within a programming
> process space.

It depends.  You could do "dynamic" caching where you kept all updates in
memory.  When saving, you'd either load all the info into memory, or use a
temp file with a copy of the content as the source, and write back to the
original.  Or you could establish the temp file when opening or parsing the
file, and use it to save updates.  Since the data is found through a pointer
in memory, you can save updates to the end of the file.  The structure kept
in memory would represent the "proper" order, and the temp file would be
somewhat jumbled.

<snip>

>>   Second, I'm not sure I like the use of a collection indexed by the full
>> path
>>   of each node, thereby representing a linked structure in a flat media.
>> Why
>>   not just maintain node links, where each node has a collection of
>> children.
>>   This should allow for more flexibility and take all the trouble out of
>>   moving the branch of nodes (just modify or add link to the root of the
>>   subtree, and everything else follows).
>
>    Yes, regarding linked structure, node links, this is what a node is. A
> node contains node children.  The linked tree would stay in memory until
> destroyed.  Having an additional separate collection enables the use of
> getNode(XPath) which is I suppose a convenience to the alternative of
coding
> out a path of...
>
>
Node("ChildName").Node("GrandchildName").Node("GreatGrandchildName...
>
>   An example XPath I used in Vox2Data:
>   Private Const cXPathVox2DataSplit =
> "\UserProfile\Applications\Vox2Data\Metrics\frmVox2Data\Split"
>
>   The above constant would represent a lot of typing using explicit
> Node.NodeChild qualified property access.
>
>   And perhaps a more important use of a collection of references to nodes
is
> that access to any particular node is accomplished with a string vs.
> hard-coding into the node tree.  Strings can be modified at runtime
whereas
> hard-coding is hard-coding.

I didn't say anything about hardcoding.  And you can still use a path string
notation.  It is just that instead of having the full path as the key in a
single collection, and referencing the node by direct reference to the
collection, you use a property of the node object:

Assume you have an "empty" object which in itself does not have a path, but
holds the collection of actual root nodes (if more than one allowed), and we
want a reference to the node represented by
"\UserProfile\Applications\Vox2Data\Metrics\frmVox2Data\Split":

(The leading "\" was used to determine that we wanted to start from the
"empty" root collection instead of some current node).


gobjEmpty and objSplit are both references to a clsNode, with a method
GetNode, and a property Children (collection):
----
Set objSplit =
gobjEmpty.GetNode("UserProfile\Applications\Vox2Data\Metrics\frmVox2Data\Spl
it")
----
Public Function GetNode(ByVal sPath as String) as clsNode

Dim vaKeys as Variant

     vaKeys = Split(sPath, "\")
     If LBound(vaKeys) < UBound(vaKeys) Then
         'Dig deeper
         Set GetNode = Me.Children(vaKeys(0)).GetNode(Mid$(sPath,
Len(vaKeys(0)) + 1))
     Else
         If LBound(vaKeys) = UBound(vaKeys) Then
             'We're at the wanted node's parent
             Set GetNode = Me.Children(vaKeys(0))
         Else
             'The path was empty - I am the node...
             Set GetNode = Me
         End If
     End If

End Function
----

<snip>

>   What is the modeling info you don't think you can code in a structured
set
>   of name, content and attributes?
>
>   I suppose anything could be modeled with node content alone, not even
> using attributes.  To avoid bogging down a modeling design, I envisioned
> more convenience qualities per node, namely, Semantics (keywords used in
> commands) and Pragmatics (linking semantics to syntactic nodes per
> algorithmic function), to complete Webster's Dictionary definition of
> Semiotics as one-node representing a Semiotic quanta.  With these added
> qualities of my convention, what I didn't explain is that a completely
> de-coupled command interpreter could manipulate a node-tree to become the
> silicon-slave, the beast-of-burden, the work-horse of information society.
> By adding extra qualities of convention, the convention would enliven
> information mobility and threshold-based-raising of algorithms. Those
things
> are all best achieved when and if a basic XML node is created, though, for
> sure, huh? :)
>
>   Cheers,
>   DonE.

Wouldn't it be natural to make those attributes (or contained nodes
depending on complexity)?

Regards,
Tore.

I haven't done any HL7 work, but I was "threatened" with it about a year or
so back, and so I have been subscribing to an HL7 list, and have a complete
archive.  There is a lot of activity in that area - and XML is key to its
future.  Would you be interested in me trying to export the email folder?

Also, MS has a healthcare "framework" that is available from their web site.
I think it may have some HL7 code in it somewhere...

Regards,
Tore.

----- Original Message -----
From: "DonEMitchell" <donemitchell@...>
To: <VBPros@yahoogroups.com>
Sent: Friday, March 23, 2001 8:58 AM
Subject: RE: [VBPros] XML Parser Project Proposal


<snip>

> P.S.  If I sound a little perturbed, I'm sorry...I just found out that my
> boss sold a contract for deliverable software, $500,000.00 per year
license
> to the hospital, done deal, signed, my deliverable software, with a 90 day
> window for production! Not a good day to tell me I need to look deeper.
And
> I truly don't know what it is they want us to do.  I only know we will be
> receiving an HL7 Message Record, and we'll populate it with dictation that
> relies on unknown data-merges from unkown data locations. I don't know if
> I'll be receiving files or IP data streams. 90 days/half a million bucks
per
> annum. My vapor ware. It may be a good day for me to call a head hunter.

<snip>

Hello All,

I've had this idea for quite a while about how to perform operations
on very large integers. Numbers that are beyond the scope of your
imagination in some cases. A few days ago someone pointed me to a
resource that 'works' for smaller 'very large integers' - but it
needs some work.

I tried mulitplying a number into the thousand digit range and it
'died' - to put it politely. And it isn't capable of some of the
features I need to implement. Anyone have any ideas on how I can add
extensive mathematical features using "very" large numbers to a
project?

Things I'll need to be able to do 'long math' including roots,
powers and stuff with numbers in the *billions* of *digits* range.
Is it even possible? How? I don't care if I have to resort to C++ -
as long as I can expose it to a VB object model.

Thanks,

Shawn
http://ReliableAnswers.com

> Shawn!  Maybe in a year or two I can hire your son for a
> small contract.  And I'm serious.  I think it would be terrific
> PR to have some code written by a child prodigy.

Sounds great, but you'd better hurry. My wife is talking about
getting him in to acting. He's got awesome red hair (you know the
only thing those people care about is how you look) and though he's
only 4 he can be very persuasive. A few days ago I said something
and he mimicked my inflection perfectly.

The first phrase I taught him was "<sigh>, women" - and it has been
a pleasure to this day!


> P.S. By the way, I get lost in a mass of emails frequently,
> and I don't want anyone to think I've ignored any messages.
> Please always consider it prudent to repost anything verses
> feeling censured, K?

That goes for me too. The last month I devoted every waking moment
to that contract and I'vegot a few thousand messages to catch up
on - so please don't think I'm ignoring you if I don't respond in a
timely manner. The way I do my email I 'click through' anything that
looks boring or unimportant and save the 'important stuff' until I
can devote enough of my attention to give it the appreciation it
deserves. If it takes me a *real long time* to answer your messages
you should be proud of having generated that much attention by me.
Or, maybe I just missed it. <g>

I've been rolling a project idea around in my head for a few years
now and maybe this is the group to bring it up in. Will send under
seperate cover.

Regards,

Shawn K. Hall
http://ReliableAnswers.com

-----Original Message-----
From: DonEMitchell [mailto:donemitchell@...]
Sent: Friday, March 23, 2001 12:26 AM
To: VBPros@yahoogroups.com
Subject: RE: [VBPros] Broad spectrum

Hi Will,

 

I would love to take you up on voice-automating VB through SliceAndDice.  But later.  I'm at near maximum bandwidth on the amount of interests I can allow before a meltdown.  

 

I've spent a significant amount of time with XML nodes, selfishly enough, because I'm at that exact place in my work-a-day for pay existence.  I'm looking forward to some SND reprieve, soon.

 

Are there any new releases since SND 1.8.11 from the web?

 

Cheers,

DonE

-----Original Message-----
From: Rawls, William [mailto:wrawls@...]
Sent: Thursday, March 22, 2001 12:35 PM
To: 'VBPros@yahoogroups.com'
Subject: RE: [VBPros] Broad spectrum

 <snip>   

 

Voice enabling VB is as easy as Voice enabling Slice and Dice. Slice and Dice hides the details and indeed the need to work with the VB IDE. It exposes the VB IDE to you if you want, but is directly intended to hide the complexities from you. All the Sandals (Slice and Dice External DLLs, aka add-ins) demonstrate how to manipulate Slice and Dice with simple DLLs and forms.

 

If you give me some working voice enabling code in DLL format I'll show you how easy it can be.

 

This is chaotic. Cooperation that understands that what I have and what you have work together to create more than either of us have and more than the simple sum. 

 

 <snip>

To unsubscribe from this group, send an email to:
VBPros-unsubscribe@yahoogroups.com



Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.

-----Original Message-----
From: Rawls, William [mailto:wrawls@...]
Sent: Thursday, March 22, 2001 12:35 PM
To: 'VBPros@yahoogroups.com'
Subject: RE: [VBPros] Broad spectrum

 <snip>   

 

Voice enabling VB is as easy as Voice enabling Slice and Dice. Slice and Dice hides the details and indeed the need to work with the VB IDE. It exposes the VB IDE to you if you want, but is directly intended to hide the complexities from you. All the Sandals (Slice and Dice External DLLs, aka add-ins) demonstrate how to manipulate Slice and Dice with simple DLLs and forms.

 

If you give me some working voice enabling code in DLL format I'll show you how easy it can be.

 

This is chaotic. Cooperation that understands that what I have and what you have work together to create more than either of us have and more than the simple sum. 

 

 <snip>

-----Original Message-----
From: DonEMitchell [mailto:donemitchell@...]
Sent: Thursday, March 22, 2001 1:08 AM
To: VBPros@yahoogroups.com
Subject: RE: [VBPros] Broad spectrum

Cool info, Tore.  Thanks.

 

The corp. didn't go belly, just 100% of the executives were terminated or stepped down, and one of the board directors of the parent corp. is signing pay-checks.  They are interviewing for president, CFO, etc.  Who knows, they may regroup and run the stock price back up.  It's around two cents a share now at a real bargain :) 

 

Tore, the info regarding SQL server is clear enough, and even if performance lags, it's just for my training/testing.  Frankly, I'm looking for a first-time install of it by myself, and sort-of want to know what not to do.  I'm totally aware that the server can be a career in it's own right, and I'm willing to stay on the coder-side of a connection string.  I'm willing to leave expert knowledge of the SQL server discipline to a future employee, should the parent corporation ever get off the ground floor.  Me thinks me needs to set up the minimum SQL server to be able to use the DMO against for allowing a desktop business tier to build tables just-in-time as needed.  Twue, twue, more elegant schemes can evolve, but I'm leaving that to the progenies that are yet un-hired.

 

What do you all think.  Would I be better off using MSDE for development as it's just me here with sore arms?

 

And yes, boy would I like to voice enable VB!  I'd like to talk to that puppy like Janeway talks to the Enterprise. But -- VB has managed to crash & corrupt every time I start bumbling around with creating a component with a project that is instancing the same component.  That is, when I write code using Dragon tools, it's ok.  When I write code that uses the Dragon components in an add-in, something goes wrong every time.  I've lost hope.  Blue screens, corruptions, gnashin of teeth.

 

If it were a perfect world, I'd like to team with Mr. Rawls in a manor to map the cardinal over the ordinal, and leave the understanding process that recognizes the relationships to the buzzing mind of the coder.  Render the internal mind buzzing to English grammar spoken by the coder to enact feature-based cardinalities of code, where each feature was represented by a treeView branch (the cardinal) into an ordinal list of tasks that modify an evolving project (the purpose at hand).  But Will, the world is so far from perfect, huh?  I really don't think that much time is a luxury I can afford, and then there's the problem of communication/desire etc, etc.  It's still a dream child I can't forget, though. I've had some interest form the corporate world in a similare direction, basically, design ideas extrapolated out to realize a voice-commanded document engine, where the document was supported by a schema, and schema-based self-extesible algorithms that has some amazing implications (in my brain-space, anyway).

 

Did you guys by chance see Chris Lofting's verbose reply to Will's message into the Semiosis yahoogroup?  That was awesome, Will!  Congratulations! I'm going to be chewing on that for a long time.  I read your side bar a couple of months back or so and thought it a marvelous discourse.  I'm enamored and riveted to your approach at information technologies -- a devoted fan.

 

And, I'm scraping my TagPosition parser altogether, Tore, but you can find it at vbxml.com in the snippets section.  The code from http://www.prodigywave.com/ that Will provided is so much more elegant, and is the very same technique I was using.

 

Hey fellows  -- I just had a thought (the poor lonely thing) and want to know if we could all work together to polish up a stateful DOM using the Q-XML stateless XML parser?  I have the concept already in use with one linked node and a collection of references into the node tree keyed by node-path. And hey!  maybe we could approach the project slowly and document it all the way from blue-sky to downloadable example project to exemplify the documentation process.  (I'm beyond blue sky and presently using the technique.)

 

The above paragraph is with full knowledge that I want the technique to be known that it is mine, but that it is just a technique to be surrendered to the public domain.  There's more than one way to skin a lawyer.  And we all have techniques.  The W3C is full of them.  I just want to learn how to publish and look like somebody.  It would also help my contract documentation skills.  Ideas gentlemen?

 

Shawn!  Maybe in a year or two I can hire your son for a small contract.  And I'm serious.  I think it would be terrific PR to have some code written by a child prodigy.  My boss could go miles and miles with that woven into his sales pitch.  He already has some original military footage of the stealth bomber (original voice technology research contracts started big-time with the stealth) in a flash presentation.

 

I've really enjoyed our initial dialog on this list.

 

Cheers,

DonE

 

P.S. By the way, I get lost in a mass of emails frequently, and I don't want anyone to think I've ignored any messages.  Please always consider it prudent to repost anything verses feeling censured, K?

 

-----Original Message-----
From: Tore [mailto:tbostrup@...]
Sent: Wednesday, March 21, 2001 7:42 PM
To: VBPros@yahoogroups.com
Subject: Re: [VBPros] Broad spectrum

And cheers to you.  Sorry to hear they went belly-up.  Hope your work

Don, I think you need to put a voice front-end on VB... :->  RSI =
Repetitive Stress Injury?

I'll lend a hand with the group if you want, but won't be offended if
somebody else wants it.

While the computer you have should run SQL Server 2000, you'd probably be
better off with 256 MB or more.  If you wind up with a huge database, you
may want even more RAM.

The following are documented requirements (excerpts from MSDN):

      Hardware Minimum requirements
      Computer Intel® or compatible
      Pentium 166 MHz or higher.

      Memory (RAM)1 Enterprise Edition: 64 MB minimum, 128 MB or more
recommended
      Standard Edition: 64 MB minimum

      Personal Edition: 64 MB minimum on Windows 2000, 32 MB minimum on all
other operating systems

      Developer Edition: 64 MB minimum

      Desktop Engine: 64 MB minimum on Windows 2000, 32 MB minimum on all
other operating systems

      Hard disk space2  SQL Server database components: 95 to 270 MB, 250 MB
typical
      Analysis Services: 50 MB minimum, 130 MB typical

      English Query: 80 MB

      Desktop Engine only: 44 MB

      Monitor VGA or higher resolution
      800x600 or higher resolution required for the SQL Server graphical
tools

      Pointing device Microsoft Mouse or compatible
      CD-ROM drive Required


1 Additional memory may be required, depending on operating system
requirements.
2 Actual requirements will vary based on your system configuration and the
applications and features you choose to install.

-------

      SQL Server edition or component
      Operating system requirement
      Enterprise Edition Microsoft Windows NT Server 4.0, Microsoft Windows
NT Server Enterprise Edition 4.0, Windows 2000 Server, Windows 2000 Advanced
Server, and Windows 2000 Data Center Server.
      Note that Microsoft Windows 2000 Server (any version) is required for
some SQL Server 2000 features.

      Standard Edition Microsoft Windows NT Server 4.0, Windows 2000 Server,
Microsoft Windows NT Server Enterprise Edition, Windows 2000 Advanced
Server, and Windows 2000 Data Center Server.
      Personal Edition Microsoft Windows Me, Windows 98, Windows NT
Workstation 4.0, Windows 2000 Professional, Microsoft Windows NT Server 4.0,
Windows 2000 Server, and all the more advanced Windows operating systems.
      Developer Edition Microsoft Windows NT Workstation 4.0, Windows 2000
Professional, and all other Windows NT and Windows 2000 operating systems.
      Client Tools Only  Microsoft Windows NT 4.0, Windows 2000 (all
versions), Windows Me, and Windows 98.
      Connectivity Only Microsoft Windows NT 4.0, Windows 2000 (all
versions), Windows Me, Windows 98, and Windows 95.



Note  Microsoft Windows NT® Server 4.0, Service Pack 5 (SP5) or later must
be installed as a minimum requirement for all SQL Server 2000 editions.
SQL Server 2000 is not supported on Windows NT 4.0 Terminal Server.


For installations of SQL Server 2000 Personal Edition on Windows 98
computers without a network card, Windows 98 Second Edition is required.

-------

If writing XML applications:

System Requirements for IIS Virtual Directory Management
The IIS Virtual Directory Management for SQL Server utility can run on a
computer running any edition of Microsoft® Windows NT® 4.0 or Microsoft
Windows® 2000. Computers running Windows NT 4.0 require:

  a.. Microsoft Internet Information Server 4.0 or higher (or Peer Web
Services 4.0 or higher on Windows NT Workstation 4.0).


  b.. Microsoft Management Console 1.2 (installed by the Windows NT Option
Pack and by SQL Server 2000 Setup).
For computers running Microsoft Windows 2000 Professional, the
Administrative Tools pack (Adminpak.msi) must be installed. This file is
located in the %windir%\System32 folder of the Windows 2000 Server editions.

-------

I'll look at your code if you want (I'm always curious).  If you plan on
commercializing it, you may want to send it through private email rather
than allowing it to wind up in a list archive somewhere...

As for compact communication - the English language contains about 75%
redundancy IIRC.  However, the redundancy is apparently important since we
usually don't concentrate 100%.

My favorite saying:

"The trouble with communication is the illusion that it has been achieved."

Regards,
Tore.

----- Original Message -----
From: "DonEMitchell" <donemitchell@...>
To: "VBPros@yahoogroups. com" <VBPros@yahoogroups.com>
Sent: Wednesday, March 21, 2001 12:24 AM
Subject: [VBPros] Broad spectrum


> Hi gentlemen, [Say, this got long, sorry:]
>
> My, I'm humbled by the presence of all three; Will, Tore, Shawn.
>
> To catch up a wee bit with more of my profile, my strongest forte would be
> with architecture designs.  My weakest point is also architectural designs
> :)  Restated, maybe: I can dream up lots of designs, and to date, I've
been
> the only fellow available to implement the designs, so I live with the
> nightmare.  <chuckles>
>
> Well for one, I'm very interested in everyone's talents.  Wow!  I'm not
sure
> where to start, there is definitely a very broad representation from
diverse
> points in our profession as VBers.  I would like to say that, though I
> realize there are tons of literature available on new technologies, I'd
hope
> to be able to glean a perspective on what's available that is within reach
> of a one-man department.  That's me.  My bottom line is I truly need the
> advice and guidance from other professionals that have 'the tee-shirt' so
I
> don't suffer re-inventing a perfect way to do it wrong.  I suppose that's
my
> motivation for wanting to begin a cross-discipline group.  Another point
> that makes it difficult is my broken arms -- chronic RSI...I have to do
> everything through some level of pain, so the luxury of writing
ten-thousand
> lines to trial-and-error the first implementation punch through is nearly
> beyond my envelope.  I'm slowly getting better with my techniques.  Will's
> SliceAndDice addin has been a big part of my recovery, though, what hurts
> the most is to think with my fingers at the ready...an old hard-to-break
> habit.  Typing email isn't so bad. (I should be voice-dictating it, huh?)
>
> To continue my profile, I've skull-scraped up a new personal paradigm that
> fits like a glove with voice-commands which act as a user interface, a
> linguistic UI...the LUI.  The magic of being at the right place at the
right
> time has rather propelled me into being owned by a corporation that
couldn't
> even fund the startup...they just wanted the glitter of the new
technology.
> (They went flop, Tore...the corp. executives were all released by the
board
> last week).  But the glitter of the new technology is still attracting big
> fish in the pond, and another startup is underway, while my true
> genius-of-a-marketing-boss is marketing full-steam, presently
demonstrating
> voice technologies to yet another hospital chain in California.  We have
one
> hospital sold already in waiting (our second, the first Mayo-clinic had to
> close and re-open, and our contract went out with their CEO.)
>
> And Shawn, I don't have any definite future myself, but if the creek
doesn't
> rise, I'll cross over to some lucrative possibilities any year now :)  I
> would like to consider VBPros to be a place to gather for all the news
about
> who's hiring, what's available, etc, as well as practical help.
>
> Enough 'bout me for one email :)
>
> I feel I know each of you here, and have spoken with each of you by phone.
> Nice coincidence.  Would anyone want to be co-moderator so the features of
> yahoogroups can be used to set up files areas, polls, etc.  Of course, a
> swollen membership would make this more interesting.  Anyone that wants to
> just speak up.  I'm just another wheel here.
>
> Before closing, have you found Semiosis@yahoogroups.com, Will?  I have had
> more head-stretching from lurking that list than I though was possible.
It
> seems they are attempting to punch-through with a redefinition of
> information technologies, starting with a loosely used term,
> quantum-neuro-psychology.
>
> I'll post a copy of Hitchhiker's Guide to Semiosis, written by Chris
Lofting
> to the files section.  Fun reading.  It makes me realize that a wizard
> approach to syntactic metonymy (assigning symbols to meaning) could likely
> facilitate a thought process being converted to compilable code in project
> form.
>
> But back to the practical...I really need to set up my home/business
systems
> with SQL 2000 and get familiar with XML returned as a query.  How much
> machine do I need to run SQL 2000.  I have a 266/128 meg box available.
Is
> that enough?
>
> I also need to perfect an XML tag parser I wrote, or bail to the 1
megabyte
> MSDOM, which I'd rather not.  Would anyone want to see and critique the
> code?...it's brute force VB string stuff.  I found a couple of logical
flaws
> and would like to re-visit the design.  I've got the time.  What would be
a
> workable parsing technique.  Links? Suggestions?  The present design is at
> vbxml.com in the snippets section under VB.
>
> Lot's of words.  I'll try to learn compact precision for better
> communication skills, too. :)
>
> I'm glad to be surrounded by, as is so happens, my top three choices among
> any VB coders I know.  That's factual, rather incredible.  It's twue it's
> twue.
>
> Thank you gentlemen.  I'm looking forward to how our future evolves.
>
> Cheers,
>
> DonEMitchell
>
>
>
> To unsubscribe from this group, send an email to:
> VBPros-unsubscribe@yahoogroups.com
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>



To unsubscribe from this group, send an email to:
VBPros-unsubscribe@yahoogroups.com



Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.

To unsubscribe from this group, send an email to:
VBPros-unsubscribe@yahoogroups.com



Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.

-----Original Message-----
From: Tore [mailto:tbostrup@...]
Sent: Wednesday, March 21, 2001 7:42 PM
To: VBPros@yahoogroups.com
Subject: Re: [VBPros] Broad spectrum

And cheers to you.  Sorry to hear they went belly-up.  Hope your work

Don, I think you need to put a voice front-end on VB... :->  RSI =
Repetitive Stress Injury?

I'll lend a hand with the group if you want, but won't be offended if
somebody else wants it.

While the computer you have should run SQL Server 2000, you'd probably be
better off with 256 MB or more.  If you wind up with a huge database, you
may want even more RAM.

The following are documented requirements (excerpts from MSDN):

      Hardware Minimum requirements
      Computer Intel® or compatible
      Pentium 166 MHz or higher.

      Memory (RAM)1 Enterprise Edition: 64 MB minimum, 128 MB or more
recommended
      Standard Edition: 64 MB minimum

      Personal Edition: 64 MB minimum on Windows 2000, 32 MB minimum on all
other operating systems

      Developer Edition: 64 MB minimum

      Desktop Engine: 64 MB minimum on Windows 2000, 32 MB minimum on all
other operating systems

      Hard disk space2  SQL Server database components: 95 to 270 MB, 250 MB
typical
      Analysis Services: 50 MB minimum, 130 MB typical

      English Query: 80 MB

      Desktop Engine only: 44 MB

      Monitor VGA or higher resolution
      800x600 or higher resolution required for the SQL Server graphical
tools

      Pointing device Microsoft Mouse or compatible
      CD-ROM drive Required


1 Additional memory may be required, depending on operating system
requirements.
2 Actual requirements will vary based on your system configuration and the
applications and features you choose to install.

-------

      SQL Server edition or component
      Operating system requirement
      Enterprise Edition Microsoft Windows NT Server 4.0, Microsoft Windows
NT Server Enterprise Edition 4.0, Windows 2000 Server, Windows 2000 Advanced
Server, and Windows 2000 Data Center Server.
      Note that Microsoft Windows 2000 Server (any version) is required for
some SQL Server 2000 features.

      Standard Edition Microsoft Windows NT Server 4.0, Windows 2000 Server,
Microsoft Windows NT Server Enterprise Edition, Windows 2000 Advanced
Server, and Windows 2000 Data Center Server.
      Personal Edition Microsoft Windows Me, Windows 98, Windows NT
Workstation 4.0, Windows 2000 Professional, Microsoft Windows NT Server 4.0,
Windows 2000 Server, and all the more advanced Windows operating systems.
      Developer Edition Microsoft Windows NT Workstation 4.0, Windows 2000
Professional, and all other Windows NT and Windows 2000 operating systems.
      Client Tools Only  Microsoft Windows NT 4.0, Windows 2000 (all
versions), Windows Me, and Windows 98.
      Connectivity Only Microsoft Windows NT 4.0, Windows 2000 (all
versions), Windows Me, Windows 98, and Windows 95.



Note  Microsoft Windows NT® Server 4.0, Service Pack 5 (SP5) or later must
be installed as a minimum requirement for all SQL Server 2000 editions.
SQL Server 2000 is not supported on Windows NT 4.0 Terminal Server.


For installations of SQL Server 2000 Personal Edition on Windows 98
computers without a network card, Windows 98 Second Edition is required.

-------

If writing XML applications:

System Requirements for IIS Virtual Directory Management
The IIS Virtual Directory Management for SQL Server utility can run on a
computer running any edition of Microsoft® Windows NT® 4.0 or Microsoft
Windows® 2000. Computers running Windows NT 4.0 require:

  a.. Microsoft Internet Information Server 4.0 or higher (or Peer Web
Services 4.0 or higher on Windows NT Workstation 4.0).


  b.. Microsoft Management Console 1.2 (installed by the Windows NT Option
Pack and by SQL Server 2000 Setup).
For computers running Microsoft Windows 2000 Professional, the
Administrative Tools pack (Adminpak.msi) must be installed. This file is
located in the %windir%\System32 folder of the Windows 2000 Server editions.

-------

I'll look at your code if you want (I'm always curious).  If you plan on
commercializing it, you may want to send it through private email rather
than allowing it to wind up in a list archive somewhere...

As for compact communication - the English language contains about 75%
redundancy IIRC.  However, the redundancy is apparently important since we
usually don't concentrate 100%.

My favorite saying:

"The trouble with communication is the illusion that it has been achieved."

Regards,
Tore.

----- Original Message -----
From: "DonEMitchell" <donemitchell@...>
To: "VBPros@yahoogroups. com" <VBPros@yahoogroups.com>
Sent: Wednesday, March 21, 2001 12:24 AM
Subject: [VBPros] Broad spectrum


> Hi gentlemen, [Say, this got long, sorry:]
>
> My, I'm humbled by the presence of all three; Will, Tore, Shawn.
>
> To catch up a wee bit with more of my profile, my strongest forte would be
> with architecture designs.  My weakest point is also architectural designs
> :)  Restated, maybe: I can dream up lots of designs, and to date, I've
been
> the only fellow available to implement the designs, so I live with the
> nightmare.  <chuckles>
>
> Well for one, I'm very interested in everyone's talents.  Wow!  I'm not
sure
> where to start, there is definitely a very broad representation from
diverse
> points in our profession as VBers.  I would like to say that, though I
> realize there are tons of literature available on new technologies, I'd
hope
> to be able to glean a perspective on what's available that is within reach
> of a one-man department.  That's me.  My bottom line is I truly need the
> advice and guidance from other professionals that have 'the tee-shirt' so
I
> don't suffer re-inventing a perfect way to do it wrong.  I suppose that's
my
> motivation for wanting to begin a cross-discipline group.  Another point
> that makes it difficult is my broken arms -- chronic RSI...I have to do
> everything through some level of pain, so the luxury of writing
ten-thousand
> lines to trial-and-error the first implementation punch through is nearly
> beyond my envelope.  I'm slowly getting better with my techniques.  Will's
> SliceAndDice addin has been a big part of my recovery, though, what hurts
> the most is to think with my fingers at the ready...an old hard-to-break
> habit.  Typing email isn't so bad. (I should be voice-dictating it, huh?)
>
> To continue my profile, I've skull-scraped up a new personal paradigm that
> fits like a glove with voice-commands which act as a user interface, a
> linguistic UI...the LUI.  The magic of being at the right place at the
right
> time has rather propelled me into being owned by a corporation that
couldn't
> even fund the startup...they just wanted the glitter of the new
technology.
> (They went flop, Tore...the corp. executives were all released by the
board
> last week).  But the glitter of the new technology is still attracting big
> fish in the pond, and another startup is underway, while my true
> genius-of-a-marketing-boss is marketing full-steam, presently
demonstrating
> voice technologies to yet another hospital chain in California.  We have
one
> hospital sold already in waiting (our second, the first Mayo-clinic had to
> close and re-open, and our contract went out with their CEO.)
>
> And Shawn, I don't have any definite future myself, but if the creek
doesn't
> rise, I'll cross over to some lucrative possibilities any year now :)  I
> would like to consider VBPros to be a place to gather for all the news
about
> who's hiring, what's available, etc, as well as practical help.
>
> Enough 'bout me for one email :)
>
> I feel I know each of you here, and have spoken with each of you by phone.
> Nice coincidence.  Would anyone want to be co-moderator so the features of
> yahoogroups can be used to set up files areas, polls, etc.  Of course, a
> swollen membership would make this more interesting.  Anyone that wants to
> just speak up.  I'm just another wheel here.
>
> Before closing, have you found Semiosis@yahoogroups.com, Will?  I have had
> more head-stretching from lurking that list than I though was possible.
It
> seems they are attempting to punch-through with a redefinition of
> information technologies, starting with a loosely used term,
> quantum-neuro-psychology.
>
> I'll post a copy of Hitchhiker's Guide to Semiosis, written by Chris
Lofting
> to the files section.  Fun reading.  It makes me realize that a wizard
> approach to syntactic metonymy (assigning symbols to meaning) could likely
> facilitate a thought process being converted to compilable code in project
> form.
>
> But back to the practical...I really need to set up my home/business
systems
> with SQL 2000 and get familiar with XML returned as a query.  How much
> machine do I need to run SQL 2000.  I have a 266/128 meg box available.
Is
> that enough?
>
> I also need to perfect an XML tag parser I wrote, or bail to the 1
megabyte
> MSDOM, which I'd rather not.  Would anyone want to see and critique the
> code?...it's brute force VB string stuff.  I found a couple of logical
flaws
> and would like to re-visit the design.  I've got the time.  What would be
a
> workable parsing technique.  Links? Suggestions?  The present design is at
> vbxml.com in the snippets section under VB.
>
> Lot's of words.  I'll try to learn compact precision for better
> communication skills, too. :)
>
> I'm glad to be surrounded by, as is so happens, my top three choices among
> any VB coders I know.  That's factual, rather incredible.  It's twue it's
> twue.
>
> Thank you gentlemen.  I'm looking forward to how our future evolves.
>
> Cheers,
>
> DonEMitchell
>
>
>
> To unsubscribe from this group, send an email to:
> VBPros-unsubscribe@yahoogroups.com
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>



To unsubscribe from this group, send an email to:
VBPros-unsubscribe@yahoogroups.com



Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.

-----Original Message-----
From: DonEMitchell [mailto:donemitchell@...]
Sent: Tuesday, March 20, 2001 11:50 PM
To: VBPros@yahoogroups. com
Subject: [VBPros] Diagramming projects

Hi group,

I have a need -- to diagrammatically represent a project in proper method to
contract out the work.

Background:

I've worked largely alone throughout my programming career.  My brief team
effort was with the FAA, where I was shown four circles with five lines,
given a functional expert that didn't know answers, and left to produce.
Beyond that I've been a successful one-man desktop application factory,
where I program VB code as I think it up.

Present:

Now I have a need to rapidly fulfill the first of a possible cascade of
lucrative contracts in the medical market.  These contracts need to be
substantiated by concise documentation that will underlie binding contracts.

Tools at hand:

I'd like to use MS Visio 2000 and/or PowerPoint.


Next step:

Seek advice.  That's why I'm here.  Any advice that could focus my research
into the ways and means of documenting a contract would be really
appreciated.

Questions:

Should I hire a consultant before I begin that could assist/mentor my
learning?  Or is the typical contract done by so many pages of prose, with
lots of telephone conversations/one-on-one dialog to substantiate what is
needed?


Many thanks guys.

Cheers,

DonE



To unsubscribe from this group, send an email to:
VBPros-unsubscribe@yahoogroups.com



Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.

And cheers to you.  Sorry to hear they went belly-up.  Hope your work

Don, I think you need to put a voice front-end on VB... :->  RSI =
Repetitive Stress Injury?

I'll lend a hand with the group if you want, but won't be offended if
somebody else wants it.

While the computer you have should run SQL Server 2000, you'd probably be
better off with 256 MB or more.  If you wind up with a huge database, you
may want even more RAM.

The following are documented requirements (excerpts from MSDN):

       Hardware Minimum requirements
       Computer Intel® or compatible
       Pentium 166 MHz or higher.

       Memory (RAM)1 Enterprise Edition: 64 MB minimum, 128 MB or more
recommended
       Standard Edition: 64 MB minimum

       Personal Edition: 64 MB minimum on Windows 2000, 32 MB minimum on all
other operating systems

       Developer Edition: 64 MB minimum

       Desktop Engine: 64 MB minimum on Windows 2000, 32 MB minimum on all
other operating systems

       Hard disk space2  SQL Server database components: 95 to 270 MB, 250 MB
typical
       Analysis Services: 50 MB minimum, 130 MB typical

       English Query: 80 MB

       Desktop Engine only: 44 MB

       Monitor VGA or higher resolution
       800x600 or higher resolution required for the SQL Server graphical
tools

       Pointing device Microsoft Mouse or compatible
       CD-ROM drive Required


1 Additional memory may be required, depending on operating system
requirements.
2 Actual requirements will vary based on your system configuration and the
applications and features you choose to install.

-------

       SQL Server edition or component
       Operating system requirement
       Enterprise Edition Microsoft Windows NT Server 4.0, Microsoft Windows
NT Server Enterprise Edition 4.0, Windows 2000 Server, Windows 2000 Advanced
Server, and Windows 2000 Data Center Server.
       Note that Microsoft Windows 2000 Server (any version) is required for
some SQL Server 2000 features.

       Standard Edition Microsoft Windows NT Server 4.0, Windows 2000 Server,
Microsoft Windows NT Server Enterprise Edition, Windows 2000 Advanced
Server, and Windows 2000 Data Center Server.
       Personal Edition Microsoft Windows Me, Windows 98, Windows NT
Workstation 4.0, Windows 2000 Professional, Microsoft Windows NT Server 4.0,
Windows 2000 Server, and all the more advanced Windows operating systems.
       Developer Edition Microsoft Windows NT Workstation 4.0, Windows 2000
Professional, and all other Windows NT and Windows 2000 operating systems.
       Client Tools Only  Microsoft Windows NT 4.0, Windows 2000 (all
versions), Windows Me, and Windows 98.
       Connectivity Only Microsoft Windows NT 4.0, Windows 2000 (all
versions), Windows Me, Windows 98, and Windows 95.



Note  Microsoft Windows NT® Server 4.0, Service Pack 5 (SP5) or later must
be installed as a minimum requirement for all SQL Server 2000 editions.
SQL Server 2000 is not supported on Windows NT 4.0 Terminal Server.


For installations of SQL Server 2000 Personal Edition on Windows 98
computers without a network card, Windows 98 Second Edition is required.

-------

If writing XML applications:

System Requirements for IIS Virtual Directory Management
The IIS Virtual Directory Management for SQL Server utility can run on a
computer running any edition of Microsoft® Windows NT® 4.0 or Microsoft
Windows® 2000. Computers running Windows NT 4.0 require:

   a.. Microsoft Internet Information Server 4.0 or higher (or Peer Web
Services 4.0 or higher on Windows NT Workstation 4.0).


   b.. Microsoft Management Console 1.2 (installed by the Windows NT Option
Pack and by SQL Server 2000 Setup).
For computers running Microsoft Windows 2000 Professional, the
Administrative Tools pack (Adminpak.msi) must be installed. This file is
located in the %windir%\System32 folder of the Windows 2000 Server editions.

-------

I'll look at your code if you want (I'm always curious).  If you plan on
commercializing it, you may want to send it through private email rather
than allowing it to wind up in a list archive somewhere...

As for compact communication - the English language contains about 75%
redundancy IIRC.  However, the redundancy is apparently important since we
usually don't concentrate 100%.

My favorite saying:

"The trouble with communication is the illusion that it has been achieved."

Regards,
Tore.

----- Original Message -----
From: "DonEMitchell" <donemitchell@...>
To: "VBPros@yahoogroups. com" <VBPros@yahoogroups.com>
Sent: Wednesday, March 21, 2001 12:24 AM
Subject: [VBPros] Broad spectrum


> Hi gentlemen, [Say, this got long, sorry:]
>
> My, I'm humbled by the presence of all three; Will, Tore, Shawn.
>
> To catch up a wee bit with more of my profile, my strongest forte would be
> with architecture designs.  My weakest point is also architectural designs
> :)  Restated, maybe: I can dream up lots of designs, and to date, I've
been
> the only fellow available to implement the designs, so I live with the
> nightmare.  <chuckles>
>
> Well for one, I'm very interested in everyone's talents.  Wow!  I'm not
sure
> where to start, there is definitely a very broad representation from
diverse
> points in our profession as VBers.  I would like to say that, though I
> realize there are tons of literature available on new technologies, I'd
hope
> to be able to glean a perspective on what's available that is within reach
> of a one-man department.  That's me.  My bottom line is I truly need the
> advice and guidance from other professionals that have 'the tee-shirt' so
I
> don't suffer re-inventing a perfect way to do it wrong.  I suppose that's
my
> motivation for wanting to begin a cross-discipline group.  Another point
> that makes it difficult is my broken arms -- chronic RSI...I have to do
> everything through some level of pain, so the luxury of writing
ten-thousand
> lines to trial-and-error the first implementation punch through is nearly
> beyond my envelope.  I'm slowly getting better with my techniques.  Will's
> SliceAndDice addin has been a big part of my recovery, though, what hurts
> the most is to think with my fingers at the ready...an old hard-to-break
> habit.  Typing email isn't so bad. (I should be voice-dictating it, huh?)
>
> To continue my profile, I've skull-scraped up a new personal paradigm that
> fits like a glove with voice-commands which act as a user interface, a
> linguistic UI...the LUI.  The magic of being at the right place at the
right
> time has rather propelled me into being owned by a corporation that
couldn't
> even fund the startup...they just wanted the glitter of the new
technology.
> (They went flop, Tore...the corp. executives were all released by the
board
> last week).  But the glitter of the new technology is still attracting big
> fish in the pond, and another startup is underway, while my true
> genius-of-a-marketing-boss is marketing full-steam, presently
demonstrating
> voice technologies to yet another hospital chain in California.  We have
one
> hospital sold already in waiting (our second, the first Mayo-clinic had to
> close and re-open, and our contract went out with their CEO.)
>
> And Shawn, I don't have any definite future myself, but if the creek
doesn't
> rise, I'll cross over to some lucrative possibilities any year now :)  I
> would like to consider VBPros to be a place to gather for all the news
about
> who's hiring, what's available, etc, as well as practical help.
>
> Enough 'bout me for one email :)
>
> I feel I know each of you here, and have spoken with each of you by phone.
> Nice coincidence.  Would anyone want to be co-moderator so the features of
> yahoogroups can be used to set up files areas, polls, etc.  Of course, a
> swollen membership would make this more interesting.  Anyone that wants to
> just speak up.  I'm just another wheel here.
>
> Before closing, have you found Semiosis@yahoogroups.com, Will?  I have had
> more head-stretching from lurking that list than I though was possible.
It
> seems they are attempting to punch-through with a redefinition of
> information technologies, starting with a loosely used term,
> quantum-neuro-psychology.
>
> I'll post a copy of Hitchhiker's Guide to Semiosis, written by Chris
Lofting
> to the files section.  Fun reading.  It makes me realize that a wizard
> approach to syntactic metonymy (assigning symbols to meaning) could likely
> facilitate a thought process being converted to compilable code in project
> form.
>
> But back to the practical...I really need to set up my home/business
systems
> with SQL 2000 and get familiar with XML returned as a query.  How much
> machine do I need to run SQL 2000.  I have a 266/128 meg box available.
Is
> that enough?
>
> I also need to perfect an XML tag parser I wrote, or bail to the 1
megabyte
> MSDOM, which I'd rather not.  Would anyone want to see and critique the
> code?...it's brute force VB string stuff.  I found a couple of logical
flaws
> and would like to re-visit the design.  I've got the time.  What would be
a
> workable parsing technique.  Links? Suggestions?  The present design is at
> vbxml.com in the snippets section under VB.
>
> Lot's of words.  I'll try to learn compact precision for better
> communication skills, too. :)
>
> I'm glad to be surrounded by, as is so happens, my top three choices among
> any VB coders I know.  That's factual, rather incredible.  It's twue it's
> twue.
>
> Thank you gentlemen.  I'm looking forward to how our future evolves.
>
> Cheers,
>
> DonEMitchell
>
>
>
> To unsubscribe from this group, send an email to:
> VBPros-unsubscribe@yahoogroups.com
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>

Boy, talk about coincidence.  Just a day or two ago, I came across a
download called VisioModeler 3.1 on the MSDN web site.

I think your best bet is to look to this and/or Visio, using UML notations
if you plan to diagram for requirements.  I also happen to know one of the
gurus in the UML arena who has also looked at using it with VB and other
"non-OO" languages.  His name is Gary Evans (gkevans@...), and he
does short-term contracts.  If you decide to contact him, be sure to send
him my regards.

As far as writing the contract, you'll need someone to handle the legalese,
of course.  However, if the legalese becomes important, you already have a
problem.  In order to ensure success, you need to make sure the requirements
are complete and non-ambiguous - which is impossible.  So you need to manage
the contract during execution to make sure that development follows the
appropriate course, and you can catch problems at an early stage.  Require
frequent reports and/or reviews (weekly is usually good).  Try to use an
iterative approach with those delivering to you, and do the same toward
those for whom you develop - either a specific client or a "focus group".
The sooner you and the end user sees what things will look like - the sooner
questions will arise and issues uncovered.

In addition to requirements, define the deliverables as well as the
acceptance criteria, and how deviations are handled.  Any changes, whether
resulting in a financial or schedule impact, should be implemented through a
documented change request.  Include some quality criteria on the delivered
components, and define whether any post-delivery support is provided, and
under what arrangements.

Most contracts, except those that are way too large to consider, are
burdened with some set of problems, usually a lack of detailed definition.
Project success is usually despite the contract, because both client
expectations and developer understanding was properly managed.

My favorite project was driven forward by weekly presentations of selected
design issues on a weekly basis, up until the documentation was written.
The code came last, and I had to consult the documentation several times.
The coding (and testing) itself was probably only about a third of the
effort.  The project was my Master's thesis, done as a collaborative project
with two other people, and we met with our professor weekly to discuss
whatever we had spent our time on that week.  We would usually bring about
1-3 pages documenting a specific issue or design, and review it with him.

Regards and HTH,
Tore.

----- Original Message -----
From: "DonEMitchell" <donemitchell@...>
To: "VBPros@yahoogroups. com" <VBPros@yahoogroups.com>
Sent: Wednesday, March 21, 2001 12:49 AM
Subject: [VBPros] Diagramming projects


> Hi group,
>
> I have a need -- to diagrammatically represent a project in proper method
to
> contract out the work.
>
> Background:
>
> I've worked largely alone throughout my programming career.  My brief team
> effort was with the FAA, where I was shown four circles with five lines,
> given a functional expert that didn't know answers, and left to produce.
> Beyond that I've been a successful one-man desktop application factory,
> where I program VB code as I think it up.
>
> Present:
>
> Now I have a need to rapidly fulfill the first of a possible cascade of
> lucrative contracts in the medical market.  These contracts need to be
> substantiated by concise documentation that will underlie binding
contracts.
>
> Tools at hand:
>
> I'd like to use MS Visio 2000 and/or PowerPoint.
>
>
> Next step:
>
> Seek advice.  That's why I'm here.  Any advice that could focus my
research
> into the ways and means of documenting a contract would be really
> appreciated.
>
> Questions:
>
> Should I hire a consultant before I begin that could assist/mentor my
> learning?  Or is the typical contract done by so many pages of prose, with
> lots of telephone conversations/one-on-one dialog to substantiate what is
> needed?
>
>
> Many thanks guys.
>
> Cheers,
>
> DonE
>
>
>
> To unsubscribe from this group, send an email to:
> VBPros-unsubscribe@yahoogroups.com
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>

Hi group,

I have a need -- to diagrammatically represent a project in proper method to
contract out the work.

Background:

I've worked largely alone throughout my programming career.  My brief team
effort was with the FAA, where I was shown four circles with five lines,
given a functional expert that didn't know answers, and left to produce.
Beyond that I've been a successful one-man desktop application factory,
where I program VB code as I think it up.

Present:

Now I have a need to rapidly fulfill the first of a possible cascade of
lucrative contracts in the medical market.  These contracts need to be
substantiated by concise documentation that will underlie binding contracts.

Tools at hand:

I'd like to use MS Visio 2000 and/or PowerPoint.


Next step:

Seek advice.  That's why I'm here.  Any advice that could focus my research
into the ways and means of documenting a contract would be really
appreciated.

Questions:

Should I hire a consultant before I begin that could assist/mentor my
learning?  Or is the typical contract done by so many pages of prose, with
lots of telephone conversations/one-on-one dialog to substantiate what is
needed?


Many thanks guys.

Cheers,

DonE

Hi gentlemen, [Say, this got long, sorry:]

My, I'm humbled by the presence of all three; Will, Tore, Shawn.

To catch up a wee bit with more of my profile, my strongest forte would be
with architecture designs.  My weakest point is also architectural designs
:)  Restated, maybe: I can dream up lots of designs, and to date, I've been
the only fellow available to implement the designs, so I live with the
nightmare.  <chuckles>

Well for one, I'm very interested in everyone's talents.  Wow!  I'm not sure
where to start, there is definitely a very broad representation from diverse
points in our profession as VBers.  I would like to say that, though I
realize there are tons of literature available on new technologies, I'd hope
to be able to glean a perspective on what's available that is within reach
of a one-man department.  That's me.  My bottom line is I truly need the
advice and guidance from other professionals that have 'the tee-shirt' so I
don't suffer re-inventing a perfect way to do it wrong.  I suppose that's my
motivation for wanting to begin a cross-discipline group.  Another point
that makes it difficult is my broken arms -- chronic RSI...I have to do
everything through some level of pain, so the luxury of writing ten-thousand
lines to trial-and-error the first implementation punch through is nearly
beyond my envelope.  I'm slowly getting better with my techniques.  Will's
SliceAndDice addin has been a big part of my recovery, though, what hurts
the most is to think with my fingers at the ready...an old hard-to-break
habit.  Typing email isn't so bad. (I should be voice-dictating it, huh?)

To continue my profile, I've skull-scraped up a new personal paradigm that
fits like a glove with voice-commands which act as a user interface, a
linguistic UI...the LUI.  The magic of being at the right place at the right
time has rather propelled me into being owned by a corporation that couldn't
even fund the startup...they just wanted the glitter of the new technology.
(They went flop, Tore...the corp. executives were all released by the board
last week).  But the glitter of the new technology is still attracting big
fish in the pond, and another startup is underway, while my true
genius-of-a-marketing-boss is marketing full-steam, presently demonstrating
voice technologies to yet another hospital chain in California.  We have one
hospital sold already in waiting (our second, the first Mayo-clinic had to
close and re-open, and our contract went out with their CEO.)

And Shawn, I don't have any definite future myself, but if the creek doesn't
rise, I'll cross over to some lucrative possibilities any year now :)  I
would like to consider VBPros to be a place to gather for all the news about
who's hiring, what's available, etc, as well as practical help.

Enough 'bout me for one email :)

I feel I know each of you here, and have spoken with each of you by phone.
Nice coincidence.  Would anyone want to be co-moderator so the features of
yahoogroups can be used to set up files areas, polls, etc.  Of course, a
swollen membership would make this more interesting.  Anyone that wants to
just speak up.  I'm just another wheel here.

Before closing, have you found Semiosis@yahoogroups.com, Will?  I have had
more head-stretching from lurking that list than I though was possible.  It
seems they are attempting to punch-through with a redefinition of
information technologies, starting with a loosely used term,
quantum-neuro-psychology.

I'll post a copy of Hitchhiker's Guide to Semiosis, written by Chris Lofting
to the files section.  Fun reading.  It makes me realize that a wizard
approach to syntactic metonymy (assigning symbols to meaning) could likely
facilitate a thought process being converted to compilable code in project
form.

But back to the practical...I really need to set up my home/business systems
with SQL 2000 and get familiar with XML returned as a query.  How much
machine do I need to run SQL 2000.  I have a 266/128 meg box available.  Is
that enough?

I also need to perfect an XML tag parser I wrote, or bail to the 1 megabyte
MSDOM, which I'd rather not.  Would anyone want to see and critique the
code?...it's brute force VB string stuff.  I found a couple of logical flaws
and would like to re-visit the design.  I've got the time.  What would be a
workable parsing technique.  Links? Suggestions?  The present design is at
vbxml.com in the snippets section under VB.

Lot's of words.  I'll try to learn compact precision for better
communication skills, too. :)

I'm glad to be surrounded by, as is so happens, my top three choices among
any VB coders I know.  That's factual, rather incredible.  It's twue it's
twue.

Thank you gentlemen.  I'm looking forward to how our future evolves.

Cheers,

DonEMitchell