**** What's oopsla up to?

The leading forum for innovative and thought-provoking ideas in
object-oriented programming for 21 years, oopsla is taking on today's
software challenges, attracting people interested in:

    * programmer productivity
    * secure and reliable software
    * changing hardware platforms
    * ultra-large scale systems

oopsla is the premier conference for publishing state of the art research, for
seeking comment on works in progress, for sharing practical experiences,  and
(frequently, we're proud to say) for presenting Turing Award lectures on
significant works. It is where industry experts and their academic peers
gather to

    * improve programming languages
    * refine the practice of software development
    * explore new paradigms

Through collaboration, diversity, and incubation we are overcoming the
challenges facing the software industry.

**** The conference isn't until October, what's happening now?

Right now we're looking for submissions for:
   * Posters
   * Demonstrations
   * Doctoral Symposium
   * Onward! Tools and Films
   * Student Research Competition

Visit http://www.oopsla.org/2007/submissions to see how you can contribute.

Furthermore, the accepted workshops are looking for submissions. Visit
http://www.oopsla.org/2007/workshops to see the list of accepted workshops.
Submissions to workshops should be to the workshop organizer.

All year long the conference committee works hard to organize the program.
People like you — developers, researchers, industry experts, and academics
from around the world — contribute most of the content. Contributing provides
tremendous satisfaction and pride in sharing your knowledge and insight. A
bonus is the chance to meet a like (or unlike) mind who's worked on similar
problems, providing an opportunity for feedback and brainstorming solutions
to sticky problems.

Contribute to oopsla and you will be enriched as you enrich the world of
software.

**** And when are those submissions due?

July 2, 2007 - Submission Deadline for Posters, Demonstrations, Doctoral
Symposium, Onward! Tools and Films, Student Research Competition, and Student
Volunteers

October 21, 2007 - Conference starts

Contact the workshop organizer for deadlines of that specific workshop.

In addition, there are four symposia that will be co-located with OOPSLA 2007.
These may be appropriate if you want to target a very specific audience. Here
are their web sites:

* Dynamic Languages Symposium
   --- http://www.swa.hpi.uni-potsdam.de/dls07/
* International Symposium on Memory Management
   --- http://www.eecs.harvard.edu/~greg/ismm07/
* WikiSym
   --- http://www.wikisym.org/ws2007/
* Library-Centric Software Design Symposium
   --- http://lcsd.cs.tamu.edu/2007/

--
For more information, visit http://oopsla.org

Program Chair
David F. Bacon, IBM
papers@...

Onward! Chair
Cristina Videira Lopes, UC Irvine
onward!@...

Conference Chair
Richard P. Gabriel, USA
chair@...

Are you still meeting every third friday during the summer?

detour11 wrote:
> Are you still meeting every third friday during the summer?
>

The meetings are on informal hiatus at the moment; we seems to have run
out of speakers / topics.  Also, agile/XP ideas seems to be spreading
through the industry enough that they come in to play at other groups also.

Kyle

> The meetings are on informal hiatus at the moment; we seems to
> have run out of speakers / topics.  Also, agile/XP ideas seems
> to be spreading through the industry enough that they come in
> to play at other groups also.

I wonder if perhaps this group should morph into something else then.
I was thinking that there should be a group concentrating on advanced
programming topics. Things like source control, frameworks, patterns,
languages, etc.

Craig

So I ran across this the other day.  I saw it at http://behaviour-
driven.org/.  Would anyone venture an opinion on this?

It looked like an interesting addition to how TDD defines solving a
problem.  Possibly, it could help with customer interaction?  Offering
an approachable set of terms for someone not familiar with agile?

What does everyone else think?

Oliver

--- In xpstl@yahoogroups.com, "Craig Buchek" <yahoo@...> wrote:
>
> > The meetings are on informal hiatus at the moment; we seems to
> > have run out of speakers / topics.  Also, agile/XP ideas seems
> > to be spreading through the industry enough that they come in
> > to play at other groups also.
>
> I wonder if perhaps this group should morph into something else then.
> I was thinking that there should be a group concentrating on advanced
> programming topics. Things like source control, frameworks, patterns,
> languages, etc.
>
> Craig
>

As a young developer i would love to share ideas on advanced
programming topics.  I'm unfortunately in a work position where i have
to be the lead in terms of implementing technologies such as source
control, frameworks, & patterns.  On a side note: isn't it frightening
how many developers STILL ignore architecture, testing frameworks,
source control, and application layering.  It leads me to be disgusted
most days.

Wes

"detour11" <w_Drone@...> writes:

> On a side note: isn't it frightening how many developers STILL
> ignore architecture, testing frameworks, source control, and
> application layering.  It leads me to be disgusted most days.

In part, that's because so many of us have seen all of that done
wrongly.  Testing frameworks and source control, when done poorly, can
be a nuisance, but "architecture" and application layering can
absolutely destroy maintainability if put into the wrong hands.  I'd
say that most of this stuff is done wrong more often than it's done
right.

For instance, how many teams are still using CVS or SVN, despite their
known inadequacies?

--
Alan Shutko <ats@...> - I am the rocks.
My teacher can always tell when I start on a project the night before.

"For instance, how many teams are still using CVS or SVN, despite their
known inadequacies?"

Heck, we just switched from CVS to Subversion... what are the known
inadequacies?

aaron

> For instance, how many teams are still using CVS or SVN, despite their
> known inadequacies?

I'm using SVN - what known inadequacies?  We seem to be doing just
fine with it.  The only problem we ever run into is when two
developers edit a winform at the same time with VS 2005 because the
Windows Form Designer will rewrite huge sections of code depending on
where you place components, order columns, or style anything.  We are
a small team not needing or utilizing any branching features - perhaps
this is what you are referring to.

But hey.... any source control is better than no source control!!!

Aaron Hinni <aaron@...> writes:

> Heck, we just switched from CVS to Subversion... what are the known
> inadequacies?

The big one is merge tracking, which was on the docket for the first
release, but then deferred because it was hard.  It's now slated to be
included in v1.5, hopefully.  I'll believe it when I see it.

I'd personally include some other stuff like the lack of any defined
policy around branching and whatnot, but without merge tracking, the
fact you can do O(1) branches is not very useful.

For small teams, this doesn't matter much.  For larger teams, it's a
major problem.  CVS and SVN basically make branching for feature
development a huge hassle because merging is so hard to do.  In
contrast, the newer systems like arch, bazaar, mercurial, darcs, etc
make merging so easy you don't have to waste a tech lead's time doing
it.

--
Alan Shutko <ats@...> - I am the rocks.
If you obey all the rules, you miss all the fun. - K. Hepburn

detour11 wrote:
>> For instance, how many teams are still using CVS or SVN, despite their
>> known inadequacies?
>>
>
> I'm using SVN - what known inadequacies?  We seem to be doing just
>
> ....
>
>  We are
> a small team not needing or utilizing any branching features - perhaps
> this is what you are referring to.
>
>

Branching and merging is far more useful than it first appears, if you
have a tool that does both operations quickly and well.  Something that
might not be immediately obvious is that as you develop, regardless of
what tool you use, you are in fact branching (by editing your local
files, for example).

SVN branching is OK, but not great.  SVN's "cheap copy" branches are
much cheaper than in some other systems, but much slower (more
"expensive") than others.  SVN merging is very weak, because SVN does
not actually know you merged, so it can't help re-merge later changes
from the same branch,  etc.

I have some idea flopping around for a talk on a more modern whizbang
source control tool, I don't know which one is most amenable to such a talk.

In the meantime, you might have a look at what Linus Torvalds has to
say, I wrote about that here:

http://kylecordes.com/2007/05/17/linux-git-distributed/

Beware that he really, really does not like SVN.

> fine with it.  The only problem we ever run into is when two
> developers edit a winform at the same time with VS 2005 because the
> Windows Form Designer will rewrite huge sections of code depending on
> where you place components, order columns, or style anything.
>
>

Unfortunately, there isn't much any source control tool can do about
this.  It seems that many dev tool file formats are hostile to something
as simple as "diffing" the history of change by one person to one file.
A rule that tools should follow is that a small change in the thing
being designed, results in a small change in the file which describes
the thing.  Tools that produce big changes in the file for small changes
in the artifact get in the way of many sound practices, including
obviously important ones like code review.

--
Kyle Cordes
http://kylecordes.com

--- In xpstl@yahoogroups.com, Kyle Cordes <kyle@...> wrote:
>
> detour11 wrote:
> >> For instance, how many teams are still using CVS or SVN, despite
their
> >> known inadequacies?
> >>
> >
> > I'm using SVN - what known inadequacies?  We seem to be doing just
> >
> > ....
> >
> >  We are
> > a small team not needing or utilizing any branching features - perhaps
> > this is what you are referring to.
> >
> >
>
> Branching and merging is far more useful than it first appears, if you
> have a tool that does both operations quickly and well.  Something that
> might not be immediately obvious is that as you develop, regardless of
> what tool you use, you are in fact branching (by editing your local
> files, for example).
>
> SVN branching is OK, but not great.  SVN's "cheap copy" branches are
> much cheaper than in some other systems, but much slower (more
> "expensive") than others.  SVN merging is very weak, because SVN does
> not actually know you merged, so it can't help re-merge later changes
> from the same branch,  etc.
>
> I have some idea flopping around for a talk on a more modern whizbang
> source control tool, I don't know which one is most amenable to such
a talk.
>
> In the meantime, you might have a look at what Linus Torvalds has to
> say, I wrote about that here:
>
> http://kylecordes.com/2007/05/17/linux-git-distributed/
>
> Beware that he really, really does not like SVN.
>
> > fine with it.  The only problem we ever run into is when two
> > developers edit a winform at the same time with VS 2005 because the
> > Windows Form Designer will rewrite huge sections of code depending on
> > where you place components, order columns, or style anything.
> >
> >
>
> Unfortunately, there isn't much any source control tool can do about
> this.  It seems that many dev tool file formats are hostile to
something
> as simple as "diffing" the history of change by one person to one
file.
> A rule that tools should follow is that a small change in the thing
> being designed, results in a small change in the file which describes
> the thing.  Tools that produce big changes in the file for small
changes
> in the artifact get in the way of many sound practices, including
> obviously important ones like code review.
>
> --
> Kyle Cordes
> http://kylecordes.com
>

Thanks for the input.  Sometimes all of this just seems overwhelming.

Kyle Cordes wrote:
> SVN branching is OK, but not great. ... SVN merging is very weak, because SVN
does
> not actually know you merged, so it can't help re-merge later changes
> from the same branch,  etc.


My team has been using svnmerge for a couple years and it does exactly
that: track which revisions have been merged from each branch. You can
tell it to merge specific revisions or all that haven't been merged. It
tracks which revisions have been merged in an svn property called
svnmerge-integrated.

http://www.orcaware.com/svn/wiki/Svnmerge.py

It's a little crude but works well enough for simple branching strategy.

That said, I'm interested in learning more about git after watching
the video of Linus's talk.

Mike Jorgensen

Mike Jorgensen wrote:
> My team has been using svnmerge for a couple years and it does exactly
> that: track which revisions have been merged from each branch. You can
> tell it to merge specific revisions or all that haven't been merged. It
> tracks which revisions have been merged in an svn property called
> svnmerge-integrated.
>
> http://www.orcaware.com/svn/wiki/Svnmerge.py
>
>

Thanks.  At first look, SVN+svnmerge appears to be a much more pleasant
toolset than SVn alone.  I'm going to point it out to our team here and
also on my web site.

--
Kyle Cordes
http://kylecordes.com

This is exactly the kind of discussion I was hoping for. It seems like
there's quite a bit of interest in it too.

What do people think about starting a new group to discuss such
advanced programming topics? Would anyone be interested in getting
together to do the work to start up the group?

Craig

On 6/27/07, Mike Jorgensen <mikejorgensen@...> wrote:
>
>
>
>
>
>
> Kyle Cordes wrote:
>  > SVN branching is OK, but not great. ... SVN merging is very weak, because
SVN does
>  > not actually know you merged, so it can't help re-merge later changes
>  > from the same branch,  etc.
>
>  That said, I'm interested in learning more about git after watching
>  the video of Linus's talk.

Mike, can you post the link to Linus' video about git?

Thanks
Ed

--
Ed Howland
http://greenprogrammer.blogspot.com
"The information transmitted is intended only for the person or entity
to which it is addressed and may contain proprietary, confidential
and/or legally privileged material. Any review, retransmission,
dissemination or other use of, or taking of any action in reliance
upon, this information by persons or entities other than the intended
recipient is prohibited. If you received this in error, please contact
the sender and delete the material from all computers."

On 6/27/07, Kyle Cordes <kyle@...> wrote:
>
>  > fine with it.  The only problem we ever run into is when two
>  > developers edit a winform at the same time with VS 2005 because the
>  > Windows Form Designer will rewrite huge sections of code depending on
>  > where you place components, order columns, or style anything.
>  >
>  >
>
>  Unfortunately, there isn't much any source control tool can do about
>  this.  It seems that many dev tool file formats are hostile to something
>  as simple as "diffing" the history of change by one person to one file.
>  A rule that tools should follow is that a small change in the thing
>  being designed, results in a small change in the file which describes
>  the thing.  Tools that produce big changes in the file for small changes
>  in the artifact get in the way of many sound practices, including
>  obviously important ones like code review.
>

<dons asbestos suit>
This is a huge problem with VB6. We essentially ended up with only one
PC and abandoned CVS other than to store backup copies of the source.
This leads me to the belief that VB6 is just a toy language not meant
to be used in professional development shops. Yet I know that many
projects exist which say otherwise.
</dons asbestos suit>

Ed

>  --
>  Kyle Cordes
>  http://kylecordes.com
>
--
Ed Howland
http://greenprogrammer.blogspot.com
"The information transmitted is intended only for the person or entity
to which it is addressed and may contain proprietary, confidential
and/or legally privileged material. Any review, retransmission,
dissemination or other use of, or taking of any action in reliance
upon, this information by persons or entities other than the intended
recipient is prohibited. If you received this in error, please contact
the sender and delete the material from all computers."

On Jun 27, 2007, at 11:23 PM, Ed Howland wrote:
> Mike, can you post the link to Linus' video about git?

Kyle already did, with lots of analysis and notes:

http://kylecordes.com/2007/05/17/linux-git-distributed/

http://www.youtube.com/watch?v=4XpnKHJAok8

Kyle,

Thanks for the information re: SVN, git and mercurial.
It was a very interesting video and now I'm going to have to check out git.
Linus makes a pretty persuasive argument.  A number of the things he talks
about pertain to any source control system.

Thanks,

Bryan
http://www.dbBear.com/

----- Original Message -----

From: oferrigni

To: xpstl@yahoogroups.com

Sent: Tuesday, June 26, 2007 6:08 PM

Subject: [!! SPAM] [xpstl] Behavior Driven Development

 

So I ran across this the other day. I saw it at http://behaviour-
driven.org/. Would anyone venture an opinion on this?

It looked like an interesting addition to how TDD defines solving a
problem. Possibly, it could help with customer interaction? Offering
an approachable set of terms for someone not familiar with agile?

What does everyone else think?

Oliver

----- Original Message -----

From: Bryan - K0EMT

To: xpstl@yahoogroups.com

Sent: Monday, July 02, 2007 2:02 AM

Subject: RE: [xpstl] Re: Meetings?

 

Kyle,

Thanks for the information re: SVN, git and mercurial.
It was a very interesting video and now I'm going to have to check out git.
Linus makes a pretty persuasive argument. A number of the things he talks
about pertain to any source control system.

Thanks,

Bryan
http://www.dbBear.com/

Jeff Grigg wrote:
> I think that the downside to 'git' and related distributed tools is
> that you have to accept the idea of summary checkins -- "This is the
> checking of everything those people over there have done over the past
> <hopefully short period of time>."  Should people

This is not true - rather, what you have described as a problem with git
is actually a key problem that git solves (where git = git itself and
other similar tools).

* Using a distributed source control system, everyone can use the tool
end to end all the time. With a tool like git, you will be *more* able,
rather than less able, to track in detail who made what change.

* Because these tools work so much faster, they facilitate much more
detailed checkins, which encourages smaller, more details checkins, (and
thus more auditable detail of the change steps), rather than the
coarse-grained checkins that so often happen with traditional
centralized tools.

* With much stronger branch support that you are used to, it's more
feasible to keep the changes for feature 123 separate from those for
feature 456. This a a very good thing if you're trying to work on
several features at once while maintaining traceability.

* With the tools that have good "signing" support (git, for example),
there will be much more reason to trust that the audit trail is
correct.  Most centralized tools can have history changed (!) by anyone
with sufficient access to the centralized server.  A trail of SHA1s is
far more trustworthy than a vague notion, or even a pile of signed
affidavits, of "we think noone has ever modified the history".


(By the way, there are countless proposed good ideas that are shot at
with "but we can't do that because of SOX / HIPPA / DOD #nnn / CMM /
pick-your-acronym"... as a general rule, attacks of this kind are FUD.
Powerful tools can be very *helpful* in meeting onerous requirements.)


--
Kyle Cordes
http://kylecordes.com

"Jeff Grigg" <jeffgrigg@...> writes:

> I think that the downside to 'git' and related distributed tools is
> that you have to accept the idea of summary checkins -- "This is the
> checking of everything those people over there have done over the
> past <hopefully short period of time>."

That seems to be specific to only certain distributed tools.  Arch
(ie, tla and perhaps derivatives like bazaar) do this sort of thing,
where if you're merging changesets from other repositories you can
diminish history to summary.  (Iirc, tla recommends that as a way to
reduce history explosion.)

Darcs, on the other hand, keeps the full knowledge of which person
originally created which changeset.  So you can still trace
line-by-line.

In a corporate environment, this is still quite doable.  You won't be
importing from completely untrusted repositories.  More likely, the
developers you employ will be able to keep branches and you'll be able
to control the merges that way.

--
Alan Shutko <ats@...> - I am the rocks.
Would You Like A Vienna Sausage! - Stimpy.

Kyle Cordes <kyle@...> writes:

> * With much stronger branch support that you are used to, it's more
> feasible to keep the changes for feature 123 separate from those for
> feature 456. This a a very good thing if you're trying to work on
> several features at once while maintaining traceability.

This is the key feature I wish I had right now at work.  We've got

* Feature A, targeted at the June release that got backed out at the
   last moment and retracted for July.

* Feature B, originally targeted for July, but now split off for a
   separate release (July part deux).

* Feature C, D, E, F, G all targeted at August.  Feature D may not
   actually make August.

Naturally, these are _all_ working around the same set of core
functionality from one of the most successful projects last year, and
each feature has 2-5 developers working on it, or will once it gets
kicked off.  Some features have dependencies on other features.

With CVS, this is a mess.  Feature A was manually backed out since it
was worked on HEAD.  Feature B is on a branch from the beginning, but
is fortunately not requiring any of the fixes currently under
development.  The rest all require up to date fixes from HEAD, but
some may need to be delayed... as a result, we've got a few branches
and they're all dreading the point of merge.

With a distributed vcs, each of these projects could be on a branch.
They could easily and safely import distinct changesets from each
other (for example, infrastructure fixes as each project finds them).
They could do repeated merges from the HEAD to keep track of mainline
development so that they don't diverge much.  Then, when each is
stable and we know it can fit in the release, it can be merged into
HEAD but it won't cause the oceans of conflicts CVS does.

--
Alan Shutko <ats@...> - I am the rocks.
Men come with instructions, they're just written in pig latin.

I know Lean ain't XP but there are many synergies.

I am doing a webinar next week that should prove to be a good
introduction to Lean Software Development. Please go to
http://www.netobjectives.com/free-seminar-schedule/intro-lean-software-
development-webinar-jul-2007 to get more information and to register.
There is no charge for this, but you must register to attend.

Alan Shalloway
CEO, Net Objectives
Maximizing Product Development ROI through training, coaching and
consulting.

Greetings!

You now have the opportunity to attend the Gateway Software Symposium
either in March
or September every year!  The Fall Edition of the Gateway Software
Symposium will be
held in St. Louis on September 28-30th.  Join Us!

Event Name: Gateway Software Symposium 2007
Dates: September 28-30, 2007
Venue: Renaissance St. Louis Airport Hotel
URL: www.nofluffjuststuff.com/sh/2007-09-stlouis

The 2007 Gateway Software Symposium is coming to St. Louis
on September 28-30th. GSS 2007 will offer five (5) concurrent sessions
for you to choose from. The hot topics covered at GSS 2007 include:

* Spring 2.0
* Groovy/Grails
* OSGI
* Annotations
* Enterprise Performance & Scalability
* REST
* Behavior Driven Development
* GWT
* Java 6.0
* SOA/ESB
* Continuous Integration
* SEAM
* JRuby
* Agility and many more...

We have a great set of speakers lined up for you to enjoy featuring:

Andrew Glover, co-author of "Groovy in Action"
Kito Mann, author of "JSF in Action"
Neal Ford, author of Application Architect w/Thoughtworks
Ted Neward, author of "Effective Enterprise Java"
Howard Lewis Ship, creator of the Tapestry Framework
Jeff Brown, Grails Committer
John Heintz, AOP Expert
Michael Nygard, author of "Release It"
Mark Volkmann, Software Architect
David Husmman, Agility Expert & Coach
Scott Delap, author of "Desktop Java Live"


The No Fluff Just Stuff Symposium series is regarded as the premier
Java/Agility event series anywhere serving over 19,000 attendees with
some 115 events since 2002. The popularity of the NFJS symposium series
can be traced to the following:

1). Exceptional Speakers
2). Limited Attendance - capped at 250 people
3). No Vendors, No Sales Pitches, no Marketecture
4). Excellent networking opportunities with speakers and fellow attendees
5). The Best Value in the Java conferencing space period

********************************************************************************\
**********************

Early Bird Registration: $750/person good thru 9/10/07

Special $50 discount available to all XPSTL Members, use the discount
code, nfjsusergroup50 when registering.

Excellent Group Discounts Available - bring your entire development team
to the show good thru 9/10/07

5-9 Attendees: $675/person
10-14 Attendees: $650person
15-24 Attendees: $625/person
25-over Attendees: $600/person

Great Swag in 2007 - all attendees receive a new 2007 NFJS laptop bag &
custom leather binder.

Excellent Giveaways - Sony Playstation 3 and Apple iPods

Join us for a great show! We appreciate your support and patronage!!

Gateway Software Symposium 2007:
www.nofluffjuststuff.com/sh/2007-09-stlouis

Venue: Renaissance St. Louis Airport Hotel

All the Best,

Jay Zimmerman
NFJS 2007 Symposium Director
jzimmerman@...

I have been a member on here since 2002 but never really introduced
myself.
Currently I am a recruiter at Stockell Consulting and have been here
for a little over a year.  One thing that has jumped out at me in my
time here at Stockell is the tenure of the account managers and the
depth of the relationships they have with their clients. One of them
has been here for 10 years and the other 8 years and as many of you
know that is a long time in consulting.  We have
employees/consultants that have been with us 20 years.  So what I am
trying to do is find individuals that are interested in seeing what
is available in the XP field here in St. Louis. We need people from
mid to senior level for contract and perm positions and can pay up to
110k with full benefits based on experience, more if you are looking
for contract positions.

Thanks for taking the time to read through this and please email with
any questions.


Jeremy Locke
Senior Technical Recruiter
Stockell Consulting
15400 South Outer Forty Dr.
Chesterfield, MO 63017
jlocke@...
636-777-7154 Direct
636-537-9100 ext. 167
314-223-4847 Cell

Good afternoon.  My name is Jason Monastra and I am a new member of
this group.  I am a partner at United Global Technologies
(www.ugtechnologies.com) and a veteran recruiter of more than 9 yrs.
Currently I head business development and partnerships for new
recruiting venues over the net.

Our practice has seen a growth in the need for experienced
professionals with XP understanding and expertise.  Our client base
continues to grow in this area and we are seeking solid professionals
to both join our team on a full time basis as well as contract
opportunities.

Please be advised that we are currently looking for the following
professionals:

All of the below must have an understanding or background in agile
development practices.

Mid Level developers
Technical Team Leads
Business Analysts (functional and technical)
Project Managers

Pay is dependant on experience and location varies with a current
focus on Columbus, Cincinnati and Chicago.

Please feel free to contact me at:

United Global Technologies
10612-D Providence Road, Suite 320
Charlotte NC 28277
Jason R Monastra
Partner
704 644 5521 (direct)
704 200 6463 (cell)
888 831 2189 (toll free)

>>>  REGISTRATION NOW OPEN  <<<

                              OOPSLA 2007

                 22nd Annual ACM SIGPLAN Conference on
   object-oriented Programming, Systems, Languages, and Applications

                          October 21-25, 2007
                    Palais des Congres de Montreal
                       Montreal, Quebec, Canada

                           http://oopsla.org/

-----------------------------------------------------------------------

Registration is now open for OOPSLA 2007, convening October 21 through
25 in chic, sophisticated, metropolitan Montreal, Quebec.

NO ORDINARY CONFERENCE

OOPSLA is the premier conference for innovative and thought-provoking
ideas, for seeking comment on works in progress, and (frequently, we're
proud to say) for presenting Turing Award lectures on significant works.
It is where industry experts and their academic peers gather to:

     * Improve programming languages
     * Refine the practice of software development
     * Explore new paradigms

It’s the place where people devoted to the art and craft of software
development gather to share ideas, to push the boundaries, and to
learn and create.

WHAT'S GOING ON?

John McCarthy. Patti Maes. Gregor Kiczales. Fred Brooks. We all recognize
these as the names of researchers who have made profound contributions
to computing and software. They are also the headline speakers at this year’s
ooPSLA, where you’ll hear what they are thinking about in 2007. ooPSLA
invited talks have a well-deserved reputation for energizing software
developers with ideas that stretch our collective mind.

Jim Purbrick and Mark Lentczner. You may not recognize these names,
but you might know their alter egos: Babbage Linden and Zero Linden,
of Linden Labs. They are the creators of Second Life, a 3D virtual
world that is changing how we think about the real world, how we live
in multiple worlds, and how we do business in the 21st century. Purbrick
and Lentczner are the Onward! keynote speakers.

Peter Turchi. He is the director of the MFA Program for Writers at Warren
Wilson College. You certainly don’t recognize his name. Turchi is the
award-winning author of Maps of the Imagination, an exploration of the
relationships between creative writing, making maps, and creativity. He is
our keynote speaker, and perhaps the most mind-bending of all our speakers.

That's just a sampling of what makes OOPSLA the conference of choice
for software technologists---from recognized academics to
undergraduate students, from industrial researchers to developers and
managers, from the creators of technology to its users.

We are proud to offer you ooPSLA and look forward to sharing the experience
with you in Montréal this fall.

-----------------------------------------------------------------------
Co-located events:

DLS 2007 (http://www.swa.hpi.uni-potsdam.de/dls07/)
ISMM 2007 (http://www.eecs.harvard.edu/~greg/ismm07/)
WikiSym 2007 (http://www.wikisym.org/ws2007/)
LCSD 2007 (http://lcsd.cs.tamu.edu/2007/)
APL 2007 (http://www.sigapl.org/apl2007.html)
Mini PloP (http://refactory.com/OOPSLA_Workshop.html)
CompFrame 2007 (http://www.compframe.org/compframe2007/)

-----------------------------------------------------------------------

For more information and to register, go to http://oopsla.org/.

Gail E. Harris, OOPSLA 2007 Advertising Chair

I've got a web site (in PHP) that has a countdown timer that shows up
the last 3 days before a deadline for user submissions. It was written
by someone else, and the time/date calculations were done very poorly
-- it didn't work correctly last session. So I'm refactoring it to
clean it up a lot and fix the errors.

I'd like to add some tests to the code, to make sure I get it right
this time -- BEFORE I deploy it. But how can I test it thoroughly (and
automatically) if the code depends on when "NOW" is? It seems wrong to
have to pass in "NOW" to all the functions I'm using.

Any ideas or pointers?

Thanks,
Craig