RfD: S>F and F>S version 1.1 First Revision 09 April 2012 Charles G. Montgomery <cgm@...> Discussions in response to the original version of...
810
Gerry
yipperoo
Apr 25, 2012 1:05 pm
Following a post on comp.lang.forth a few weeks ago, I've collected some floating point test programs together from different sources and placed them on the...
811
Peter Knaggs
pknaggs
May 8, 2012 7:19 pm
Parse Rational ============== 20120316 PjK Original Text Problem ======= The rationale for 6.2.2008 PARSE includes a discussion of the non standard word...
812
Anton Ertl
anton@...
May 9, 2012 11:58 am
... [...] Looks good. - anton...
813
Bernd Paysan
berndpaysan
May 9, 2012 5:11 pm
Problem: LEAVE and UNLOOP are defined to have no interpretation semantics, but an execution semantics. This implies that ['] LEAVE EXECUTE is equivalent to ...
814
David Harralson
msd1107
May 9, 2012 6:11 pm
Bernd brings up an interesting situation. Is LEAVE (and UNLOOP) compilation only? Interpreting the wording of the standard says no. It is up to the...
815
Bernd Paysan
berndpaysan
May 9, 2012 6:25 pm
... And since both provide meaningful error messages, and therefore catch the behavior, its users are happy with the current situation. ... It fails for the...
816
BruceMcF
brucemcf...
May 9, 2012 9:14 pm
... (20 years ... the return ... EXECUTE to ... Yes, I would rather that a small, compact implementation *be able to* rake the leaves during compilation and...
817
David Harralson
msd1107
May 9, 2012 10:00 pm
Bernd, I think the point here is that the error messages are, in actuality, the error. It appears that these particular implementors are assuming that LEAVE...
818
Bernd Paysan
berndpaysan
May 9, 2012 10:42 pm
... Actually, as defined in the standard, it is not - LEAVE has no interpretation semantics. ... LEAVE changes the control flow. Maybe you have found another...
819
Stephen Pelc
sfprem
May 10, 2012 5:07 am
Peter said ... The need for WORD has largely been eliminated by PARSE and PARSE-NAME. WORD is retained for backward compatibility. It was difficult to parse...
820
Andrew Haley
haleyandrew13
May 10, 2012 9:17 am
... Is this correct? ... Presumably not, because of the phrase "Continue execution immediately following the innermost syntactically enclosing DO ... LOOP or...
821
Anton Ertl
anton@...
May 10, 2012 11:58 am
... EXIT is also defined that way (and there are others). I asked the Forth-94 TC about ticking EXIT, and got an informal answer that a standard program...
822
Bernd Paysan
berndpaysan
May 10, 2012 1:54 pm
... Hm, funny reasoning. When we two argue about state-smartness, the fact that COMPILE, has no interpretation semantics is discussed away by saying that ... ...
823
Bernd Paysan
berndpaysan
May 10, 2012 9:21 pm
... D.6.7 implies that "compilation words" (which are not well specified there) can't be ticked, so that probably also includes all words with non-standard ...
824
mark wills
markrobertwills
May 10, 2012 9:33 pm
Making clear that something is ambiguous is an oxymoron. I can execute EXIT from the command line on my system just fine. No ambiguity. Nothing goes bang. ... ...
825
Bernd Paysan
berndpaysan
May 10, 2012 10:50 pm
... So you say "I don't know" conveys no meaning? A standard is a guide for system implementers, and it should be pretty unambiguous (i.e. not subject of long...
826
Peter Knaggs
pknaggs
May 10, 2012 11:39 pm
... Annex D is informative and therefore not actually part of the standard. I would not relay on any Annex to determine the meaning of the normative part of...
827
Mitch Bradley
wmb@...
May 11, 2012 2:15 am
... Except that the annexes do contain explanatory material that can be useful for understanding the thought processes that resulted in the normative text....
828
Elizabeth D Rather
erather
May 11, 2012 6:58 am
On 5/10/12 11:33 AM, mark wills wrote: Making clear that something is ambiguous is an oxymoron. "Ambiguous" in the context of a Forth standard is a technical...
829
Anton Ertl
anton@...
May 11, 2012 2:49 pm
... Yes, we can define ... and the result will have the same compilation semantics as the standard COMPILE, and interpretation semantics that are the execution...
830
Bernd Paysan
berndpaysan
May 11, 2012 7:57 pm
... No, of course not. The self-defined compile, inherits the same deficiencies the original compile, has. You are allowed to define this, but you can't ...
831
Hugh Aguilar
hughaguilar96
May 11, 2012 10:35 pm
________________________________ From:Â mark wills <markrobertwills@...> To:Â "forth200x@yahoogroups.com" <forth200x@yahoogroups.com>Â ...
832
Dennis Ruffer
daruffer
May 11, 2012 10:53 pm
Hugh, Ray is certain NOT dead! I wish you would realize how much your words confirm your ignorance. DaR From: forth200x@yahoogroups.com...
833
Hugh Aguilar
hughaguilar96
May 13, 2012 7:32 am
Well, my apologies to Ray Duncan then --- I had heard that he had passed away. Testra has the source-code for UR/Forth. I had believed that this was the only...
834
Peter Knaggs
pknaggs
May 13, 2012 7:34 am
... Giving this attitude, why are you even posting to this mail list. Why are you even a member? -- Peter Knaggs...
835
Anton Ertl
anton@...
May 14, 2012 12:55 pm
... There is also a little bit of run-time semantics of the colon definition and of ";", but that has no visible effect outside. ... Yes, if there is no...
836
Bernd Paysan
berndpaysan
May 14, 2012 1:43 pm
... Yes, and the execution semantics of COMPILE, does not have any meaningful interpretation semantics. The idea behind these words that have no ...
837
David Kuehling
dvdkhlng
May 14, 2012 2:06 pm
... You mean ... would not be meaningful and would not neccessarily work? I'd never have thought that. /me doesn't think that COMPILE, should be part of your...
838
Bernd Paysan
berndpaysan
May 14, 2012 2:44 pm
... Well, it currently is (since ANS Forth). In 1994, we had systems with state- smart words, we had systems with separate vocabularies for compilation (like ...
