Minutes of FIGGY BAR RT Conference.

Date: 07/05/90   Time: 22:27EDT

Attendees:
[GARY-S]
[[Kent] K.PETERSON5]
[[Kevin] APPERT]
[[Dennis] D.RUFFER]
[[Wil] W.BADEN1]
       
   The question of the night: Was Mitch Bradley right ? Are some of us 
   making too big a deal about 'minor' changes as proposed by ANS Forth ?
   That comprised 90% of the discussion. Also discussed were a bug found
   in Pygmy (TIB) and Speech Synthasizers.

Minutes:
<K.PETERSON5> is here.
<APPERT> is here.
<GARY-S> Hi - just posting the notice
<GARY-S> Are some of us making too big a deal out of 'minor' changes as 
proposed by ANS Forth ?
<GARY-S> Kev, Kent ?
<[Kent] K.PETERSON5> Well, a lot of people are reluctant to change what they 
are used to...
<[Kent] K.PETERSON5> or what feels "right" to them, but for the sake of getting 
SOME standard...
<[Kent] K.PETERSON5> comprimises have to be made.
<[Kevin] APPERT> it has always been my feeling that there will be other Forths 
besides ANS Forth...
<[Kevin] APPERT> and it's hard for me to get excited about the ANS effort.  
What's an example of a 'minor change' ?
<GARY-S> NOT is not not - it is invert
<GARY-S>   the other bitch is /
<[Kevin] APPERT> for all the emotions and effort put in to both sides, as far 
back as Brodie's "starting Forth"  and before, I don't think you can call
that minor
<GARY-S> Mitch does
<[Kevin] APPERT> floored division?
<[Kent] K.PETERSON5> But we could take the standard, load a bunch of our own 
favorite words on top
<[Kent] K.PETERSON5> and feel fairly comfortable, I think
<[Kevin] APPERT> a wise man once told me that he defined NOT  in the following 
fashion:
<[Kevin] APPERT> : NOT true abort" Don't use this, nobody knows what it means " 
;
<[Kevin] APPERT> this was about 10 years ago, or so
<[Kent] K.PETERSON5> I like it! :-)
<GARY-S> The question really is, 'Is there too much wailing and gnashing of 
teeth ? OR... is this an effort to squelch opposition - opposition required
by ANS
<[Kent] K.PETERSON5> A friend of mine said he gave up monitoring the ASNI 
forth traffic. Too much volume from griping.
<GARY-S> Is it all that different from 79 and 83, in regards to complaints ?
<D.RUFFER> is here.
<[Kent] K.PETERSON5> I think people believe (or hope) ANSI carries more weight.
<GARY-S> answer the question Dennis
<[Dennis] D.RUFFER> YES!!!!!!!!!
<GARY-S> Don't hedge, say what you really mean . :-)
<[Dennis] D.RUFFER> the problem I see was created with the 83 Std!
<[Dennis] D.RUFFER> the ANS efforts are merely trying to recover and resolve 
them
<[Wil] W.BADEN1> is here.
<GARY-S> By going back to 78 (and before ) ?
<[Dennis] D.RUFFER> no, by making the controversies go away, by getting rid of 
the controversial words or limiting their functions to what we can agre
e on
<GARY-S> 1's compliment ? Get real !
<[Dennis] D.RUFFER> ie, NOT was controversial so it is gone
<[Dennis] D.RUFFER>  / is controversial so it is also gone
<[Dennis] D.RUFFER> where is 1's complement machines used or even considered in 
the standard?
<[Kent] K.PETERSON5> What will be left when all the controversy is done? DUP & 
SWAP??
<GARY-S> two keystrokes is better for divide ?
<[Dennis] D.RUFFER> there are some on the TC who would love to standardize DUP 
and leave it at that
<[Dennis] D.RUFFER> two keystrokes?? How do you mean gary?
<GARY-S> what is the new definition for divide ?
<GARY-S> Hello wil - care to jump in ?
<[Dennis] D.RUFFER>  / is defined for positive operators just as it has always 
been
<GARY-S> did you not say it was gone three replies before
<[Kent] K.PETERSON5> You said / is gone. Did you mean \ ?
<[Dennis] D.RUFFER> gone in the negative domains
<[Kent] K.PETERSON5> Oh
<[Kent] K.PETERSON5> So, the flooring arguments have been brushed aside?
<GARY-S> I should have been more explicit - is there another language that uses 
different operators for the same function if negative/positive ?
<[Wil] W.BADEN1> Mitch is right that fuss shudnt be made over small matters 
(like name changes.) But CATCH, THROW, etc., are major matters.
<[Dennis] D.RUFFER> if you want to use negative operators in / you must define 
it yourself using one of the 3 divide operators
<GARY-S> I understand Len Morgenstern's proposal wasn't even considered - only 
CATCH and THROW
<[Dennis] D.RUFFER> Len's proposal was not even submitted as far as I know (I 
have not seen it)
<[Wil] W.BADEN1> In C division is undefined for negative numbers.
<[Wil] W.BADEN1> C therefore has 3 operations for integer division: /  div  
ldiv.
<GARY-S> What would you like to see with regards to CATCH/THROW ?
<[Dennis] D.RUFFER> I would like to see something more constructive than just 
saying that it is bad
<[Dennis] D.RUFFER> it is NOT true that the concept of error control is new to 
Forth
<[Dennis] D.RUFFER> I use 'ILDE and now 'ABORT vectored words in polyFORTH all 
the time
<[Dennis] D.RUFFER> CATCH/THROW just give me a more generic format for it
<[Wil] W.BADEN1> I was going to present @JUMP and !JUMP (analogous to setjmp 
set) but decided that no such proposal can be portable.
<[Wil] W.BADEN1> All the objections to SP@ SP! RP@ RP! apply to CATCH THROW and 
setjmp.
<GARY-S> There seems to be some strong sentiment for setjmp in clf, Wil
<[Dennis] D.RUFFER> sentiment, no matter how strong, is just not good enough.
<[Kevin] APPERT>  how 'bout saving that stuf for FORTH 2000, when we can call 
it existing practice!
<[Dennis] D.RUFFER> we need some well thought out proposals for an alternative
<[Wil] W.BADEN1> You cannot save context with just pointers.
<[Dennis] D.RUFFER> CATCH/THROW meet the existing practice requirements in some 
systems
<[Dennis] D.RUFFER> just because you don't use something does not mean that it 
does not exist
<[Wil] W.BADEN1> Not Johns Hopkin, and RTX.
<GARY-S> what makes you think I have , or is this a generic argument, Dennis?
<[Dennis] D.RUFFER> the lack of existing practice argument has been bandied 
about quite often recently, I don't think any of us (or anyone for that ma
tter) can make that claim without examining ALL version of Forth
<[Wil] W.BADEN1> Forth 2000 is going to be "Tiny C With Postfix Operators."
<GARY-S> I believe that !
<[Kevin] APPERT> I believe Wil is trying to get a rise
<[Wil] W.BADEN1> No, just extrapolating ANSI proposals.
<GARY-S> Dennis, since when is it required that ANYONE examine ALL existing 
systems before having a right to voice an opinion about X3J14 ?
<[Dennis] D.RUFFER> opinions are fine Gary, some of the comments that I have 
read here in the BB have just gone beyond the line.
<[Wil] W.BADEN1> The applications for which Forth is The Best do not need CATCH 
& THROW, or local variables.
<[Kevin] APPERT> do they need a standard, Wil
<[Kevin] APPERT>  ?
<GARY-S> Excuse me, but this is an open forum and ANS is an open committee. I 
don't know if you are refering to me or comments in general; but I believe
they have a place here. Dark corners yield bad decisions.
<[Wil] W.BADEN1> Yes, to stop a recurrence of the Forth 83 debacle.  F83 was 
one of the best things to happen to Forth; Forth 83 The Worst.
<[Wil] W.BADEN1> All this / MOD NOT controversy comes right from Forth 83.
<[Dennis] D.RUFFER> I'm sorry if I seem to be yelling at you specifically Gary. 
 I do not mean to. I just have seen enough Forth to know that the exis
ting practice argument is stupid when it comes to Forth
<GARY-S> I'm not taking personal offense either. I just maintain discussion is 
healthy.
<[Dennis] D.RUFFER> I do not know of one application I have written in Forth 
that did not need an error control mechanism Wil.
<[Wil] W.BADEN1> True, but ABORT" will get you far.
<[Kevin] APPERT> why is it in the charter of the committy, Dennis?  Existing 
practice would have stopped the floored div. mess of F83!
<[Dennis] D.RUFFER> discussion is more than just health Gary.  It is ESSENTIAL! 
 That is what was missing with the 83 Std.  It was done in one week by
 a bunch of self proclaimed experts who did not even ask for input.
<[Dennis] D.RUFFER> not far enough Wil, when the task has no terminal
<[Kevin] APPERT> took much longer than a week!  there was input solicited.
<[Kevin] APPERT> (in some cases ignored)
<[Dennis] D.RUFFER> that is exactly why it is in the charter kevin, to protect 
the interests of the vendors
<GARY-S> So - right or wrong, some of the 'fodder' you are seeing in the BB 
does have a place.
<[Dennis] D.RUFFER> yes, I encourage the debates Gary.  I just don't like the 
"no existing practice" argument.
<[Wil] W.BADEN1> Error recovery is easy to do when you know the particular 
architecture.
<[Dennis] D.RUFFER> I agree Wil, I do it all the time.  It would just be nice 
to be able to do it in a portable way.
<GARY-S> I *really* doubt portability will be the end result of X3J14
<[Dennis] D.RUFFER> it is not in the charter Gary, but it will promote it as a 
side effect
<[Wil] W.BADEN1> Just say that at most one value on each stack may be kept in a 
register.
<[Wil] W.BADEN1> Then restore SP@ SP! RP@ RP!
<GARY-S> I hope you are right
<[Dennis] D.RUFFER> That would remove all of Chuck's chips from complying Wil
<GARY-S> ABORT works for me, but I'm not trying to switch circuit loads or 
runway guidance
<[Dennis] D.RUFFER> I don't think I've ever been able to get away with the 
standard ABORT" in any useful application
<[Wil] W.BADEN1> I've said how to do it, and shown why CATCH THROW is wrong.
<[Dennis] D.RUFFER> write it in a proposal Wil, they have to at least talk 
about it then
<[Wil] W.BADEN1> Propose SP@ etc.?
<[Dennis] D.RUFFER> you will need to include rational as to how it can be done 
p'ortably and why the alternatives are bad for it to get any kind of su
pport
<[Wil] W.BADEN1> (I have a speech synthesizer program attached and I'm 
LISTENING to you.  You all talk funny.)
<[Dennis] D.RUFFER> well then squabble squabble lahfahfaskldhds
<GARY-S> set mine with a drawl, wil :-)
<[Kent] K.PETERSON5> How's the pronunciation of a name like Yvonne?
<[Wil] W.BADEN1> "eevun"
<[Kent] K.PETERSON5> Not bad. Mine did Yavun till I put in a filter.
<GARY-S> The phonetic abbreviations you use, probably suit a synthesizer very 
well, don't they ?
<[Wil] W.BADEN1> Yes.
<[Dennis] D.RUFFER> now all you need is the faces to go along with them
<GARY-S> I posted a complete list of known faces in the UNIX RT
<[Wil] W.BADEN1> Hey, get back to Forth.
<[Dennis] D.RUFFER> not the smiley kind, the ones VPC (I think it's called) 
displays in conferences like this when you talk
<[Kent] K.PETERSON5> I found a weird bug/feature in Pygmy today.
<[Dennis] D.RUFFER> all that's Forth, what bug Kent?
<[Kent] K.PETERSON5> TIB has an extra level of indirection. You have to do TIB 
@ to get the F83 TIB.
<[Dennis] D.RUFFER> ouch, that could be a nasty one for conversions
<[Kent] K.PETERSON5> It baffled me for a while.
<GARY-S> Leave e-mail to Frank, Kent. He REALLY does want that sort of input.
<[Dennis] D.RUFFER> I wonder who's standard he followed for that one
<[Kent] K.PETERSON5> I sent him a note just before I came here.
<GARY-S> great
<[Dennis] D.RUFFER> although, making it a variable is better than the games 
I've seen done with #TIB and >IN to reposition TIB
<GARY-S> Things are quiet - let's go home
<[Kent] K.PETERSON5> OK. Bye.
<[Dennis] D.RUFFER> did I shout everyone down?  I'm sorry!
<[Wil] W.BADEN1> What did we decide about the question of the day?
<GARY-S> Too much bitchin' but it has a place
<[Dennis] D.RUFFER> I'm glad to see Mitch is going toe to toe with it at least
<[Wil] W.BADEN1> Well, that's a positive note to end the  evening.
<GARY-S> Hey, we can't always run on sunshine and lollypops
=== End of Steno notes. ===