[comp.lang.modula2] JPI V3.0... will they get it right?????

zerr@schaefer.math.wisc.edu (Troy Zerr) (02/18/91)

Those of you who do development on PC's are probably familiar with the 
problems with JPI's Modula-2 V. 2.0 compiler.  The compiler had far too
many flaws.  Some were just plain bugs, and the documentation was the
worst compiler documentation that I have ever seen,
but the most severe problems were design defects which made using the 
compiler with non-JPI ANYTHING almost impossible.
 
Those of you that gave up on v.2 and went back to v.1.17 might find
the following blurb from the JPAM (JPI) bbs interesting:

------------------------------------------------------------------------------
   I am pleased to be able to relay JPI's announcement, made just today, of
the V3 family; new M-2 and PASCAL and C++ joining the family.  Spoke with
Jesper, and the various comments we all made seem to have made it across to
London and been taken into account.  Jesper recognised me by my namecard.
   I have great hopes for V3 stuff once the final production hurdles are
overcome.  ANd this is not a premature release.  The Version 3 family was
offically announced today, C++ manuals and so on were in evidence.  Version
3 is very real, and not a rumor.
   What I am hoping is that after the Versino 2 hiccup, Version 3 is very
real and eliminates the errors that were made with Version 2.  Jesper
indicated that a *lot* of work has gone not only into TopSpeed interoperability
but also into interoperability of TopSpeed languages with the Microsoft .OBJ
and Microsoft .LIB world.
   More Version 3 news: Classes now support multiple inheritance, and other
beef has been added into things.
   In short, the comments we have made here were heard.
   So  Version 3 has been promised to ship within a month, with Software
Development '91 being the announcement place and time.
   If it measures up to the development team's expectation, you'll be able
to waltz from language to language as the need arises, and get aftermarket
libraries from just about anywhere.  Also, I saw a demo running under Windows
3.0.  That seems to be part of the new release package.
   All in all, very promising.
------------------------------------------------------------------------------

Could it be true??  Could JPI have gotten it right??  Do you remember 
where you left your "Free Upgrade" coupon??  

Stay tuned.


-- Troy Zerr
zerr@math.wisc.edu
uunet!math.wisc.edu!zerr
 

GRANGERG@VTVM1.BITNET (Greg Granger) (02/18/91)

On Sun, 17 Feb 91 18:55:57 GMT <editor%ucf1vm.BITNET@LILAC.BERKELEY.EDU> said:
>Those of you who do development on PC's are probably familiar with the
>problems with JPI's Modula-2 V. 2.0 compiler.  The compiler had far too
>...
>Could it be true??  Could JPI have gotten it right??  Do you remember
>where you left your "Free Upgrade" coupon??
>...

I use version 2 of the JPI M2 compiler and for the most part it does
work.  I doubt "JPI could have gotten it right", because:
   1) They haven't had time to make the kinds of changes needed.
   2) The biggest problem they have can never now be addressed
      Version 1 of there compiler was a simple and elegance
      product, it could have used a few 'enchancements' but all
      in all it was what M2 is all about.  Version 2 added
      mega-media-hype-crap (not what M2 is about), and did a
      poor job of it at that.  Course once it's in there it
      can't be removed...  I will upgrade and give it a chance
      (I do like the product, I'm just disappointed with it's
      lack of direction/innovation), but if it's more (and more
      and more) of the same I'm likely to nuke it and put version
      1 back on (and save 2 meg of disk space).
      <big sigh>
      It's a shame the could find the people who produced version
      1 and hire them back <grin> ...
Greg