From mboxrd@z Thu Jan 1 00:00:00 1970 X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on polar.synack.me X-Spam-Level: X-Spam-Status: No, score=-0.4 required=5.0 tests=AC_FROM_MANY_DOTS,BAYES_00 autolearn=no autolearn_force=no version=3.4.4 X-Google-Language: ENGLISH,ASCII-7-bit X-Google-Thread: 103376,cec20777e0d41ea0 X-Google-Attributes: gid103376,public X-Google-ArrivalTime: 2001-03-17 01:38:53 PST Path: nntp.stanford.edu!newsfeed.stanford.edu!news.tele.dk!195.25.12.36!oleane.net!oleane!isdnet!psinet-france!psiuk-f4!psiuk-p4!uknet!psiuk-n!news.pace.co.uk!nh.pace.co.uk!not-for-mail From: "Marin David Condic" Newsgroups: comp.lang.ada Subject: Re: Increased Interest In Ada? Date: Thu, 15 Mar 2001 13:18:57 -0500 Organization: Posted on a server owned by Pace Micro Technology plc Message-ID: <98r13a$6k9$1@nh.pace.co.uk> References: <3A82EFA2.C8756B09@acm.org> <970ma1$1l7$1@nh.pace.co.uk> <98inu2$1fr$1@nh.pace.co.uk> <98kjgv$3j69@news.kvaerner.com> <98lc03$23l$1@nh.pace.co.uk> <3AAED3D6.79CEAA7F@acm.org> <3AAFE454.B079E0F1@averstar.com> <98op0t$c4g$1@nh.pace.co.uk> <3AB0E994.A2B2D3C6@averstar.com> NNTP-Posting-Host: 136.170.200.133 X-Trace: nh.pace.co.uk 984680362 6793 136.170.200.133 (15 Mar 2001 18:19:22 GMT) X-Complaints-To: newsmaster@pace.co.uk NNTP-Posting-Date: 15 Mar 2001 18:19:22 GMT X-Priority: 3 X-MSMail-Priority: Normal X-Newsreader: Microsoft Outlook Express 5.50.4522.1200 X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4522.1200 Xref: nntp.stanford.edu comp.lang.ada:91491 Date: 2001-03-15T18:19:22+00:00 List-Id: Sounds interesting. Obviously, if the programmer doesn't see the C intermediary steps and just gets code out the back end, then its just another Ada compiler. However, I'd still have concerns about gluing it onto someone's embedded C development environment because of the post-compilation stuff. You need to play with the linkage editor & loader and it remains a question in my mind if that is just going to invisibly look like Ada or if you're going to need to treat it more from the intermediate C perspective. After that, you've got to deal with the possible existence of source-level debuggers and other testing tools. Are these going to see the Ada code - or go referencing back to the C code? I am not saying it can't be done or that doing it may not be of value. What I'm saying is that it would likely require quite a bit of investigation just to see what would be involved in doing the job and what quality the finished product would have. After that there is certainly a non-trivial amount of work in pulling the pieces together. If I was actively using the Z-World board and familiar with its development environment, a lot of the assessment would be already in my head. Without a funded mandate to go forth and kill, its hard to justify the time committment. :-) If we had a vendor interested in adding Ada as an additional front-end and was willing to pay to merge the tools, that would really be something. Unless I can talk my current employers into programming Cable-TV boxes in Ada, I'm not likely to find the time to look into it. (Work is the curse of the Language Advocate? :-) MDC -- Marin David Condic Senior Software Engineer Pace Micro Technology Americas www.pacemicro.com Enabling the digital revolution e-Mail: marin.condic@pacemicro.com Web: http://www.mcondic.com/ "Tucker Taft" wrote in message news:3AB0E994.A2B2D3C6@averstar.com... > Actually, one of the interesting things working with the version > of our Ada 95 technology that uses C as an intermediate is how > it illustrates exactly what you are giving up by going to C. > All of the consistency checks performed by Ada at compile-time, > plus the additional checks which are performed at run-time when > they can't be proved safe by the compiler, are generally all lost > when you write in C "by hand." By looking at the generated > C you immediately see all the run-time checks that remain, and > it makes me cringe to think that people writing by hand in C > don't have any of those safety checks performed. And that is *after* > our optimizer has already eliminated many of the provably-safe checks. > > For what it is worth, the "glued together" compiler is actually > quite easy to use, and it "feels" like a regular Ada compiler, except > that if you want to check up on what the compiler is doing, you > can save and look at the generated C code rather than having > to look at the generated machine code. By default, the intermediate > C code is deleted, so all that is left is a ".obj"/".o" just like > a "regular" compiler. >