home *** CD-ROM | disk | FTP | other *** search
- Path: fohnix.metronet.com!not-for-mail
- From: milam@fohnix.metronet.com (Stan Milam)
- Newsgroups: comp.lang.c
- Subject: Re: Free Pacific C problem
- Date: 20 Jan 1996 18:57:55 -0600
- Organization: Texas Metronet, Inc (login info (214/488-2590 - 817/571-0400))
- Message-ID: <4ds32j$qor@fohnix.metronet.com>
- References: <4d9msn$k9p@fohnix.metronet.com> <tcpnntpd.16.1.14.14.45.42.2781597121.340150@the-fix.sos.on.ca> <4de4ik$n5o@fohnix.metronet.com> <3100d885.204542784@news.bne.aone.net.au>
- NNTP-Posting-Host: fohnix.metronet.com
- X-Newsreader: TIN [version 1.2 PL2]
-
- Clyde Smith-Stubbs (clyde@hitech.com.au) wrote:
- : On 15 Jan 1996 11:57:40 -0600, milam@fohnix.metronet.com (Stan Milam)
- : wrote:
-
-
- : >I agree with you. Once you figure it out and learn where the holes are it
- : >is a nice compiler. It also produces really tight, fast code, and the
- : >executables are *very* small. It would go a long way to:
- : >
- : >1. Get all of the ANSI manifest constants defined in the header files.
- : >2. Implement all of the ANSI required functions in the library (strtol(),
- : > memmove(), etc.).
-
- : I'm happy to fix any that are pointed out - I don't have the time at
- : present to go through the standard page by page (a lot of the library
- : was done back in the late '80s when the standard was in its late
- : drafts - the final standard added a few extra functions and macros -
- : this is the main reason they're not there).
-
- Seven years. Seems like enough time to me.
-
- : >3. Allow initialization of data structures such as arrays and structures
- : > that are of the auto storage class, as ANSI requires.
-
- : This is being done, chiefly because it happens to have become an issue
- : with one of our cross compilers. I personally suggest avoiding auto
- : aggregrate initialization - it's rarely the best solution to a given
- : problem.
-
- Oh really? I personally like it and use it a lot. Why is it "rarely the
- best solution"?
-