• Squish under Linux

    From Serguei Trouchelle@2:550/357.1 to Bob Jones on Friday, November 14, 2003 13:23:10
    Would you try adding a subdirectory 'sqafix' (without the quotes)
    at the same level that the squish, slib, util, etc.

    Another question.
    Does sqafix sources exist in original Pete's package? I know it can be compiled
    with Watcom, but CVS doesn't have original makefiles.

    If anybody can send me these sources, please write to <stro@isd.dp.ua>

    Serguei Trouchelle
    --- timEd/386 1.10.y2k+
    * Origin: Was sollen wir trinken Sieben Tage lang... (2:550/357.1)
  • From Bob Jones@1:343/41 to Mark Lewis on Monday, November 17, 2003 16:32:06
    Mark:

    Appologies for the long delay in replying..... Unfortunately, it's not going to get better over the next 2 months.....

    Would you try adding a subdirectory 'sqafix' (without the
    quotes) at the same level that the squish, slib, util, etc.
    (typically .../maximus/sqafix ) already exist in your source
    code tree and try running your CVS update again? I suspect
    that you may have a user ID issue with creating
    sub-directories in the location where you are compiling the
    code and updating from CVS. :( Let me know what happens.

    added the directory with the same results... running
    everything as root so that should not be a problem...
    ended up removing everything and reinstalling maximus-
    3.03-etc.tar.gz and running

    Drats..... I may have to have you roll back ./configure to the version before I added sqafix. :(

    cvs -z3 -d
    :pserver:anonymoys@cvs.sourceforge.net:/cvsroot/maximus co maximus

    (instead of cvs up)

    and then cd into the maximus directory and run
    ./configure and then make build with a similar error...
    my redirection logging isn't working as i'd like but
    the screen output without redirection ends like so...

    sqafix.c:2180: warning: passing arg 2 of `DoDumpArea'
    from incompatible pointer type
    sqafix.c:2182: warning: passing arg 2 of
    `DoRunManualMode' from incompatible pointer type

    Ok. Someone is going to have to dig into DoDumpArea and DoRunManualMode calls and see what's up. I am guessing that I'll find a define or macro definition isn't correct for your platform. :(

    sqafix.c: At top level:
    sqafix.c:218: warning: `DoCheckKey' defined but not used
    sqafix.c:228: warning: `DoGetKey' defined but not used
    gmake[1]: *** [sqafix.o] Error 1
    make: *** [sqafix] Error 2

    i'm currently seeking info on redirecting everything to
    the log file... seems that

    program 2>&1 > logfile

    try 'program 1>name.log 2>name.err'... Unfortunately, that will seperate stderr from stdout lines.....

    Hmmmm......

    Thanks for the note. If you find a solution, let me know.

    Take care.....

    Bob Jones, 1:343/41



    --- Maximus/2 3.01
    * Origin: Top Hat 2 BBS (1:343/41)
  • From Bob Jones@1:343/41 to Serguei Trouchelle on Monday, November 17, 2003 16:37:58
    Would you try adding a subdirectory 'sqafix' (without the quotes)
    at the same level that the squish, slib, util, etc.

    Another question.
    Does sqafix sources exist in original Pete's package? I
    know it can be compiled with Watcom, but CVS doesn't
    have original makefiles.

    If anybody can send me these sources, please write to <stro@isd.dp.ua>

    For original sources, please contact Pete. He is registered on on the Maximus Sourceforge page, and I think e-mail to his sourceforge accout does get to him.
    If you do get source from Pete, please make sure you keep what is on the sourceforge site legal under the GNU public license.

    The main problem with the original SqaFix code was a couple of copyrighted source files that Pete had. From memory, I had to replace four files. One file was the CRC-16 / CRC-32 calculation. That file was believe to be PD, but I didn't have clear knowledge of that. So, I replaced the CRC calculation with
    one from another GNU package. The current CRC code should be OK with the Watcom compiler, but I haven't tried it yet. The (two or) three other files contained Borland copyrights if I remember correctly. So, if you had the appropriate Borland C(?) compiler, you could use the code..... I wouldn't be supprised if at least part of the Borland code has been released for public consumption, but I didn't have the time to track it down. From memory, these files dealt with OS issues, such as file name / directory issues, etc. Bo and I had to replace these files with other code or new code, and that new code was
    only tested in a Linux / Unix based environment. I'm sure there will be some changes needed in those (two or) three files to get the code to compile again for a Win32 / OS/2 / DOS target.

    And I did drop a few files from being checked in to the SqaFix subdirectory on sourceforge from Pete's original code because I didn't have a way to use or check the file's contents. I think the Watcom IDE's control files (i.e. make stuff) was part of that.

    Take care.....

    Bob Jones, 1:343/41



    --- Maximus/2 3.01
    * Origin: Top Hat 2 BBS (1:343/41)
  • From mark lewis@1:3634/12 to Bob Jones on Tuesday, November 18, 2003 14:10:56
    i'm currently seeking info on redirecting everything to
    the log file... seems that

    program 2>&1 > logfile

    try 'program 1>name.log 2>name.err'... Unfortunately, that
    will seperate stderr from stdout lines.....

    that's what i've ended up doing... i guess it does make it easier to see the errors ;-)

    Hmmmm......

    Thanks for the note. If you find a solution, let me know.

    hopefully you've read my other messages by now... i did finally get a complete compile... seems that there are some ^Ms in a couple of the sqafix files and my
    compiler doesn't like lines that continue on subsequent lines with a \... i've had to recombine all those onto one line as use tr to change the ^Ms to spaces...

    )\/(ark


    * Origin: (1:3634/12)
  • From Serguei Trouchelle@2:550/4077.101 to Bob Jones on Wednesday, November 19, 2003 00:18:11
    program 2>&1 > logfile

    try 'program 1>name.log 2>name.err'... Unfortunately, that will
    seperate stderr from stdout lines.....

    It can be used as 'program 1>>name.log 2>>name.log', and you'll have it in timing order, but you cannot recognize, which line is stderr.

    Serguei Trouchelle
    --- timEd/386 1.10.y2k+
    * Origin: Send more complaints! (2:550/4077.101)