Patches item #1781994, was opened at 2007-08-26 12:19
Message generated for change (Comment added) made by
pmiller
You can respond by visiting:
https://sourcefor
ge.net/tracker/?func=detail&atid=559968&aid=1781994&
amp;group_id=80503
Please note that this message will contain a full copy of
the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: liblicense
Group: None
Status: Open
Resolution: None
Priority: 5
Private: No
Submitted By: Peter Miller (pmiller)
Assigned to: Scott Shawcroft (tannewt)
Summary: liblicense.0.4 - gcc -ansi
Initial Comment:
This change set increases the portability of the C code - to
compilers other than GCC - by avoiding post-1989 features of
the C language definition. Yes: 18 years later and they
STILL haven't caught up.
I haven't committed this to my repo yet, I'd like to know
what folks think.
Personally, I'd rather be coding in C++, where mixed
declarations and statements are OK. They make a lot of
sense. Problem is that most compilers except GCC are, well,
behind the times.
The no-c++ comments thing is important programmer
psychology. It provides the reader with an almost
subliminal clue that they are dealing with C, and not
anything else.
------------------------------------------------------------
----------
>Comment By: Peter Miller (pmiller)
Date: 2007-08-29 00:35
Message:
Logged In: YES
user_id=160
Originator: YES
Attached is a new improved patch. It gets rid of alloca,
and it fixes the
flat_file.c problem.
File Added: liblicense.0.4.c24.patch.txt
------------------------------------------------------------
----------
Comment By: Peter Miller (pmiller)
Date: 2007-08-29 00:32
Message:
Logged In: YES
user_id=160
Originator: YES
sf.net has it still assigned to you, scott.
please commit the patch, if you agree that I can fix the
alloca in another
change set (soon!)
------------------------------------------------------------
----------
Comment By: Jason Kivlighn (jakin44)
Date: 2007-08-28 23:40
Message:
Logged In: YES
user_id=1625120
Originator: NO
I see a bug in flat_file.c. If fgets() doesn't read
anything, tmp should
explicitly be set to NULL (or is it the empty string
(Scott?)). Otherwise,
tmp remains uninitialized and we return jibberish. The file
will be empty
when the default license has been unset.
------------------------------------------------------------
----------
Comment By: Peter Miller (pmiller)
Date: 2007-08-28 22:15
Message:
Logged In: YES
user_id=160
Originator: YES
sf.net has it still assigned to you, scott.
please commit the patch, if you agree that I can fix the
alloca in another
change set (soon!)
------------------------------------------------------------
----------
Comment By: Jason Kivlighn (jakin44)
Date: 2007-08-28 19:42
Message:
Logged In: YES
user_id=1625120
Originator: NO
Getting rid of c++ comments is a clear OK. Getting rid of
mixed
declarations and statements couldn't hurt (although I agree,
it's lame and
annoying that there are still compilers out there that
haven't gotten their
act together).
I'm unsure of using alloca(). I'm not all that familiar
with it, but it
sounds like it's discouraged and has many buggy
implementations. I'd
prefer to use malloc() instead. ll_new_license_chooser is
typically only
called once during the entire session, so we're not looking
at a
significant performance hit allocating on the heap verses
the stack.
------------------------------------------------------------
----------
Comment By: Jon Phillips (kidproto)
Date: 2007-08-28 19:28
Message:
Logged In: YES
user_id=914868
Originator: NO
Go for it...pmiller can commit to!
------------------------------------------------------------
----------
Comment By: Scott Shawcroft (tannewt)
Date: 2007-08-28 19:19
Message:
Logged In: YES
user_id=1634298
Originator: NO
Looks good. Compiler portability is not something I've
dealt with so I'll
take your word for it. Shall I commit it?
------------------------------------------------------------
----------
You can respond by visiting:
https://sourcefor
ge.net/tracker/?func=detail&atid=559968&aid=1781994&
amp;group_id=80503
_______________________________________________
cc-devel mailing list
cc-devel lists.ibiblio.org
ht
tp://lists.ibiblio.org/mailman/listinfo/cc-devel
|
