<<< Date Index >>>     <<< Thread Index >>>

Re: feature freeze?



On Mon, Aug 14, 2006 at 21:38:29 +0200, Thomas Roessler wrote:
> So, what else -- except for a bit more public exposure of this or
> the next version -- do we need before 1.6 is ready?

Just for sanity's sake: Try to get rid of warnings?

Linux 2.6.17, glibc 2.4, gcc 4.1.1 20060525 (Red Hat 4.1.1-1)
./configure --enable-imap
make

[Sorry for long lines]
message.c: In function ‘imap_read_headers’:
message.c:303: warning: dereferencing type-punned pointer will break 
strict-aliasing rules
message.c: In function ‘msg_cache_open’:
message.c:872: warning: value computed is not used
message.c:873: warning: value computed is not used
copy.c: In function ‘mutt_copy_header’:
copy.c:473: warning: ISO C90 does not support the ‘ll’ printf length modifier
copy.c: In function ‘_mutt_copy_message’:
copy.c:582: warning: ISO C90 does not support the ‘ll’ printf length modifier
copy.c: In function ‘copy_delete_attach’:
copy.c:825: warning: ISO C90 does not support the ‘ll’ printf length modifier
sendlib.c: In function ‘mutt_write_fcc’:
sendlib.c:2431: warning: ISO C90 does not support the ‘ll’ printf length 
modifier
muttlib.c: In function ‘mutt_buffer_add’:
muttlib.c:1488: warning: dereferencing type-punned pointer will break 
strict-aliasing rules
utf8.c:3: warning: ISO C forbids an empty source file
wcwidth.c:14: warning: ISO C forbids an empty source file
muttlib.o: In function `mutt_adv_mktemp':
/usr/local/new/tools/networking/mail/mutt/mutt-1.5.13/muttlib.c:78: warning: 
the use of `mktemp' is dangerous, better use `mkstemp'

* The 'll' issue is not important, and I have seen some build
  environments hack around this.

* The 'ISO C forbids an empty source file' issue is very peculiar.
  config.h is _not_ empty, AFAICT.

* The 'mktemp' vs. 'mkstemp' is a not so recent glibc warning. Could be
  worked around by checking for mkstemp() and using it if available.

I know if you fix it for one compiler (version), it might break for
another. Just wondering. :)

I might try to build with icc (Intel C++ Compiler) and see what that
coughs up. ;-))

Just a low prio thought,
Moritz