Hi List!
In my experience having a compiler warning free code is a good way to
prevent very subtle bugs and would also be a good addition to the
programs security - and BTW more pleasant to work with ;-)
You will often hear the following excuse on this topic: "but you cannot
write code which won't produce warnings on all those compilers out
there". While there are cases where this is true (which has to be
handled individually), it's much more often a sign of lazy / ignorant /
unskilled developers IMHO.
Unfortunately, we're currently having a clear trend to get more and more
compiler warnings instead of less. Years ago, I've taken the work to
remove warnings of the Win32 build to about 10, today we're having
hundreds again :-(
While I would be willing to remove code warnings for Win32, I first
thought about how to prevent new warnings to rush in - otherwise any
work on this topic would be annoying by seeing new warnings rushing in
shortly.
In my experience, preventing yourself to add new warnings to the WS is
currently difficult because of the multi platform support. Even if you
provide code that doesn't produce any warnings on "your" working
platform, you won't simply notice any new warnings on other platforms
(unless you take a look at the buildbot logs) - and compiling on
multiple platforms all by yourself is not on option for most of the
developers IMO. In the end this is what the buildbot is for.
So here comes the buildbot into the scene. If we would use a compiler
option like "stop on warnings" (or "treat warnings as errors" or alike),
it would become at least much more obvious if new warnings were added -
the buildbot will get "red". This will also make the time when a warning
is noticed much nearer to the time the code was added/changed -
currently fixing a warning once added is often done much later than it
was introduced (making the fix unnecessarily difficult).
Simply add such a compiler option to the current code state (with it's
hundreds of current warnings) would make the buildbot "red" on all
platforms for a very long time - which is probably not a good idea.
An incremental way to introduce this could be:
a) cleanup the warnings for a specific submodule (e.g. wiretap) and
platform (e.g. Win32)
b) set the compiler option for this submodule only to get a "barrier"
for new warnings to rush in
c) continue with a) for next submodule :-)
As usual, this is my "Win32 point of view". I'm pretty sure the above is
possible to do for the Win32 platform. I'm not sure if it's possible
with the automake foo for the different unix/linux platform builds ...
In effect, the above is to introduce the new coding policy "write
warning free code" and a way to "enforce" this.
So what's the opinion about this way to improve the Wireshark code base?
Are we willing to produce only warning free code and fixing warnings
that appear on the buildbot?
While I would take a look on the Win32 warnings, are the unix/linux
developers willing to spend some time to remove warnings that don't
appear on Win32 (or would this be a "Win32 only" show)?
Regards, ULFL
