Boost logo

Boost Users :

From: Wu Yinghui, Freddie (yhwu_at_[hidden])
Date: 2006-08-16 11:24:43


> Hi Freddie.
>
>> Thanks for your quick reply. I knew this trick. But what I want is
>> probably some more elegant ways of doing this. (It seems quite a bit of
>> a trouble since those pragmas will generate lots of warnings when I
>> build using gcc. And if I have to surround each of those pragma with
>> platform detection macros, it will end up quite a few lines for each of
>> my hundred header/source files.)
>>
>> Any easier-to-adopt ideas? :)
>
> Hmmm... I do not think I explained myself clearly enough the last
> time. :-)
>
> If you look at the header files I sent you they already have platform
> detection guards in them. And the whole point of those headers is to
> have them and only them contain the platform detection and the
> necessary pragmas. After that you just use those BoostWrappers
> headers instead of the original ones from boost. That means that you
> add NO EXTRA LINES to any of your own hundred header/source
> files. :-)
>
> I also told you that we at first had the same fears as you, but as it
> turns
> out actually very few boost libraries cause warnings and so there are
> actually very few of those BoostWrappers headers. Look the archive
> I attached in the last message - there are less maybe 10-20 files in there
> and their contents is very stable. Adding new ones is rare and when it
> does happen it happens natually - compiling your other code tells you
> directly which warnings to disable and in which header files.
>
> Hope this clears things up... :-)
>
> Best regards,
> Jurko Gospodnetic

Sure it does. I guess this is one of the best approach I can see for
now. But I had problem getting the attachment you mentioned from your
previous post. (I didn't see any attachment.)

Is it possible that you send the header files again? Sorry for the trouble.

Cheers,

Freddie





Boost-users list run by williamkempf at hotmail.com, kalb at libertysoft.com, bjorn.karlsson at readsoft.com, gregod at cs.rpi.edu, wekempf at cox.net