lkml on kernel headers and glibc (et al)

Donald Smith dss-lfs at cfl.rr.com
Sat May 31 18:46:31 PDT 2003


Seth W.Klein wrote:
> Donald Smith <dss-lfs at cfl.rr.com> wrote:
> 
>>And I still don't understand the kernel guys' stance on this. They 
>>provide the interface so why not the header files? In my opinion, 
>>they're just lazy.
> 
> 
> I think it's not like that.
> 
> In that thread, Linus expressed no opinion on future direction (only
> saying that the existing headers were not clean for external use) and
> Alan said that contributions were welcome. Basically, one gets the
> impression that they're not fixing it because they have more broken
> things to fix. If it's not obvious, these guys have enough work that
> they must prioritize heavily.
> 
> Of those with the time to make a difference on something not totally
> broken, which this is, the ones with the technical understanding were
> in favor of the kernel providing headers for its public API.
> 
> Of the few who clearly favored a separate package, the foremost was
> the maintainer of the RedHat package. He has a job at stake.
> 
> Unfortunately, no one claimed to have the time to produce a good
> solution right now. I think the direction this will go is clear;
> the speed, however, is not.
> 
> cheers,
> Seth W. Klein

Yes, but instead of leaving the interim solution that has worked, they 
allow developers to make changes that break the glibc compile without 
providing the correct alternative. The proper solution should have been 
provided first and THEN break the old method.

Of course, I am of the old school that knows if there are two separate 
copies of something that needs to be maintained, one will be forgotten 
at some point. Then again it *is* an OS interface...

-- 
Unsubscribe: send email to listar at linuxfromscratch.org
and put 'unsubscribe lfs-dev' in the subject header of the message



More information about the lfs-dev mailing list