[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: The future (dependencies on libc{5,6,6.1} &c)



On Monday, 29 Sep, Michael Alan Dorman wrote:
> 
> The 6.1 isn't just an arbitrary number---it's the soname of the
> library.  We should keep the package name reflecting reality.
> 

That's true.  However, everywhere in Debian (and in Linux world) libc6
means glibc2.  It's probably more confusion to move away from that than
to allow a package name not to reflect the soname.  Examples:
  - libc6-doc is a binary-all package; shall we maintain a binary-alpha
    copy of it under the name libc6.1-doc or say somewhere in FAQ that
    libc6-doc is actually the documentation for libc6.1 package?
  - the source package for libc6.1 is called libc6_<version>; although
    it actually doesn't have to have anything in common with the binary
    package name, this situation is confusing.  

Although it is very good practice to put the soname of a library into
that library's package name, it's not a policy requirement.  We may
deviate from this when it's justified.

> I realize that it's confusing, since it is 6.1 only on the alpha, and
> I'll agree that ought not to have been done (it was done to accomodate
> RedHat), but we shouldn't mangle it.

I think we have a perfect chance to undo what ought not to have been done.
I believe there are no Debian packages left which are linked against
glibc-1.xx.  So at the present moment we can change libc6.1 to libc6
without any problems, we don't even have to maintain compatibility with
previous releases.  Later it will be much harder to accomplish.

At the moment there is no libc6 package in Debian/Alpha.  I doubt there
ever will be.  Xquake requires glibc-1.93 - OK, they can be packaged
together in the same package, glibc-1.93 being "xquake-private" rather than
system-wide library.  Why can't we rename libc6.1 to libc6?

	Nikita


--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
debian-alpha-request@lists.debian.org . 
Trouble?  e-mail to templin@bucknell.edu .


Reply to: