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

Re: RFS: libconfig (requires transition)

Hi, Jonathan

My apologies if you did send an e-mail and I didn't read it. My inbox is severely overloaded as of lately...

Isn't it good style to notify the maintainer and/or coordinate with them before an NMU ?

On 31/01/12 00:08, Jonathan McCrohan wrote:
Julien Cristau<jcristau@debian.org>  wrote:

Please don't change the -dev package name.


All of the packages except one have versioned Build-depends
on libconfig8-dev. Surely this needs to be replaced with
libconfig-dev or at least libconfig9-dev?

No it doesn't?  You can rename the -dev package to
libconfig-dev if you want, but certainly don't *need* to, and
if you do it, then it would be way better from our point of
view to keep building libconfig8-dev as a transitional package
until the reverse deps are updated, and to do that separately
from the SONAME bump.
Yes, please.
Being bitten a couple times already after not checking buildability of r-deps... it is the library maintainer's responsibility, after all.
If its ok, I'll leave the package as is.

I can change it if it makes it easier for you so. Your paragraph above
made it sound as if it didn't matter which way it was done.

To clarify, what is the process for this transition? Will the
package be uploaded to experimental to allow me to report bug
reports and patches against dependant packages?

I'm not sure I understand what you're asking.
This is my first upload which requires a transition, and I am unsure
of what happens next.

Please read the library maintainer's guide first (or re-read if needed).
It does avoid many a headache...

  It seems common for packages to be uploaded to
experimental for a time prior to the actual transistion to allow other
maintainers update their packages accordingly. I was wondering will
this be the case with this transition?
Well, unfortunately for the world (some would say ;) ), not too many packages depend on libconfig; Nor are they very complex. Therefore, a full transition via experimental and involving the RM is not needed, AFAIK. Just notifying the depending maintainers should suffice (it would be different during freeze, of course)

Just drop me a line if I can be of any help.



Reply to: