[Shr-User] Problem with e libs in shr testing

[Robin Paulson]

2009/5/10 Carsten Haitzler <raster at rasterman.com>:
> all you are doing is to revert the work upstream is doing to try and improve
> stability - but if that is what you want to do - feel free. we will also feel
> free to ignore any bug reports or issues as we no long know what release your
> built against.
>
> we explicitly released a shanpshot with that version name/tage so we know what
> you build against. don't blame upstream. we announced:
>
> http://www.enlightenment.org/p.php?p=news/show&l=en&news_id=3
>
> on enlightenment.org's front page. if you choose to update to a sevnrev
> somewhere in between you will get arbitrary soames as if you are using svn we
> assume you are a developer and are regularly updating and rebuilding. if you
> use a "Blessed" svnrev (or release tarballs) then you will be doing well.

the problem wasn't with the enlightenment releases per se, it was with
shr blithely packaging and releasing the new version of e in
shr-testing, without any warning of breakages or an opportunity for
devs to rebuild packages against the new scheme

several users pointed out it had broken a load of e apps., and we were
told 'it's upstream's fault, they released e with a new naming scheme,
don't blame us. sort it out for yourselves'. faced with a system that
didn't work, or doing what we did - i.e. a slew of hacky symlinks - we
went with the latter

> i will do a similar rename of sonames and release of elementary later this
> month along with a new snap of core efl as before (this time it will be 02
> instead of 01.)