Re: elinks to be REMOVED from squeeze?

[Freeman <evenso@worldwidehtml.com>]

On Fri, Jan 22, 2010 at 03:24:43PM -0800, Todd A. Jacobs wrote:
> On Thu, Jan 21, 2010 at 01:21:24PM +0000, Jeffrey Cao wrote:
> 
> > sure elinks has not been updated for a long time. So, what is changed
> > to make elinks depends on libtre5?
> 
> I don't know about this particular case, but in general, one of two
> things probably happened:
> 
>     - The upstream author changed a library dependency. This happens a
>       lot, and is only a bug if the package doesn't track the dependency
>       correctly.
>     - The maintainer made an error in a configuration file when building
>       the deb. If you've ever tried to build your own deb files, you'd
>       understand how easy it is for this to happen.
> 
> Either way, you will typically get zero traction with "Why did someone
> break my favorite package?" The correct thing to do is report a bug, if
> no one has already done so, and then either wait for a new package to
> migrate to your chosen repository (such as testing), install from
> somewhere else (e.g. sid or experimental), or roll your own from source.
> 
> If you're tracking packages from testing or unstable, then expect this
> sort of thing to happen from time to time. The only cure is bug reports
> and the occasional invocation of "aptitude dist-upgrade" to ensure that
> you're upgrading things that have problematic library dependencies.
> 
> -- 
> "Oh, look: rocks!"
> 	-- Doctor Who, "Destiny of the Daleks"
> 
> 

Is there really even a point to filing a bug report against something when
the fix is upstream?

Testing is my opportunity as a peon to contribute to the Debian effort by
filing reports when I can correctly assess a bug's existence.  However,
caution toward filing reports necessarily also seems important.

I usually wait a month before thinking much of it when something breaks. 
Most apps start working correctly following an update, sooner than later. 

As I understand, buggy packages occasionally are pushed downstream to avoid
dependency logjams in release development.

I booted into the 2.6.26 kernel to run Google-Earth for 6 months. When it
quit working in that kernel too, I waited another 4 months before
discovering my card was dropped from fglrx. :)

-- 
Kind Regards,
Freeman