2014/10/06 10:44 "Chris Bannister" <cbannister@slingshot.co.nz>:
>
> On Sun, Oct 05, 2014 at 10:34:57PM +0900, Joel Rees wrote:
> >
> > > If you think you have found a problem, then as a user of testing you
> > > should report it. You may have found an obscure bug which should be
> > > fixed *before* Jessie becomes the new stable.
> > >
> > > As users of testing, we should take a more active role with regards to
> > > bugs, after all one of the reasons to run testing is to help make the
> > > stable release as bug free as possible.
> >
> > Chris, I'm not understanding your reasoning here.
>
> How do you think bugs get fixed if they're not reported?
> What else can you read into the above paragraphs?
>
> > Are you thinking he should be considering the possibility he's been
> > hit by a MITM or something? Do you want him to try to roll back the
> > testing repository or a mirror so he can do a full binary compare?
>
> *WHAT*? I was thinking nothing of the sort. I'm not the person
> responsible for producing the netinst image, but if I was that would be
> the last thing I'd be thinking --- more likely just a silly mistake BUT
> how would I get to hear about it if whomever found the mistake didn't
> let me know, *Hrrrmmm*?
Okay, looking at the OP's descriptions of the problems, should we ask the OP to go ask the release team whether an old netinstall image for testing should, in general, be able to recover from having the kernel on the CD out of sync with the current repository? And would they like him to image the old netinstall CD and upload it somewhere for analysis? Or would they like him to try to replay the install, maybe from a mirror where they can watch?
I don't know how slow the friend's connection is, but unless there were some specific bug I were chasing, I'd hesitate to ask for a replay if it's going to take two of them a good part of a day, and erasing a working system. Especially considering that my record on installs leaves me with about one in five install CDs botched for various reasons.
Joel Rees