On Fri, Sep 24, 2010 at 11:43:13AM +0100, Tim Abell wrote:
> 

Just few comments to your dump :).

> I've been thinking about how to manage bugs that occur in one or more 
> versions of shr, as I'm thinking about how shr-t should be run from here on.
> 
> Given that we currently just have shr-u and shr-t, if a bug exists in 
> both, how should it be categorized in trac? The version field only 
> supports marking a single version. I see we have "fixed in shr-u" in the 

"fixed in shr-u" was introduced to trace such fixes which should be
interesting to shr-t maintainer, who should cherry-pick them from shr-u
if they are easy (doesn't need many depending changes) or apply them
with next shr-u -> shr-t merge and then mark this bug as "fixed" which
would mean fixed in both.

so in other words "fixed in shr-u" is something like fixed in shr-u,
waiting to be fixed in shr-t too.

> resolutions, which I guess implies not in shr-t, but some bugs might 
> never show in shr-t so if a bug is marked as version shr-u it will be 
> hard to tell (for the shr-t maintainer) if the issue affects shr-t or 
> not. If it is marked as shr-t it might not be clear if it still affects 
> shr-u. I don't really have an answer but it's something that's been 
> bothering me. It might be worth having a look at how 
> https://bugs.launchpad.net/ manages bugs against the many currently 
> supported versions of ubuntu. Would it be possible to configure trac to 
> have multiple "affects x version of shr" and "fixed in x version of shr" 
> fields or similar, so that we can track exactly which versions of shr 
> are currently affected and which still require attention? (This is of 
> particular interest to me as a potential shr-t maintainer as I would 
> need to know which bugs affect shr-t specifically and which have fixes 
> available in shr-u, as well as what the impact is etc).
> 
> Incidentally I'd also like to see shr move away from pulling new 
> features as a drip feed directly into shr-t and instead move to the 
> debian/ubuntu model where at regular(ish) intervals a snapshot of shr-u 
> is made into the beta of the next shr-t, and once that shr-t becomes 
> semi-stable and officially "released" the old shr-t just moves down one 

yup, that's why there is newer shr-t I've built long ago
http://build.shr-project.org/tests/shr-testing/
but never merged to normal feed in
http://build.shr-project.org/shr-testing/
because request for testing didn't show clearly if it's worth it and if
somebody is interested in this upgrade. And because I'm not using shr-t
myself at all, I'm not going to spend much time testing it..

but for next shr-u -> shr-t merge I would like to see something like
shr-u was just before switch to 2.6.32, because I see lots of complains
about uSD instability etc after we switched.. so it could be good
next-shr-t base.

> but continues to be maintained for a while. I suggest we attempt to 
> mimic ubuntu's schedule https://wiki.ubuntu.com/Releases as I think 
> fixed dates work well, and the ubuntu people are trying to persuade all 
> the distros to rally around the same dates (so that the kernel, kde, 
> firefox etc all produce stable/unstable at the right moments for 
> integration into distro releases, which I think is very sensible). If 
> no-one objects I'll document this plan in the page I added to the wiki 
> http://www.shr-project.org/trac/wiki/ShrMaintainerHowTo and attempt to 
> run the shr-t releases myself (though I am interested in the way debian 
> appoints maintainer responsibility etc even if I haven't got my head 
> round it yet!).

Great that you're still interested in shr-t maintaining.. we really
need shr-t maintainer as those images are really going to be too old :).

-- 
Martin 'JaMa' Jansa     jabber: [email protected]
_______________________________________________
Shr-User mailing list
[email protected]
http://lists.shr-project.org/mailman/listinfo/shr-user

Reply via email to