Edubuntu 7.10 - A Released Debacle and a Practice in Failure - Response

[Jim Kronebusch]

> #6 Releases too often for Education
> 
> Yes, twice a year is perhaps too frequent for a typical education
> environment, especially seeing as typical school administrators are hard
> pressed for time and resources, and an unexpected spanner in the wheel
> can cause many headaches for the teaching environment.
> 
> At the 6.10 UDS, we discussed the concept of:
> 
> * an annual edubuntu release for education deployment
> 
> * a 6 month edubuntu gap release for administrators/decision
>   makers to evaluate and get an idea of where we are
>   heading
> 
> What complicates this is that the school calendar for the north/south
> hemispheres are out of sync due to the structuring of the school year
> around the summer long-break vacation.
> 
> Here in South Africa, we run with a calendar year ... the school year
> begins in January after the December summer break. Up north, it
> commences in September.
> 
> The Ubuntu release cycle is April & October
> 
> So which release is a good month for schools in both groups ?
> 
> * Comments welcome on this one. *

I have a lot of thoughts on all topics, but I need more time to respond to some many of
them.  But I have some quick thoughts on this one.  The release times are tied directly
to Ubuntu, so I assume that won't change.  Now for me, I don't want to upgrade during
the school year, but I would like the release to be available during the summer break
for a ton of testing before the school year starts and the production servers are
upgraded.  So for me the use of the April release is best, this allows 4 months of
testing before it goes into service.  That is enough time for the gotchas to be worked
out and most bugs to be fixed, making what is available by the time of actual service to
be very reliable.  But for those who have their break in December, this might be a long
wait for them to release.  However do they really want to go production on an OS that
was only released 2 months prior?  That might be enough time for bug fixes and testing,
but maybe not.  I really don't usually ever use an OS immediately on release, just
doesn't seem like a good idea as there are always problems no matter what distro
(Windows, Mac, Linux flavors).  I will be waiting for Hardy and will skip Gutsy
completely.  Not because I think Gutsy is bad, just that I wouldn't dare upgrade the
central server that runs all 110 student computers in the middle of the school year, and
I don't want any major changes that they would have to adjust to half way through the
year.  I'll wait for Hardy in April, start my testing, watch the bugs and hopefully
report my own, then when all is tested and safe, I'll upgrade the production server in
the end of August and test for a couple weeks before students show up.  

So long story short, I vote for a single release in April.  If anyone wants to they can
watch the beta releases and test new features and such.  But this makes it safe for all
end users to test...test...test...and bug fix so that all kinks are worked out before
implementation.  This also would relieve a lot of hassle that the limited developers
(full-time and volunteer) just to make a release.  The extra release cycle I'm sure
takes away from a lot of programming and bug fixes and feature additions.  A sensible
school (in my opinion) should do once a year upgrades during the yearly transition
(summer break for us).

My 2 cents,
Jim

-- 
This message has been scanned for viruses and
dangerous content by the Cotter Technology 
Department, and is believed to be clean.