--- On Mon, 4/12/10, Peter Robinson pbrobinson@gmail.com wrote:
From: Peter Robinson pbrobinson@gmail.com Subject: Re: New XO-1.5 10.2.0 build 119 To: "Yioryos Asprobounitis" mavrothal@yahoo.com Cc: "Bernie Innocenti" bernie@codewiz.org, "Devel" devel@lists.laptop.org, "Fedora OLPC" fedora-olpc-list@redhat.com Date: Monday, April 12, 2010, 3:56 PM On Mon, Apr 12, 2010 at 8:40 PM, Yioryos Asprobounitis mavrothal@yahoo.com wrote:
--- On Mon, 4/12/10, Bernie Innocenti bernie@codewiz.org
wrote:
From: Bernie Innocenti bernie@codewiz.org Subject: Re: New XO-1.5 10.2.0 build 119 To: "Peter Robinson" pbrobinson@gmail.com Cc: "Devel" devel@lists.laptop.org,
"Fedora OLPC" fedora-olpc-list@redhat.com
Date: Monday, April 12, 2010, 9:15 AM On Sat, 2010-04-10 at 17:57 +0100, Peter Robinson wrote:
Finally, I guess you have thought of
it, but
by the
time 10.2 will be out F11
repositories will
be down
and thus the builds totally frozen
software-wise.
I think it would have been better to
rebase on
F12 6 months ago.
Now it's way too close to the release
date :-(
I recommended F-12 which was in beta when
this process
started but was
ignored. I noticed the other day that dsd has
created
a F-12 branch in
git but I think we should be aiming straight
for F-13
now. It'll be
out in a little over a month, is quite stable
already
and have will be
supported for another 14 months.
+1
F13, btw, seems like a very solid release to me.
I do not know why re-basing on F13 will reach
deployment status much faster than F11 and will not be at the same point a year from now.
The XO is a _production_ machine. Makes no sense to
run development/short-lived OS. Maybe the RHEL/CentOS idea should not be dismissed, if feasible.
In the longer term its likely that it will be. The libraries in CentOS 5 are too old. The problem is that RHEL-6 isn't out yet, nor is there any announcements when it will be, then from there the CentOS team need to review, engineer, QA and release CentOS 6. Then you need to build/test/QA sugar etc for it.
It will then be OK for a while until Sugar needs a newer version of a library that is static on CentOS due to the support requirements. And then it reverts to Fedora. And before you mention Ubuntu LTS, the previous LTS has the same version issues, the current one will still have the same problems going forward.
Maybe so, but the fact is that "going forward" actually means that the XOs are still running F9, and hopping to get an EOL F11 in the next few months. So at the end of the day this "needed library" is not actually used by the vast majority of the actual users. I repeat that the XO, and education in general, is/needs _production_ machines. Given the resources and the development rate, a 3-4 years OS support should be the minimum I think, even if the feature requiring "that library" is not implemented. Just my 2c.
Peter