RE: [Axiom-developer] backups

[Page, Bill]

On Wednesday, April 12, 2006 7:36 Bob McElrath wrote:
> 
> Is the backup at http://axiom.risc.uni-linz.ac.at/ and the
> main mathaction wiki synchronized?  i.e. if someone makes
> a change on the axiom.risc.uni-linz.ac.at site, does it
> get propagated to the main site?

Yes, I am using ZSyncer on both sites. Right now the process
is still manual - mostly due to a problem involving
"last_modified_date" on ZWiki pages separated by 7 timezones.
I am planning to synchronize the two sites at least once per
week. In the future I think this can be largely automated
(except for edit conflicts). Eventually we might get software
that makes this easier.

> If not, it is probably best not to advertise the backup
> site, because it will just cause headaches if people use
> it.  (In fact, it is probably best to keep it offline to
> prevent google discovering it and leading people to use
> it.)
>

An offline backup has it's merits and I think we should
continue to do that on a regular basis. Note that Tim
Daly has recently extended the service contract with
RoseHosting to include a nightly backup of the contents
of axiom-developer.

I have thought about how best to use the new mirror site
and I strongly believe that it should be a full and
completely accessible mirror of the main Axiom site.
Keeping a live mirror provides the incentive to keep the
backup site up to date. It is ready to go and immediately
in case of any problems with the other site. In the future
if they are allowed to diverge in content I also think
that is not necessarily such a bad thing - provided that
both sites remain open and accessible.

Perhaps some users will feel a greater affinity for a web
server that is physically closer to them? and as a result
they might devote more time to it than otherwise would be
the case. Also inevitably there may be a diversity of
interests in how Axiom is used and how it can be further
developed. Since this is an open source project, I think
we should not be afraid of such a possibility, although
given our current very limited resources it might be
prudent to try to concentrate our efforts on less ambitious
goals.

Regards,
Bill Page.