[BlueOnyx:00290] Re: Questions about BlueOnyx...

[Michael Stauber]

Hi Darrell,

> > As on BlueQuartz the server name cannot be identical to that of a site.
> > This has been the case on BlueQuartz for like 2-3 years now - if I recall
> > correctly. The issue here is that Sendmail otherwise has issues resolving
> > local users.
>
> What do you recommend someone doing if they have that arrangement? 
> Changing the name of the machine and then performing the export?  Or
> changing the name of the first site?

During export there is nothing special to do there. Just run CMUexport. On 
import on BlueOnyx just make sure that the name of the server itself is 
different than that of any site.

> When you CMUexport, where does the files go? 

Typically /home/cmu, unless you specify a different directory with the -d 
switch.

# cmuExport -h
Reading config file: /etc/cmu/cmuConfig.xml
usage:   cmuExport [OPTIONS]
         -a export admin's files
         -c export configuration only
         -d build directory, this is where export will place all exported 
files, the default is /home/cmu/FQDN
         -i export all virtual sites with this IP address (RaQ only)
         -n export these virtual sites, ie "www.foo.com,www.bar.com" (RaQ 
only)
         -p do not export user passwords
         -v verbose, print all messages to stdout
         -h help, this help text

So if you run ...

cmuExport -d /home/export 

... it will store the CMU files in /home/export instead.
 
> Would it be easier to RaqBackup to the new machine and then restore that?

Doesn't really matter. RaQbackup.sh may save you a step or three, but in the 
end they both do the same steps.

> Still doesn't answer the question of what is the status of BO's Majordomo.
> Was it changed, left as is, will it import the old to the new?  I assume
> CMU will move the Majordomo lists, settings, etc., but has the product
> changed at all since the original BQ implementation?

Majordomo hasn't changed. It's the very same old clunker as on BlueQuartz - 
without any differences.

> > ChiliSoft ASP
>
> I wonder if anyone is really using it.  

Yeah, am wondering about that myself, too. I'd be hard pressed to think of 
anyone who's really (still) using it. JSP / Tomcat is a different matter, so 
we went great lengths to make sure that this works out of the box already.

> Interesting.  And I suppose FrontPage usage has discontinued or at least
> waned as well?

There are probably some die hard fans who still might use it, because they've 
always used it and didn't look further. But it really might be high time to 
nudge them onwards and get them moved to something more modern. :p

> Michael, what is the status of the comment that an upgrade procedure is
> forthcoming?  While an upgrade is tricky and convoluted, I think people
> want to know there is some formalized process before they begin.

Am still working on that and I don't really want to set a fixed date on when 
it'll be finished. If I say "a week" it could as well turn into two. But 
around that figure. But as said before, this upgrade procedure will be tricky 
and it cannot be guaranteed that it won't end in a total trainwreck in some 
isolated cases. Using CMU instead will be much, much safer and cleaner.

> Also, Michael, can you comment on what the impact of Brian's departure will
> have on the BO project?  I know a lot of people were/are excited to see
> something happen with the old BQ project, but with Brian's departure,
> wonder if the time is right for a fork off BQ.  In other words, wouldn't
> this be a good time to partner back up with Hisao and label this project BQ
> for CentOS 5?

Let me put it this way: In almost two years that Brian and I were part of 
BlueQuartz there was literally no visible progress whatsoever. There was lots 
of "noise" on the developers list when we tried to nudge things along, but 
very little action. This wasn't due to malice or underlying conflicts or 
resentiments. For the most part it may simply have been a mentality and also a 
language problem that we could never really overcome. None of that has 
changed. So re-merging BlueOnyx with BlueQuartz would solve none of those 
issues or improve anything.

As for the impact that Brian's break has on project BlueOnyx: He sure will be 
sadly missed for however long his departure takes. However, he's a maker, a 
coder and very ressourceful and eventually he may miss strangling the code and 
may come back with new ideas, concepts and contributions. Who knows? Now we 
set the project up in a way that both Brian and I had equal responsibilities 
and abilities when it came to managing the resources and scheduling releases. 
That way the continuity of the project would even be ensured if one of us got 
run over by the proverbial bus. 

Of course that "chain" of responsibilities (and abilities) is currently 
interrupted, but we're working on getting that redundancy back. Team BlueOnyx 
currently consists of four active developers and we're also willing to take on 
more. Of course it takes time for people to get acquainted with the code or 
some intricacies of the underlying system that they typically had not to work 
on. But we're getting there.

So no worries: BlueOnyx isn't a a bushfire and we'll do our best to expand on 
the foundation that has been laid out so far.

If the web servers traffic statistics for January are any indication of a 
sucessful release, then I must say that it - already - has exceeded our 
wildest expectations by several magnitudes. Uh ... Chris? Still think it was a 
good idea to sponsor the bandwith? :o)

-- 
With best regards

Michael Stauber