One issue I see with the automatic update is that we're subject to
spam being automatically propagated to our main web site.
True, although looking at http://cwiki.apache.org/confluence/
signup.action , I notice that Confluence uses a word-image human-
detector, which I think is usually pretty robust a keeping spambots
away. Since we've been using the same system so far for the OpenJPA
incubator pages and haven't yet seen any spam, maybe we can wait and
see if we run into any problems, and if so, we can look into forcing
a user to run a script to do the update.
Also, the chain of Confulence Autoexport -> ~jefft rsync -> /www/
openjpa.apache.org rsync -> Apache webserver rsync can mean that
hours can pass between a change being made to the wiki pages and them
showing up on http://openjpa.apache.org/. While this is unfortunate
from the standpoint of not getting instant editing gratification, it
also means that there is some time to review the Confluence "page
changed" e-mails before they wind up going out on the exported front
pages.
On May 26, 2007, at 4:56 AM, Craig L Russell wrote:
So far so good.
Thanks for handling this.
One issue I see with the automatic update is that we're subject to
spam being automatically propagated to our main web site. If we
leave the mprudhom cron job off we are more in control, although it
would mean we would have to remember to rsync the site whenever one
of us made a change to the wiki that we wanted to propagate to the
main web site. But it would give us a chance to fix the errant wiki
pages before they were published to the world from our official site.
Craig
On May 25, 2007, at 11:45 PM, Marc Prud'hommeaux wrote:
OK, I've set this up. The system is basically the same as the
convoluted setup for exporting to the old incubator.apache.org site:
1. My personal crontab runs /www/openjpa.apache.org/RSYNCSITE.sh
(it would be nice to have an "openjpa" user so this isn't tied to
my personal user account).
2. RSYNCSITE.sh runs: rsync -a /home/jefft/public_html/confluence/
openjpa/* /www/openjpa.apache.org/ (note that /home/jefft/
public_html/confluence/ appears to contain an automatic sync of
all the confluence exported sites, so I didn't see any need to set
up an additional one).
3. Once the stuff is at /www/openjpa.apache.org/, some other
process (unknown to me) seems to periodically sync that directory
content to a place where it actually shows up at http://
openjpa.apache.org/ .
I've also set up redirects from all the old html pages from http://
incubator.apache.org/openjpa/ to go to http://openjpa.apache.org ,
so links should be redirected appropriately. E.g., if you do to:
http://incubator.apache.org/openjpa/downloads.html
you should be redirected to:
http://openjpa.apache.org/downloads.html
Please let me know if anyone runs into any problems with this.
On May 25, 2007, at 11:19 AM, Craig L Russell wrote:
Thank you thank you thank you.
Can you make sure you can edit the people.apache.org /www/
openjpa.apache.org directory (need to be in the openjpa unix
group) to facilitate your work later?
All the committers should be in the unix openjpa group already
but this is part of the migration so it might not be set up yet.
Craig
On May 25, 2007, at 11:12 AM, Marc Prud'hommeaux wrote:
I'll try to look into this sometime this evening (PST).
On May 25, 2007, at 11:07 AM, Joshua Slive wrote:
On 5/25/07, Craig L Russell <[EMAIL PROTECTED]> wrote:
Hi Joshua,
I would appreciate some help here. I have no time to troll
through
all other Apache web sites looking for models.
Is there a quick 1-pager on how to symlink the OpenJPA
confluence web
site to be our www web site? Or another project that does so
with a
public www/projectnamehere that I can model?
Sorry, but your confluence guys need to come up with the
procedures
yourself. I don't use confluence. I can only tell you what is
required
to have a reliable, maintainable apache infrastructure.
But you can probably use the instructions from
http://cwiki.apache.org/CWIKI/
just replacing the ~jefft location with the location under
/www/confluence-exports.
Joshua.
Craig Russell
Architect, Sun Java Enterprise System http://java.sun.com/
products/jdo
408 276-5638 mailto:[EMAIL PROTECTED]
P.S. A good JDO? O, Gasp!
Craig Russell
Architect, Sun Java Enterprise System http://java.sun.com/products/jdo
408 276-5638 mailto:[EMAIL PROTECTED]
P.S. A good JDO? O, Gasp!
