It will… 
···
On Wednesday, July 23, 2003, at 02:27 PM, Simon Strandgaard wrote:
Curious… does the webserver have mod_ruby installed ?
repository…two distinct things…but we will want to work closely
with RAA (perhaps updating RAA automagically when you do a release on
RubyForge for example).If you want people to use rubyforge, then I think you should really
offer that.
I agree. I will contact the RAA team and see what we can do.
Me thinks there is perl in there…and python…ug…but hey, its a
tool!How hard was gforge to set up?
Its a PITA…but Tom has now done it several times, and so has learned
the challenges involved. Its better than trying to use Savannah’s
source though. There are just a lot of dependencies.
Will you be offering eruby for use on the projects websites?
We should be able to. I will let folks know when this is available.
Best,
-rich
···
On Tuesday, July 22, 2003, at 12:29 AM, maillist@bestworldweb.homelinux.com wrote:
I was just kidding. Just add yourself as RubyForge user and then
request the project…we will get back to you when we wake up (its
1:30am here…ug).
-rich
···
On Tuesday, July 22, 2003, at 01:16 AM, Gavin Sinclair wrote:
On Tuesday, July 22, 2003, at 12:41 AM, Gavin Sinclair wrote:
You are going to host my project, aren’t you?
Is it written in Ruby
Hmmmm. Actually… no.
I want a project to cover Vim configuration files for Ruby editing. In
particular, the following files:indent/ruby.vim
ftplugin/ruby.vim
syntax/ruby.vim [if the current maintainer agrees]This is obviously a small project, written in a language other than
Ruby,
but I see a necessity to have it available by CVS, so bleeding edge
features can be tested and contributed before being released.Gavin
OK, OK…major oversight
Its there now (albeit at the bottom). Of course, this is simply
because Ruby is the #.last language you will ever need 
-rich
···
On Tuesday, July 22, 2003, at 07:40 AM, Jim Weirich wrote:
I’ve noticed that the snippet library doesn’t contain an entry for
Ruby! (although Perl and Python are represented).
First code snippet!
First feature request!
First bug report!
I win!!!
Dan
Note: This post is not for the humor impaired.
Simon Strandgaard wrote:
Curious… does the webserver have mod_ruby installed ?
then Richard Kilmer wrote:
It will…
Just great!!
This is a good appetizer for other projects doing a transition to
Rubyforge.org.
BTW: The last 2 hours I been browsing Rubyforge, comparing it against
sourceforge. I have noticed many nice features, that I sourceforge is
lacking. Especialy the easy file-release interface is nice.
···
–
Simon Strandgaard
But having subversion corrupt its db all the time was too much even for us.
Yes this is unacceptable and that this happens after a few years of
development is not a good sign. Until now i never had the time to go
through this extremely unpleasent installation - i don't now why a lot
of open source projects don't take more care about this point.
Does anyone know if there is a small PHP/MySQL based SCM tool
(commerical or free) . For small projects with only a handfull of users
i don't like to install a CVS server, throwing a few PHP pages and run
a database script on a cheap webhoster would be much nicer.
Did you fly any of these database problems by the subversion
developers, to see if they had any ideas as to the causes? I
haven’t used subversion at all, but one of my friends works a
lot with it. I’m certainly hoping to see problems like this
addressed.
When I asked my friend about this, he said there were several
cases of problems when some user runs ‘svnadmin recover’ on
a repository that people were still accessing. Is it possible
that you would have run into that? (I don’t even know what
the “recover” would be used for…)
···
At 4:08 AM +0900 7/24/03, Brett H. Williams wrote:
I’ve been following subversion development for two years, and
we finally took it out for a more complete test drive a couple
of months ago, using it for some small projects. …… Also, with a single tiny repository and 3 people using it
lightly, we got several database corruptions in a single week
which needed to be repaired. One of these wasn’t even
repairable with the recover command.
But having subversion corrupt its db all the time was too much
even for us.Do others have differing experiences with it? I’d love to hear
we’ve just done something really dumb because we could really
use subversion here…
–
Garance Alistair Drosehn = gad@gilead.netel.rpi.edu
Senior Systems Programmer or gad@freebsd.org
Rensselaer Polytechnic Institute or drosih@rpi.edu
Brett H. Williams wrote:
I’ve been following subversion development for two years, and we finally
took it out for a more complete test drive a couple of months ago, using
it
At my place ( informatic dept @ univ ) we have been using Subversion for
about 9month now for several internal development projects as well as
giving students the possibility to create their own repositories. Not once
did we experience database corruption.
The only major downside (other than that compiling is a PITA) we experienced
is, that the server and client versions need to be somewhat in synchro,
otherwise their communication breaks in unexpected ways (like XML errors or
plain “permissin denied” messages).
First code snippet!
and a very nice snippet indeed!
First feature request!
First bug report!
I win!!!
Excellent…you have won the ability to freely host code at
RubyForge!..Yeah!!!
···
On Tuesday, July 22, 2003, at 10:09 AM, Daniel Berger wrote:
Thinking of RubyForge + nice to have features:
The berlios project has one ‘wiki’ per project, eg:
http://developer.berlios.de/projects/ncurses-ruby
Berlios also has a ‘Screenshots’ tab.
Maybe its bloat, but both would be nice to have.
BTW: 3 different CVS-browsers … nice
···
On Wed, 23 Jul 2003 23:00:48 +0200, Simon Strandgaard wrote:
Simon Strandgaard wrote:
Curious… does the webserver have mod_ruby installed ?
then Richard Kilmer wrote:
It will…
Just great!!
This is a good appetizer for other projects doing a transition to
Rubyforge.org.BTW: The last 2 hours I been browsing Rubyforge, comparing it against
sourceforge. I have noticed many nice features, that I sourceforge is
lacking. Especialy the easy file-release interface is nice.
–
Simon Strandgaard
BTW: The last 2 hours I been browsing Rubyforge, comparing it against
sourceforge. I have noticed many nice features, that I
sourceforge is
lacking. Especialy the easy file-release interface is nice.
Glad to hear it! Thanks much for your patience as we continue to shake
out the inevitable initial issues…
Yours,
Tom Copeland
Did you fly any of these database problems by the subversion
developers, to see if they had any ideas as to the causes? I
haven’t used subversion at all, but one of my friends works a
lot with it. I’m certainly hoping to see problems like this
addressed.
We didn’t, as the problems are documented there as known.
When I asked my friend about this, he said there were several
cases of problems when some user runs ‘svnadmin recover’ on
a repository that people were still accessing. Is it possible
that you would have run into that? (I don’t even know what
the “recover” would be used for…)
Yes, these are the sorts of problems we continually encountered. But
until you do that no one can do anything with the repository, and you
must run svn admin on the machine that is hosting the repository (at
least in our experience).
Basically, the current DB behind subversion is a Berkeley DB.
Apparently this DB gets corrupted.
···
On Jul 25, Garance A Drosihn wrote:
We used the same source for both platforms so the versions should have
lined up (although testing the Red Hat 9 version with our server yielded
errors not unlike what you describe). Which server did you use (webdav,
subversion, local)? Did you have multiple platforms?
···
On Jul 25, Oliver Bolzer wrote:
Brett H. Williams wrote:
I’ve been following subversion development for two years, and we finally
took it out for a more complete test drive a couple of months ago, using
itAt my place ( informatic dept @ univ ) we have been using Subversion for
about 9month now for several internal development projects as well as
giving students the possibility to create their own repositories. Not once
did we experience database corruption.The only major downside (other than that compiling is a PITA) we experienced
is, that the server and client versions need to be somewhat in synchro,
otherwise their communication breaks in unexpected ways (like XML errors or
plain “permissin denied” messages).
We are definitely going for a wiki per project.
Wikis are just too powerful a capability for distributed development to
not have them.
We are going to transition the FreeRIDE wiki over, and we may use that
wiki as the basis for RubyForge…I will let folks know.
-rich
···
On Wednesday, July 23, 2003, at 05:08 PM, Simon Strandgaard wrote:
Thinking of RubyForge + nice to have features:
The berlios project has one ‘wiki’ per project, eg:
http://developer.berlios.de/projects/ncurses-ruby
Hmm… upon a reread you are saying more than I first thought.
Basically you are saying that if someone does svnadmin recover while
someone else is attempting to access the repository, that we will have a
worse failure than DB corruption?
That may indeed have been what happened.
···
On Jul 25, Brett H. Williams wrote:
On Jul 25, Garance A Drosihn wrote:
When I asked my friend about this, he said there were several
cases of problems when some user runs ‘svnadmin recover’ on
a repository that people were still accessing. Is it possible
that you would have run into that? (I don’t even know what
the “recover” would be used for…)Yes, these are the sorts of problems we continually encountered. But
until you do that no one can do anything with the repository, and you
must run svn admin on the machine that is hosting the repository (at
least in our experience).Basically, the current DB behind subversion is a Berkeley DB.
Apparently this DB gets corrupted.
When I asked my friend about this, he said there were several
cases of problems when some user runs ‘svnadmin recover’ on
a repository that people were still accessing. Is it possible
that you would have run into that? (I don’t even know what
the “recover” would be used for…)Hmm… upon a reread you are saying more than I first thought.
Basically you are saying that if someone does svnadmin recover
while someone else is attempting to access the repository, that
we will have a worse failure than DB corruption?
Yes, that is what he told me. Well, “more DB corruption”, such
that one recover may appear to succeed, but it will set you up
for a subsequent DB corruption.
That may indeed have been what happened.
Okay.
···
At 9:02 AM +0900 7/25/03, Brett H. Williams wrote:
On Jul 25, Brett H. Williams wrote:
On Jul 25, Garance A Drosihn wrote:
–
Garance Alistair Drosehn = gad@gilead.netel.rpi.edu
Senior Systems Programmer or gad@freebsd.org
Rensselaer Polytechnic Institute or drosih@rpi.edu