-

Difference between revisions of "Server software"

From Open Bioinformatics Foundation
Jump to: navigation, search
(News Server)
m (News Server)
 
(One intermediate revision by one other user not shown)
Line 9: Line 9:
  
 
I don't think you HAVE to make a release before the changeover, but you just need to get your project coordinated so everyone knows about the ways to get code out.   
 
I don't think you HAVE to make a release before the changeover, but you just need to get your project coordinated so everyone knows about the ways to get code out.   
 +
 +
* BioPerl has a preliminary CVS to SVN [http://www.bioperl.org/wiki/CVS_to_SVN_Migration migration page] on the wiki to track progress; it just needs to be updated and eventually moved into the full SVN docs once they are up. --[[User:Cjfields|Cjfields]] 11:59, 29 October 2007 (EDT)
  
 
===How will people access the machines?===
 
===How will people access the machines?===
Line 19: Line 21:
  
 
ChrisD - what do you think?  Can we go ahead and get and start the wheels turning on testing all of these things so that once a project is ready to transition we can just implement the changeover in a few hours?  I know you have svnweb installed - is there any more testing that needs to be done?  Can you look either rsync or svnsyncing the test bioperl repository in /home/svn-repositories?
 
ChrisD - what do you think?  Can we go ahead and get and start the wheels turning on testing all of these things so that once a project is ready to transition we can just implement the changeover in a few hours?  I know you have svnweb installed - is there any more testing that needs to be done?  Can you look either rsync or svnsyncing the test bioperl repository in /home/svn-repositories?
 
  
 
==Web 2.0 @ OBF==
 
==Web 2.0 @ OBF==
Line 27: Line 28:
  
 
* I think consolidating the new feeds would be fine; we can always categorize/tag the posts.  Speaking as one who wrote them up for a brief time, mailing list summaries are quite time-consuming, so finding a way to semi-automate it might be easier (I think the Perl6 list used to do this). --[[User:Cjfields|Cjfields]] 11:56, 29 October 2007 (EDT)
 
* I think consolidating the new feeds would be fine; we can always categorize/tag the posts.  Speaking as one who wrote them up for a brief time, mailing list summaries are quite time-consuming, so finding a way to semi-automate it might be easier (I think the Perl6 list used to do this). --[[User:Cjfields|Cjfields]] 11:56, 29 October 2007 (EDT)
 +
* Consolidating them would be much better to provide a nice overview of all O|B|F projects' progress. People who require more detailed tracking of them are used to be subscribed to the corresponding lists. --[[User:Mauricio|Mauricio]] 11:09, 31 October 2007 (EDT)
  
 
===Wikis===
 
===Wikis===

Latest revision as of 15:09, 31 October 2007

This page can be dynamic so feel free to update and add your comments. We can archive parts by moving them to the discussion page. Please sign your comments though with ~~~


SVN

I would love to see the CVS to SVN migration happen by the end of the year for all of the OBF projects.

What would need from the project leaders is some sort of schedule. So you'd need to set a schedule for your project and coordinate how you want to make the migration happen and we (well George with some of the root-l people's help) can do the CVS to SVN changeover.

I don't think you HAVE to make a release before the changeover, but you just need to get your project coordinated so everyone knows about the ways to get code out.

  • BioPerl has a preliminary CVS to SVN migration page on the wiki to track progress; it just needs to be updated and eventually moved into the full SVN docs once they are up. --Cjfields 11:59, 29 October 2007 (EDT)

How will people access the machines?

  • We will still have people use SSH to get onto the dev.open-bio.org machine to make commits so we will still have to give shell accounts on the dev machine.
    • for security reasons we have chosen not run any apache or webdav on dev so we can't do SVN over https. I would be willing to consider it, but I don't know what the potential security flaws are here and our code repositories are an important asset so we have tried to keep the machine as locked down as possible.
  • We will rsync the repository from dev to code.open-bio.org (or possibly using svnsync i.e. svnsync info.
  • anonymous SVN checkout and Browsing can be done via http://code.open-bio.org


ChrisD - what do you think? Can we go ahead and get and start the wheels turning on testing all of these things so that once a project is ready to transition we can just implement the changeover in a few hours? I know you have svnweb installed - is there any more testing that needs to be done? Can you look either rsync or svnsyncing the test bioperl repository in /home/svn-repositories?

Web 2.0 @ OBF

News Server

Most projects aren't really using the news servers (bioperl.org/news). It is worth keeping these installed? Should we consolidate things to a single newsfeed (news.open-bio.org) with projects as categories instead? I was hoping that projects would be able to designate some people to post summaries of list traffic and project progress through the news system, but I know this takes a lot of time.

  • I think consolidating the new feeds would be fine; we can always categorize/tag the posts. Speaking as one who wrote them up for a brief time, mailing list summaries are quite time-consuming, so finding a way to semi-automate it might be easier (I think the Perl6 list used to do this). --Cjfields 11:56, 29 October 2007 (EDT)
  • Consolidating them would be much better to provide a nice overview of all O|B|F projects' progress. People who require more detailed tracking of them are used to be subscribed to the corresponding lists. --Mauricio 11:09, 31 October 2007 (EDT)

Wikis

I have started working to try and consolidate the installations so upgrades aren't such a big deal. I am sure the big meta-wiki sites have their way of doing this as well to make it easier, but I am not sure how I'd like to see it setup much differently.

There are quite a few orphan wiki sites right now - obda.open-bio.org, biosql.org- is there interest in still maintaining them? Otherwise we can redirect them to open-bio.org wiki and just flesh out some simple 1 page project descriptions there.

A potential crazy idea is to consolidate the content into a single wiki and just have project namespaces i.e. BioJava:Main_Page instead of the domain names (domain names would preferentially redirect to main page of a project i.e. biojava.org would link to wiki.open-bio.org/BioJava:Main_Page for example. Is this a good idea? Bad Idea? Harder to identify project branding with this I know, but some of the definition pages like multiple sequence alignment tools is harder.

  • I think keeping the wikis separate is better, though I can also see the benefit of consolidating some of the definition pages into one space, as well as moving anything which relates to more than one project (BioSQL, OBDA, etc). Maybe use the OBF wiki for that and link to those pages where needed? --Cjfields 17:51, 28 October 2007 (EDT)
  • I quite like OpenID as a single sign on system between multiple web sites. Would be nice to install the OpenID plugin. This could be used to identify the same user between our different wikis --Andreas 07:07, 29 October 2007 (EDT)

Ideas

If you have any feedback about how we move forward with web and web2.0 content for OBF let me know. We don't currently do much for networking developers and users, but we could try and emphasize things like facebook, linked in, scilink.com, or nature network as possible ways to try and connect everyone up a little more effectively beyond the mailing lists.

  • There is also the open source social networking site Ohloh. See the Biojava and BioPerl projects there. --Andreas 06:40, 29 October 2007 (EDT)

Mailing lists

We have had a steady set of volunteers dealing with mailing lists to moderate spam and other things. The Volunteers list had coordinated some of this. I'd like to check and see if this is still working well. Can we improve how moderation is done in any way - are there new software tools out there to make it easier to go to one place and see all the mail that needs to be moderated?

Recruiting drive

Part of some future stuff we'd like to do in OBF is recruit some more people to be part of the leadership. This is everything from being on the OBF Board to helping administer the machines, generate content for the websites and newsfeeds, and generally recognize the hard work that many people are doing behind the scenes. So if you are interested or want to help recruit people, please speak up.

I would also like to see us keep more of the history pages describing who and what have worked on the projects on the wikis. This is important to both give people credit for the work they did and for us to keep track of our history on the different projects.

  • you could consider installing automated tools to visualize the historic growth of CVS and SVN projects like StatCVS and StatSVN --Andreas 06:44, 29 October 2007 (EDT)

We are making strides to get the OBF as a 503(c) organization and so it will be important to show the progress we make if we solicit donations from corporations.

History

Initial edits by jason 14:19, 27 October 2007 (EDT)