[WWdevel] Re: CVS alternative

[John Jones]

First, I am copying Bill and Jeff since I don't know if they get 
openwebwork-devel e-mail.  Also, since Jeff (and some openwebwork-devel 
readers) may not know what we are talking about, here is the plan.

For the .pg files of the National Problem Library (perhaps to be renamed 
the National Problem Database), we will start using a cvs-like system.  
There will be two repositories, or maybe two directories of one 
repository.  The basic distinction is tagged vs. non-tagged problems.

Problems start in the non-tagged side, basically however we find them.  
Once this thing is initialized, I guess we can start filling that up 
with lots of pg files.  When it gets tagged, then it is moved to the 
tagged-side, which will be organized to mirror the heirarchical topic 
structure of the database.

We may not be able to "polish" every problem, but as that is done, it 
simply gives an updated version of the problem on the tagged side.  The 
setup as described above basically gives up on the notion of 
systematically polishing problems.  If we want to keep that alive, we 
should have 3 basic sub-divisions (raw, tagged, and 
tagged-and-polished).  Actually, this 3-part version might be a good way 
to go.

Thinking of the operations we will need in a cvs-like system:

    * add directories and files - I would hope all systems are good at this
    * move a file - cvs may be weaker here since it looses version
      history if you move a file, but maybe we don't really care.  We
      don't expect much revision to take place before tagging.
    * look at recent updates, and maybe the diffs.  This would be
      important as new versions are committed to the tagged files (e.g.,
      to be sure no one deleted the tags, or to see if the new version
      should be forked instead of being a new version of the same
      problem).  I don't know which is better here.  It probably hinges
      on how useful the status commands are for pulling information
      (since I wouldn't want to browse through thousands of files
      looking for recent changes).

I don't know enough about different systems to know which will be better 
for us on these things.

John


Sam Hathaway wrote:

> On Jan 10, 2005, at 12:08 PM, Michael Gage wrote:
>
>> We've been considering this, but haven't made any moves yet -- mostly 
>> because we've
>> had other things to do.  Sam, John, do you have any comments?
>
>
> Arch is more ambitious and has a greater "cool" factor, but I think 
> Subversion is the way to go right now, especially since it's possible 
> to convert a repository from CVS to Subversion: 
> <http://svnbook.red-bean.com/en/1.0/apas11.html>.


> -sam
>
>> On Jan 10, 2005, at 11:52 AM, William Ziemer wrote:
>>
>>> Most of the projects I link with are going to subversion:
>>> http://subversion.tigris.org/
>>> Maybe use this for the problem database?
>>> Good to see you all again,
>>> Bill
>>
>
>
>
> -------------------------------------------------------
> The SF.Net email is sponsored by: Beat the post-holiday blues
> Get a FREE limited edition SourceForge.net t-shirt from ThinkGeek.
> It's fun and FREE -- well, almost....http://www.thinkgeek.com/sfshirt
> _______________________________________________
> OpenWeBWorK-Devel mailing list
> OpenWeBWorK-Devel at lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/openwebwork-devel


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://webwork.maa.org/pipermail/webwork-devel/attachments/20050110/c3614a4b/attachment.html>