[Biophp-dev] CVS

Nico Stuurman biophp-dev@bioinformatics.org
Thu, 1 May 2003 09:32:10 -0700


I lost track of where things are with code integration.  However, I 
making my last plea here to integrate the different code bases in a 
single cvs.  I uploaded the stuff from Serge to the Biophp cvs server, 
but since Serge has in the mean time been changing the code and putting 
in cvs at sourceforge, that was pretty much a mute thing to do.  If you 
all want to work as a team on code, by far the easiest way is to all 
use (the same) source code management system (like cvs).

Imagine, type 'cvs -z3 update -d' and get all the changes Serge made 
overnight to the code, look at them, send an email to the list with 
comments, make changes in the code, do 'cvs -z3 commit -m "Renamed 
class Parser to IOseq_Read" parser.inc.php' and have it immediately 
available to everyone else.  You can revert back at any time to any 
version ever in existence.  It is documented exactly who made what 
change, so credits can be given to where they are due.  By having it 
all in one place we'll never run into the stupid problems we just went 
through.

As I said before, it is simply impossible to use multiple cvs 
repositories with basically the same code.  As an example, I don;t have 
Serge's latest changes, and Serge has never seen the Parsers class.

If you feel insecure about having your code under someone else's 
control, simply download the whole cvstree on a regular basis, you will 
always be able to fork off and start your own cvs at the point where 
you left the group (and since it is all GPL, or isn't it?, you will 
also be legally allowed to do so).

Please, let's now really work as a team.


Best,


Nico