I am working on setting up a really nice development environment for once, I have machines i have started work on before, but never finished. I stripped parts from a older server that was included in a the lot of equipment I purchased to combined them with an older, but still good Pentium 4.

The Pentium 4 is the original socket 423 with RDRAM. If you never had the pleasure of owning a computer with RDRAM, 1024mb is like $500 for PC800. I was able to get 1gb of it on eBay cheap, like $65 shipped so I am combining that with the Ultra160 SCSI RAID card from the other server and the 3 18gb hard drives to create a nice development server. I will of course be running Slackware, I gave CentOS a try and I just couldn’t deal with the packages. I will give it another go sometime, but I actually prefer compiling. The specs of the machine are listed on my page about my machines.

I have enjoyed the subversion integration in trac when viewing revisions and plan to use trac more often, but the problem is that I didn’t like milestone1, milestone2, etc. I decided to make a few changes to the source to allow for something that fits my projects a little better.

In PYTHON_ROOT/site-packages/trac/db_default.py

##
## Default database values
##

# (table, (column1, column2), ((row1col1, row1col2), (row2col1, row2col2)))
def get_data(db):
   return (('component',
             ('name', 'owner'),
               (('component1', 'somebody'),
                ('component2', 'somebody'))),
           ('milestone',
             ('name', 'due', 'completed'),
               (('milestone1', 0, 0),
                ('milestone2', 0, 0),
                ('milestone3', 0, 0),
                ('milestone4', 0, 0))),
                .......

Unless I am wrong chaning the values of the milestones is all you will need to do. I tested it with a new environment and it appeared to work properly.