This CleanUp branch has taken on a life of it's own. The change to DOM has
been 'cleaner' -- I trust/know everything that goes on -- but it hasn't been
easy. [It is amazing how many hours of coding/debugging I've had to put in
to get some weak approximation of what Sam achieved in w/ some Pythonic
wizardry. Oh well, at least I can maintain this.] The features of XML merge
that I know we need seem to work, although I have to more closely compare
the run output.
Since the CleanUp branch has dragged on so long, I've added a few extra,
including the Depot integration (which is pretty close & a chore for the
next week of so). I'll ensure those features are in, and complete, before I
merge this branch back in.
The LSD server hasn't worked the re-install of the platform. Since the
workspace working area root (on that machine) is above where Gump is
installed, I had to resort to a re-naming. The Gump there is checked out
into gump-install. Since Forrest is not installed on that machine, I've
checked out the CleanUp branch & will be attempting an XHTML run. I'll will
try to monitor this build.
Hmm, when I go to the LSD gump it redirects to Brutus public. Leo, this long
term? Do you want Gump off this machine, or just avoiding stale results?
The Brutus JDK15 server needs some work. I focused on XHTML with CleanUp,
but then Leo was nice enough to do the work to install the JDK1.5 build as
XDOCS (and even install JDK1.5) & it found bugs with the new 'sync as you go
code'. I'll try to work on those shortly, so we get a truer comparison of
new branch to CVS HEAD.
http://brutus.apache.org/gump/jdk15/
I've tried to keep the gump.try package environment up to date with Brutus,
and so the differences in build percentages are clearly due to code issues.
[Last night's run shows me the fix I need to apply next, for txt2html-task.
"Failed to perform mkdir <gump.model.misc.Mkdir instance at
0xb5d6792c>:'unicode' object has no attribute 'contains' ".[
http://gump.try.sybase.com/buildLog.html
BTW: Brutus doesn't appear to have rsync installed on it. Do folks think it
ought?
Next steps:
After 4th July (study) break, and my tests that following week, I'll try to
get this all cleaned up & merged back into CVS HEAD. At that point I'll
probably work on getting Python Gump into SVN & metadata in a separate CVS
module, and taking the project to that next stage (leaving traditional
further behind). I'll try to test Brett's Maven patch also.
regards,
Adam
--
Experience the Unwired Enterprise:
http://www.sybase.com/unwiredenterprise
Try Sybase: http://www.try.sybase.com
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
