So I finally got my load_every_help.py script pretty well debugged and
so it can actually survive a bunch of conditions. Basically, its a
simple automated test of every help patch. You give it a base
directory for Pd, and it opens every help patch and logs the result.
This makes it easier to see which help patches have problems. I could
see this script doing something like posting anytime there is a
"couldn't create" error.
Here's an example log, I ran it without X on Debian, so all of the Gem
and probably PDP stuff failed.
http://autobuild.puredata.info/pdlab/load_every_help-2009-06-19_01.04.42.log
I think this script could also then serve as the basis of a regression
testing script. It would basically look thru the SVN for any *-
regression.pd file. These pd patches would run some bit of code, then
print out the word "SUCCESS". If it doesn't that's a fail and it'd be
automatically mailed to the list.
.hc
If you are not part of the solution, you are part of the problem.
Hallo,
interesting, and surely a pretty useful tool. Checking the results of some of my own [list]-abs help-files, I'm happy to see only two unexpected errors (a third one is deliberate). However these two should not be reported as errors: It's about [list-abs] not created, while list-abs.pd is just a simple abstraction in the [list]-abs collection and lives right next to the respective help-files. So it should have no problems loading and there definitely are not problems on a standard Pd.
Ciao
Frank
Hans-Christoph Steiner hat gesagt: // Hans-Christoph Steiner wrote:
So I finally got my load_every_help.py script pretty well debugged and
so it can actually survive a bunch of conditions. Basically, its a
simple automated test of every help patch. You give it a base directory for Pd, and it opens every help patch and logs the result. This makes it easier to see which help patches have problems. I could see this script doing something like posting anytime there is a "couldn't create" error.Here's an example log, I ran it without X on Debian, so all of the Gem
and probably PDP stuff failed.http://autobuild.puredata.info/pdlab/load_every_help-2009-06-19_01.04.42.log
I think this script could also then serve as the basis of a regression
testing script. It would basically look thru the SVN for any *- regression.pd file. These pd patches would run some bit of code, then
print out the word "SUCCESS". If it doesn't that's a fail and it'd be
automatically mailed to the list..hc
If you are not part of the solution, you are part of the problem.
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
I just uploaded a log from running this script on Mac OS X 10.5.7/ Intel (the previous was on Debian/stable/i386)
http://autobuild.puredata.info/pdlab/load_every_help-2009-06-20_10.25.56.log
This one is in the same order as the Help Browser, so its easier to
skim. Turns out there are a number of errors in Miller's docs.
.hc
On Jun 20, 2009, at 5:07 AM, Frank Barknecht wrote:
Hallo,
interesting, and surely a pretty useful tool. Checking the results
of some of my own [list]-abs help-files, I'm happy to see only two unexpected
errors (a third one is deliberate). However these two should not be reported
as errors: It's about [list-abs] not created, while list-abs.pd is just a simple abstraction in the [list]-abs collection and lives right next to the
respective help-files. So it should have no problems loading and there
definitely are not problems on a standard Pd.Ciao
Frank
Hans-Christoph Steiner hat gesagt: // Hans-Christoph Steiner wrote:
So I finally got my load_every_help.py script pretty well debugged
and so it can actually survive a bunch of conditions. Basically, its a simple automated test of every help patch. You give it a base
directory for Pd, and it opens every help patch and logs the result. This
makes it easier to see which help patches have problems. I could see this
script doing something like posting anytime there is a "couldn't create"
error.Here's an example log, I ran it without X on Debian, so all of the
Gem and probably PDP stuff failed.http://autobuild.puredata.info/pdlab/load_every_help-2009-06-19_01.04.42.log
I think this script could also then serve as the basis of a
regression testing script. It would basically look thru the SVN for any *- regression.pd file. These pd patches would run some bit of code,
then print out the word "SUCCESS". If it doesn't that's a fail and it'd
be automatically mailed to the list..hc
If you are not part of the solution, you are part of the problem.
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
Pd-list@iem.at mailing list UNSUBSCRIBE and account-management -> http://lists.puredata.info/listinfo/pd-list
¡El pueblo unido jamás será vencido!
-
Frank Barknecht -
Hans-Christoph Steiner