On 08/ 7/10 04:48 AM, Mitesh Patel wrote:
We're in the *very* early stages of developing a new doctesting API at
http://trac.sagemath.org/sage_trac/ticket/9224
Unifying sage-test and sage-ptest does make sense.
I'm not over-impressed with the current system. I've had tests fail with zero
failures and all many of other inconsistencies. Run make ptestlong and one gets
a failure. Run the test from the command line and it passes. I've no idea why.
I've seen a comment from I believe Dan on one of the tickets that ptest has some
alful code in it.
Also useful would be if the data and time of the test was recorded. Then if a
failure does occur, one could look at the system log files and see if there was
an I/O problem, running out of swap space or similar. Just make sure it's done
in a portable way.
Other useful things would be to have at the very least the host name and
operating system stored. Perhaps once at the top, and once in the summary to
record that information.
I'd personally like to see some way of testing Sage from the browser interface.
I've posted a maxima example before, where one of the examples in the Sage
manaual was incredibly badly formatted (unusable) in the browser front end, yet
that same doctest has passed. That however would need a totally different
approach - a non-trivial one at that! But perhaps the ability to run the test
scripts from inside the Sage session would go some way to resolve this.
--
To post to this group, send an email to sage-devel@googlegroups.com
To unsubscribe from this group, send an email to
sage-devel+unsubscr...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/sage-devel
URL: http://www.sagemath.org