#14512: Use ....: doctest continuations everywhere
-------------------------------------------------------------------------------------------------------------+
Reporter: jdemeyer
| Owner: mvngu
Type: enhancement
| Status: needs_review
Priority: trivial
| Milestone: sage-5.10
Component: doctest coverage
| Resolution:
Keywords:
| Work issues:
Report Upstream: N/A
| Reviewers:
Authors: Jeroen Demeyer
| Merged in:
Dependencies: #11745 #12894 #13698 #13961 #14252 #14287 #14299 #14424
#14461 #14477 #14479 #14483 #8703 | Stopgaps:
-------------------------------------------------------------------------------------------------------------+
Comment (by nthiery):
Replying to [comment:8 jdemeyer]:
> Replying to [comment:6 nthiery]:
> > Ouch ... This is likely to be a pain and create lots of conflicts ...
Is it really worth the trouble?
> I think it would be good to change these doctests. If we disallow the
`...` syntax, then we could finally write doctests with expected output
starting with ellipsis `...`.
Well, I very much hope the "..." syntax won't be disallowed right
away: people have a *lot* of .sage files out there which include
doctests. Dropping backward compatibility will force them to update
them at once. Which means they won't be able to share them properly
with other people running older versions of Sage (ok, the files are
still usable, but one won't be able to run the tests; for example to
make sure that everything works fine with all versions in use).
There are also printed or about to be printed *books* that have been
written with this syntax and sagetex. It's useful to be able to test
the examples in those books on various versions of Sage over a non
trivial amount of time.
I appreciate the gain of tests starting with "...:. All our "Traceback
..." lines are kind of silly. But it's really not critical either. We
lived without it for years, we could wait a bit more.
> > Well, luckily it should not be too hard to rebase the patches, and it
might get less noticed in the middle of the change of workflow.
> With some small modifications, the script on this ticket could be used
to auto-rebase patches. I could run this script when merging patches and I
could also run it on the sage-combinat queue for example. I think that
should ease most of the pain.
I appreciate the offer! Please take into account that, once rebased,
the patch won't apply anymore on "older" versions of Sage. Well,
unless you duplicate the patch in the queue, with guards, and all the
usual pain of duplication for things that are still evolving.
So as much I could be ok with this change occurring right at the time
the Sage-Combinat queue gets reborn as a git tree (with e.g. the
patches being auto-rebased when they are imported into git). As much I
would oppose a change before that. I guess that means not in any 5.*.
There will be a lot of turmoil ahead for us to handle the
transition. Please don't add yet another hurdle!
Cheers,
Nicolas
--
Ticket URL: <http://trac.sagemath.org/sage_trac/ticket/14512#comment:9>
Sage <http://www.sagemath.org>
Sage: Creating a Viable Open Source Alternative to Magma, Maple, Mathematica,
and MATLAB
--
You received this message because you are subscribed to the Google Groups
"sage-trac" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/sage-trac?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.
