On Tuesday, July 29, 2014 2:24:10 PM UTC-4, Volker Braun wrote:
>
>
> What do we test with 1000 point resolution that can't be tested with 10?
>

Maybe more examples should be flagged with "# not tested" or something like 
"# optional -- demo" or "# optional -- tutorial"?

This and another recent thread have clarified for me that the doctests 
serve two goals which often overlap but are actually distinct:

1. Test that code produces the expected output
2. Document what sage can do

The 1000 point resolution is not so useful for 1, but is useful for 2 -- it 
illustrates that sage can produce very high quality graphics.  A number of 
the other long doctests are probably more useful for 2 than 1.  Rather than 
remove them, it seems better to flag them in some way so they're not tested.

Unfortunately, I think the wording of the "# not tested" flag is misleading 
for novice users (a large part of the target audience) -- it might be 
mistaken to mean that the example is not expected to work, or hasn't ever 
been tested by anyone.  I don't think there's any harm in some version of 
the "# optional" flag, except we'd have to agree on yet another convention 
that we'll instantly forget, or agree to just use "# optional -- [whatever 
strikes my fancy, so long as it doesn't conflict with the other optional 
flags]".

Surely this kind of problem has come up elsewhere in the Python community 
-- is anyone familiar with it?

-Niles

p.s.  It occurs to me that if we do agree on a new optional flag, then 
these examples could be tested every so often if one wanted.


-- 
You received this message because you are subscribed to the Google Groups 
"sage-devel" 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-devel.
For more options, visit https://groups.google.com/d/optout.

Reply via email to