On 2014-06-03, 3:34 PM, Boris Zbarsky wrote:
On 6/3/14, 2:36 PM, Gregory Szorc wrote:
There is a clear win in the ability to reuse, understand, and modify the
common code.
No one is arguing against having common harness code as far as I can see.
I can't even recall which file(s)
contain is/ok from mochitest
SimpleTest.js. Which also contains everything else from mochitest.
Assert.jsm makes logic centralized, consistent, and easy to extend.
No one is arguing against that.
I'm arguing against Assert.jsm using the commonjs API names.
Same here. Agreed on all of your other points.
FWIW, a lot of this bikeshedding about what effectively boils down to
style could be avoided if we had JavaScript code rewriting facilities
that could atomically change all references.
That wouldn't make writing new tests any easier, though it would of
course ease the "inconsistent styles during conversion" issue.
Also note that those rewriting facilities do not exist, and the current
proposal is explicitly asking for the exact opposite, which is, no
rewrites, new tests being written in one style, old tests remaining in
another style, which will mean that the inconsistency between which APIs
the test uses depending on when it was introduced (and frankly who wrote
it as well most likely) indefinitely into the future.
Nobody is opposing a proposal where someone volunteered to create the
rewriting facilities you're mentioning and using them to fix the entire
code base in one go.
Cheers,
Ehsan
_______________________________________________
dev-platform mailing list
dev-platform@lists.mozilla.org
https://lists.mozilla.org/listinfo/dev-platform
