On 14/02/2009, at 10:09 AM, Paul Davis wrote:
Tests should be identifiable obviously, but if you want literature in
your code, go write AppleScript.
Code should prioritize easy comprehensibility, especially given that
there's no separate documentation. You're obviously not a literate
programming fan :)
And if you want terseness, got write APL - I did that for two years,
and it's a heap of fun.
Our requirements are a working HTTP client implementation, and a
browser's
conformance to that isn't about it's Couch support. I think this
should be a
separate project.
I'm confused by the the argument that we should treat in browser tests
differently for a vote against treating in-browser tests differently.
We shouldn't have in-browser tests. In-browser tests are about testing
the browser, which we shouldn't do in this project. Our tests should
be about Couch, on the command-line, build-farm-integratable.
If those tests are also runnable from the browser, just for chuckles,
then fine, but that's not the canonical environment because it
introduces another variable into the mix.
Antony Blakey
-------------
CTO, Linkuistics Pty Ltd
Ph: 0438 840 787
He who would make his own liberty secure, must guard even his enemy
from repression.
-- Thomas Paine
