On 30/05/12 21:58, C Anthony Risinger wrote:
On Wed, May 30, 2012 at 3:54 PM, C Anthony Risinger<[email protected]> wrote:
On Wed, May 30, 2012 at 7:48 AM, Rich Newpol<[email protected]> wrote:
Ooo my two cents - I like the override system, especially with the new
file-based (rather than directory-based) naming convention. My only issue
with the current override scheme is that it isn't always clear what you can
put in the override file and what you can leave out but there are lots of
examples to check.
Aside from that bit of head-scratching, it's nice that overrides are clearly
seen at the directory level, and 'special' code doesn't clutter the
'mainstream' code.
Basically, I like it.
yes i like it much more now that they are all at the same level ... i
guess i don't like the fact that many overrides are very minimal and
don't really necessitate a whole new file. i think this makes it
harder to actually remove dead code and/or reconcile the differences,
because they are not next to each other.
there is room to work here ... not really critical. maybe a mix of
both, idk, but i do want to increase the visibility somehow.
... also the reliance on specialized import/merging machinery kinda
sucks, and is brittle, IMO.
I'm not particularly concerned about how the overrides work, as long as
they do. My main concern is about the fact that event the most recent
versions of IE/MSHTML are having to run overrides that they don't need,
just because we can't distinguish between IE6/7 and later versions. It
irritates me because I use pyjd/MSHTML equivalent to IE8 or 9 as my
development vehicle, so I can often see that the standard code works
fine without an override.
Adding another platform such as IE8 would perhaps be overkill at this
point, although at some later point the IE6 platform is going to have to
migrate to some later version of IE. I wondered about adding version
checks to the IE6 overrides, so they could run standard code if the
browser was capable of it. Not a very tidy idea, however, and pyjs
doesn't currently seem to give access to the navigator object at run
time, anyway.
So maybe we just have to wait a year or two more and then cut the IE6/7
stuff out of the IE platform.
Phil
