launchpad-dev team mailing list archive

[Deryck Hodge <deryck.hodge@xxxxxxxxxxxxx>]

On Sun, Feb 13, 2011 at 1:00 PM, Paul Hummer <paul.hummer@xxxxxxxxxxxxx> wrote:
> On 02/11/2011 04:09 PM, Robert Collins wrote:
>> On Sat, Feb 12, 2011 at 11:24 AM, William Grant
>> <william.grant@xxxxxxxxxxxxx> wrote:
>>> I would really prefer that we stopped including external dependencies in
>>> the tree. Is there some other way we can do it? sourcecode, maybe? Our
>>> tree is for our code, not embedding everyone else's.
>> Agreed; thats what eggs etc are for. As this is YUI, and lazr-js wraps
>> UI, perhaps lazr-js is a decent initial approximation for these third
>> party YUI widgets?
>
> I would also agree.  In fact, we already have some gallery code in lazr-js.
>

I disagree. Just to be crystal clear, if I wasn't before. :-) :-)

Ok, so I agree in spirit that gallery code should live in lazr-js
*eventually*.  But I think the "develop widgets in lp and pull back
when appropriate rule" still holds.  This work is not complete and
Brad doesn't know the value or usefulness of this widget yet.  When he
does, it will be appropriate to pull it into lazr-js.  If it doesn't
work out, it's easier to kill it from the lp tree, rather than having
to roll a new lazr-js tarball.

Also, I think we're making too much of this "don't dirty our code base
with third party modules" line of thinking.  JavaScript is not Python.
 eggs don't work for js, and it's much more work to treat lazr-js as a
bzr branch in sourcecode at this point.  Maybe that's appropriate at
some point, especially once Paul's work is done in lazr-js, but as I
said in the earlier mail, this is all premature now.

== Warning: Rant ==

Finally -- and I hope you guys receive this in the love and respect I
hold for you all :-) -- this whole thread is frustrating to me.  This
is exactly the sort of thing that that turns people off from Launchpad
development.  Brad didn't ask anyone where this code should live.  He
did some work he needed to do.  He coordinated with people working in
this area of the code.  He took *extra* effort to do it in a clean
way, a way that allows other legacy code to be moved away, and then he
notified all of us about this work so we could build on.  Then, you
all come along and tell him he did it wrong and he should now refactor
it. :-)  Okay, if he *really* screwed up, that's completely
appropriate.  But we're debating where a couple js and css files
should live on disk. ;)  I'm also completely frustrated by how few
people actually do any work to make the JavaScript development story
better, and yet, as soon as someone does some work in this area, 2-3
people pop up and deride the work for not being correct.

Cheers,
deryck


-- 
Deryck Hodge
https://launchpad.net/~deryck
http://www.devurandom.org/