I'd like to resolve the differences between the lzunit code in trunk and the lzunit code in legals also. I vaguely remember maybe Tucker
reverted to an older version from before Ben modified lzunit , in order to not depend on the lz components for presenting the results (?)
On 8/21/06,
Jim Grandy <[EMAIL PROTECTED]> wrote:
I didn't have a mechanism in mind, I just know we should be running lztest against browsers on the build machines.jim
On Aug 20, 2006, at 6:11 PM, Henry Minsky wrote:Well I will check if the debug logging-to-server stuff is still working, Is that the mechanism you are thinking of using
for running lztest in a real browser?On 8/20/06, Jim Grandy <[EMAIL PROTECTED] > wrote:I didn't know the warnings were a show-stopper -- I thought we had dealt with the fatal ones in a separate step. If they are indeed a show-stopper, then we should definitely get back to them sooner rather than later.Having said that, LPP-2406 is basically an offshoot of the class conversion task, isn't it? If so, it would make sense to do at the same time.I'm also concerned about pushing out the lztest improvements. Anyone else want to volunteer to take a pass at the lztest improvements? (Getting automated lztest runs against real browsers and Flash is another important task that should be taken on at the same time.) If I have a volunteer for the lztest work, I'm fine with having Phil work on the warnings next.We *really* need to beef up our unit test infrastructure as much as possible.jimOn Aug 20, 2006, at 2:00 PM, Henry Minsky wrote:Right I second that opinion, because Phil Apley is going to hit all of these immediatealy and be dead in the water as soon
as he tries to run any components. Although I guess it's a good way to get intimate with the LFC quickly., But we don't want three different people fixing the same issues as they encounter them indepenently.
On 8/20/06, P T Withington < [EMAIL PROTECTED] > wrote:FWIW, I would vote for moving the warning removal ahead of 2406 and
tests. My reasoning:
1) As Henry discovered, as soon as you try _anything_ other than
LZPIX, you run into warnings, which halt the app.
2) Phil already went through this exercise, so he can more
expediently fix it than anyone else.
3) The only reason we held off on Phil's patch last time was lack of
QA resources, I had already done the technical review.
[Anecdote: every time Henry asked me about a warning he hit, I knew
the answer from reviewing Phil's patch.]
On 2006-08-20, at 14:31 EDT, Jim Grandy wrote:
> Hi Henry,
>
> Here's Phil's list of priorities:
>
> - finish class conversions
> - LPP-2406
> - lztest improvements (asynchronous tests, and intentional failure)
>
> Then let's see where we are. Getting rid of LFC warnings is the
> most likely next step, but in the time is takes to do those three
> items our priorities might have changed.
>
> jim
>
> On Aug 18, 2006, at 3:32 PM, Henry Minsky wrote:
>
>> Hi Phil,
>>
>> You did some work a while back to find and work around a lot of
>> warnings in the LFC, that I think it would be a good time
>> to try to update and merge them in. I'm starting to try some of
>> the other demo apps and there are
>> dozens of references to undefined slots which are causing DHTML to
>> stop with errors.
>>
>> I think we should make a big sweep to catch a bunch of these so we
>> can move to the next step
>> with components and other apps.
>>
>> I'm going to define some placeholders for some real missing
>> classes like focus manager, mode manager, lzcommand, etc..
>>
>> --
>> Henry Minsky
>> Software Architect
>> [EMAIL PROTECTED]
>>
>
--
Henry Minsky
Software Architect
[EMAIL PROTECTED]
--
Henry Minsky
Software Architect
[EMAIL PROTECTED]
--
Henry Minsky
Software Architect
[EMAIL PROTECTED]
_______________________________________________ Laszlo-dev mailing list [email protected] http://www.openlaszlo.org/mailman/listinfo/laszlo-dev
