Let me clarify:
Telling me I'm doing it wrong and then telling me to do exactly what I
was already doing is killing my will to contribute. I'm not asking to
hold commits because it's more convenient for me, I was asking to
commit what you had outstanding so I could integrate that. As a
service to you.
Since my commit is a major refactor of each and every file in the
FlexJS JS framework, every commit will be a conflict if it's based on
the 'old' code. I asked everyone to hold their contributions for a
short time to not confuse things. Since git is no better (and often
worse) than SVN at resolving merge conflicts (and please don't mention
the legendary "merge hell", that's a function of source control, not
SVN in particular), I was trying to make it as easy as I could for
everyone involved.
While we're on the subject: it's not as if git has brought anything
useful to this project, and it has cost us a dearly. We're still
nowhere near being able to do a release of the SDK, let alone having
automated builds or contributions of newbies. And it's been months
since the move... The hordes of contributors that would come to our
help and make this the best project ever are still successfully hiding
somewhere...
Keep telling me that git will be a miracle tool that will solve all
kinds of problems - which we didn't have - makes no sense anymore. And
telling me that I need to stick with it because I just don't get it,
and that it will grow on me is plain insulting, as I by now probably
know git better than most, just so I could keep contributing. I hate
this elitist "it's more complicated to use, so it's better and if you
don't get that you're not worthy". Sometimes more complicated just
means less useful. The fact that even the people who use git for a
living can't agree on the way to use it - which by the way cost us a
valued committer - is proof enough of that.
Thoroughly demotivated, in full drama queen mode,
EdB
On Sat, May 4, 2013 at 6:18 PM, Erik de Bruin <e...@ixsoftware.nl> wrote:
> Git is killing this project.
>
> EdB
>
>
>
> On Sat, May 4, 2013 at 10:36 AM, Carlos Rovira
> <carlos.rov...@codeoscopic.com> wrote:
>> Erik,
>>
>> sorry for the inconvenience, but GIT is just about not have to wait. Your
>> mind are still in SVN and still does not switch to GIT.
>>
>> We at work are always merging and resolving conflicts in integration
>> branches. We never merge directly in develop. For this kind of brutal
>> changes you should:
>>
>> 1.- start an integration branch
>> 2.- make merges in that brach (and maintain it over time what could be
>> cumbersome as time and divergences happen)
>> 3.- test that all is ok
>> 4.- merge back into develop
>>
>> Erik, I know you are not yet a fan of GIT but think that the way that I'm
>> propossing to you could never be done in SVN and most important, you don't
>> stop other people works.
>>
>>
>>
>>
>>
>> 2013/5/4 Erik de Bruin <e...@ixsoftware.nl>
>>
>>> Thanks for waiting with this commit till after I land the
>>> 'goog.events' branch :-(
>>>
>>> EdB
>>>
>>>
>>>
>>> On Fri, May 3, 2013 at 11:48 PM, Carlos Rovira
>>> <carlos.rov...@codeoscopic.com> wrote:
>>> > Hi Alex,
>>> >
>>> > fixed in my latest commit, not JS and SWF works ok. Things look better at
>>> > daylight ;).
>>> > Only JS side need to improve applicationModel to be getter/setter.
>>> >
>>> > I change SimpleBinding fix to try catch as you suggested.
>>> >
>>> > Best,
>>> >
>>> > Carlos
>>> >
>>> >
>>> > 2013/5/3 Alex Harui <aha...@adobe.com>
>>> >
>>> >> Hi Carlos,
>>> >>
>>> >> I don't think the change to ViewBase.as was correct. The
>>> applicationModel
>>> >> property should be a getter. It looks like you reverted to a var.
>>> >>
>>> >> See if that fixes the SWF situation.
>>> >>
>>> >> -Alex
>>> >>
>>> >> On 5/2/13 6:17 PM, "Carlos Rovira" <carlos.rov...@codeoscopic.com>
>>> wrote:
>>> >>
>>> >> > Alex,
>>> >> >
>>> >> > I shared a branch with some changes
>>> "feature/applicationModel_refactor".
>>> >> >
>>> >> > applicationModel is the variable that makes createjs sample fail with
>>> the
>>> >> > change in simplebinding. I refactor to the way you show (hopefuly).
>>> >> >
>>> >> > In the JS version all is working right. But now SWF compilations
>>> fails in
>>> >> > runtime for both, _again and createjs sample (in different ways).
>>> >> >
>>> >> > Could you take a look and see if this refactor is in the way you
>>> >> > commented?. I'll take a look tomorrow and see why SWFs are failing.
>>> >> >
>>> >> >
>>> >> >
>>> >> > 2013/5/3 Carlos Rovira <carlos.rov...@codeoscopic.com>
>>> >> >
>>> >> >> Hi Alex,
>>> >> >>
>>> >> >> now I understand the problem. I was not aware of @expose and what you
>>> >> >> comment here. Very important information. Tomorrow I'll go back to
>>> the
>>> >> >> sample and see if I can change the sample to get working with that
>>> style
>>> >> >> and remove this fix since I think it's better to force people to make
>>> >> >> things in the good way.
>>> >> >>
>>> >> >>
>>> >> >>
>>> >> >> 2013/5/2 Alex Harui <aha...@adobe.com>
>>> >> >>
>>> >> >>> Hi Carlos,
>>> >> >>>
>>> >> >>> I saw the changes.
>>> >> >>>
>>> >> >>> I think it is a good fix, but you might want to add a comment that
>>> the
>>> >> >>> non-getter approach is likely to fail for non-public variables when
>>> >> >>> minified.
>>> >> >>>
>>> >> >>> Right now, binding in FlexJS only works for public properties and
>>> your
>>> >> fix
>>> >> >>> makes it work for public variables. Note that in some cases, the
>>> >> backing
>>> >> >>> variable for a public get/set is the same name because in JS the
>>> >> pattern
>>> >> >>> looks like:
>>> >> >>>
>>> >> >>> /** @private */
>>> >> >>> var foo;
>>> >> >>>
>>> >> >>> /** @expose */
>>> >> >>> function get_foo()
>>> >> >>> {
>>> >> >>> return this.foo;
>>> >> >>> }
>>> >> >>>
>>> >> >>> The question is, what did the AS look like? If it looked like
>>> >> >>>
>>> >> >>> private var _foo;
>>> >> >>> public function get foo()
>>> >> >>> {
>>> >> >>> return this._foo;
>>> >> >>> }
>>> >> >>>
>>> >> >>> Then the JS code for your failure case should be changed to look
>>> like
>>> >> the
>>> >> >>> recommended pattern and then you don't need the [this.sourceID] code
>>> >> path
>>> >> >>> as
>>> >> >>> there will be a get_ function.
>>> >> >>>
>>> >> >>> If the AS looked like:
>>> >> >>>
>>> >> >>> public var foo;
>>> >> >>>
>>> >> >>> Then the JS should be
>>> >> >>>
>>> >> >>> /** @expose */
>>> >> >>> Someclass.prototype.foo;
>>> >> >>>
>>> >> >>> If they forget to use @expose then it will likely fail when
>>> minified.
>>> >> >>>
>>> >> >>> One more minor thing: Do you happen to know if try/catch is more
>>> >> >>> efficient
>>> >> >>> than doing a property lookup to see if it is undefined before
>>> making a
>>> >> >>> call?
>>> >> >>>
>>> >> >>> Anyway, thanks for contributing. We can use all the help we can
>>> get.
>>> >> >>>
>>> >> >>>
>>> >> >>> On 5/2/13 2:35 PM, "Carlos Rovira" <carlos.rov...@codeoscopic.com>
>>> >> wrote:
>>> >> >>>
>>> >> >>>> Hi Alex,
>>> >> >>>>
>>> >> >>>> I commit a solution that fixes the problem, but I don't know if is
>>> >> >>>> compliant with the motivation you make this change (minified
>>> version).
>>> >> >>>>
>>> >> >>>> I think binding should work even if we don't have a getter/setter
>>> >> setup,
>>> >> >>>> but I'm to new to this new framework that maybe this is not what
>>> you
>>> >> >>> have
>>> >> >>>> in mind.
>>> >> >>>>
>>> >> >>>> So it's up to you to change if you see some problem with this fix.
>>> >> >>>>
>>> >> >>>>
>>> >> >>>> 2013/5/2 Alex Harui <aha...@adobe.com>
>>> >> >>>>
>>> >> >>>>>
>>> >> >>>>>
>>> >> >>>>>
>>> >> >>>>> On 5/1/13 3:40 PM, "Carlos Rovira" <carlos.rov...@codeoscopic.com
>>> >
>>> >> >>> wrote:
>>> >> >>>>>
>>> >> >>>>>> Hi Alex
>>> >> >>>>>>
>>> >> >>>>>> this latest change in SimpleBinding:
>>> >> >>>>>>
>>> >> >>>>>> - this.source = this.document[this.sourceID];
>>> >> >>>>>> + this.source = this.document['get_' + this.sourceID]();
>>> >> >>>>>>
>>> >> >>>>>>
>>> >> >>>>>> Makes the createjs sample to not generate controls far beyond a
>>> >> >>>>>> SimpleBinding.
>>> >> >>>>>>
>>> >> >>>>>> I need to change the sample code or it's a bug?
>>> >> >>>>> Well, using getters is correct. Are you binding to something that
>>> >> >>> isn't a
>>> >> >>>>> getter in AS? Maybe the JS version of that property should have
>>> >> been a
>>> >> >>>>> get/set pair.
>>> >> >>>>>
>>> >> >>>>> But if you want, you can also try to make SimpleBinding tolerant
>>> and
>>> >> >>> find
>>> >> >>>>> things that aren't getters.
>>> >> >>>>>>
>>> >> >>>>>> Thanks
>>> >> >>>>>
>>> >> >>>>> --
>>> >> >>>>> Alex Harui
>>> >> >>>>> Flex SDK Team
>>> >> >>>>> Adobe Systems, Inc.
>>> >> >>>>> http://blogs.adobe.com/aharui
>>> >> >>>>>
>>> >> >>>>>
>>> >> >>>>
>>> >> >>>
>>> >> >>> --
>>> >> >>> Alex Harui
>>> >> >>> Flex SDK Team
>>> >> >>> Adobe Systems, Inc.
>>> >> >>> http://blogs.adobe.com/aharui
>>> >> >>>
>>> >> >>>
>>> >> >>
>>> >> >>
>>> >> >> --
>>> >> >> Carlos Rovira
>>> >> >> Director de Tecnología
>>> >> >> M: +34 607 22 60 05
>>> >> >> F: +34 912 94 80 80
>>> >> >> http://www.codeoscopic.com
>>> >> >> http://www.directwriter.es
>>> >> >> http://www.avant2.es
>>> >> >>
>>> >> >
>>> >> >
>>> >>
>>> >> --
>>> >> Alex Harui
>>> >> Flex SDK Team
>>> >> Adobe Systems, Inc.
>>> >> http://blogs.adobe.com/aharui
>>> >>
>>> >>
>>> >
>>> >
>>> > --
>>> > Carlos Rovira
>>> > Director de Tecnología
>>> > M: +34 607 22 60 05
>>> > F: +34 912 94 80 80
>>> > http://www.codeoscopic.com
>>> > http://www.directwriter.es
>>> > http://www.avant2.es
>>>
>>>
>>>
>>> --
>>> Ix Multimedia Software
>>>
>>> Jan Luykenstraat 27
>>> 3521 VB Utrecht
>>>
>>> T. 06-51952295
>>> I. www.ixsoftware.nl
>>>
>>
>>
>>
>> --
>> Carlos Rovira
>> Director de Tecnología
>> M: +34 607 22 60 05
>> F: +34 912 94 80 80
>> http://www.codeoscopic.com
>> http://www.directwriter.es
>> http://www.avant2.es
>
>
>
> --
> Ix Multimedia Software
>
> Jan Luykenstraat 27
> 3521 VB Utrecht
>
> T. 06-51952295
> I. www.ixsoftware.nl
--
Ix Multimedia Software
Jan Luykenstraat 27
3521 VB Utrecht
T. 06-51952295
I. www.ixsoftware.nl
