On Wed, Jul 7, 2010 at 2:25 AM, Richard Jones <[email protected]>wrote:
> On 07/07/2010, at 3:12 PM, claudio canepa wrote: > > I think you are trowing the confidence we all built but running code for > months just for not write > > super(xxx, self).zzzz > > I think I understand where you're coming from. The problem I see is that > it's not consistent and when I look at teaching it the inconsistencies > become quite difficult. > In the *absence* of init , Scene and Layer are consistent with all their cousins: the other CocosNode subclasses ( Draw, Sprite). You plan to add to the other classes ? Will something else broke ? super( ) is a nuisance, but is common python idiom, and you must use if you override other methods. ( on_enter, on_exit, visit, draw, etc ). Then, *without* the init you have consistence between different methods. > An then, the existing uses of init in other code section are consistent with those new inits ? are compatible ? What exactly are those other inits ? I'm not telling the idea needs to be discarded, only that needs more investigation, and that hours before the planned release it is not the moment for a not so clear change. I am confident that the current cocos2d tests and samples all work. > It is a good indicator, but purposely they are very simple scripts, which don't stress too many code interactions. > > I've tested all the old cocos2d games I've got lying around and apart from > the really old ones that were broken by previous cocos2d releases they all > work (though all of those games also included their own copy of cocos2d). > > We are at in impasse here. I will state my situation and propose three ways out: My situation: - Tomorrow night I left town for three weeks (work), and the following two weeks will be also busy. So, today is the last day that I can do hard work. - A release is telling 'we wrote some code we think useful, we are familiarized with it and exercised enough to have confidence in it, we think is ready for public use'. I will need some days of going back and fort in the code to decide by myself if its broke something or not, if it gives something valuable or not, or if it forces to deform other subsystems. I don't have those days. Ways out: - Some other committer is willing to tell 'I know how the others init work, and there will be no problem with the news init' . You say: 'all the test runs'. All right , in that scenario I will trust both, pack what it is in trunk and release today. - I realize the first is hard due to time frame. It may be someone is willing to review, but need more time. Fair, but probably nobody wants to wait five weeks, and I won't have any motivation to return to this point in five weeks. So I don't do the release, Cocos nominates another person to make it happen, and hopefully in some days it can be out. Really at this point building package and documentation is easy, basically all in the release chain is updated and tested. We have a release, only that other person releases. - We settle for not introduce the features in discussion. No init for Layer and Scene. I release today. How is the decision taken ? option 3 is Richard decision option 1 is any committer decision option 2 I think its for the Cocos historical leaders, Lucio or Riq. I will be perfectly well with any decision, really. To close, I will be out now , 9:00 local, return home between 16:00 and 17:00, can work from there to max 2:00 I will be offline till 17, and will be no time for so much discussion after. If there is an explicit decision by the time, much better. cheers -- claudio -- You received this message because you are subscribed to the Google Groups "cocos2d discuss" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/cocos-discuss?hl=en.
