Re: [Pharo-dev] [Bloc] Hierarchy of BiEvent

2016-11-06 Thread stepharo
Le 27/10/16 à 10:57, Guille Polito a écrit : Something itches me from this discussion. I somehow find myself with Glenn: - should Bloc work with other than SDL/OSWindow? - should SDL/OSWindow work with other than Bloc? Merging the event hierarchy may prevent these two scenarios. And these

Re: [Pharo-dev] [Bloc] Hierarchy of BiEvent

2016-10-27 Thread Guille Polito
Something itches me from this discussion. I somehow find myself with Glenn: - should Bloc work with other than SDL/OSWindow? - should SDL/OSWindow work with other than Bloc? Merging the event hierarchy may prevent these two scenarios. And these scenarios are not only really possible: this is

Re: [Pharo-dev] [Bloc] Hierarchy of BiEvent

2016-10-26 Thread stepharo
Hi Glenn Now I understand. So can you copy and paste this mail into the class comment of the BlEvent root. We should really document such design decision. Stef Le 26/10/16 à 19:25, Glenn Cavarlé a écrit : Hi Stef, stepharo wrote OSWindow cannot be the only one backend for Bloc so Bloc

Re: [Pharo-dev] [Bloc] Hierarchy of BiEvent

2016-10-26 Thread stepharo
Le 26/10/16 à 20:40, Denis Kudriashov a écrit : 2016-10-26 19:45 GMT+02:00 Glenn Cavarlé >: Yes, your are right. What i mean is that it could be interesting if: [ SDLEvent => BlEvent ] is substantially faster than [

Re: [Pharo-dev] [Bloc] Hierarchy of BiEvent

2016-10-26 Thread Denis Kudriashov
2016-10-26 19:45 GMT+02:00 Glenn Cavarlé : > > Yes, your are right. > What i mean is that it could be interesting if: > [ SDLEvent => BlEvent ] is substantially faster than [ SDLEvent => > OSEvent => BlEvent ]. > If the difference is not significant, i think that does

Re: [Pharo-dev] [Bloc] Hierarchy of BiEvent

2016-10-26 Thread Glenn Cavarlé
Denis Kudriashov wrote > 2016-10-26 18:21 GMT+02:00 Denis Kudriashov > dionisiydk@ > : > >> No. You are not right here. Conversion code for OSEvent=>BlEvent and >> SDLEvent=>OSEvent is almost same. All events have similar interface to >> retrieve event data. > > > I mean conversion code in

Re: [Pharo-dev] [Bloc] Hierarchy of BiEvent

2016-10-26 Thread Glenn Cavarlé
Hi Stef, stepharo wrote >> OSWindow cannot be the only one backend for Bloc so Bloc events cannot be >> based on OSEvents. > Why that? Because it is OS"Window" and many devices don't have window and sometime don't have SDL too. If tomorrow, OSWindow become something like OSDirectMedia with

Re: [Pharo-dev] [Bloc] Hierarchy of BiEvent

2016-10-26 Thread Denis Kudriashov
2016-10-26 18:21 GMT+02:00 Denis Kudriashov : > No. You are not right here. Conversion code for OSEvent=>BlEvent and > SDLEvent=>OSEvent is almost same. All events have similar interface to > retrieve event data. I mean conversion code in both cases is just stupid creation

Re: [Pharo-dev] [Bloc] Hierarchy of BiEvent

2016-10-26 Thread Denis Kudriashov
2016-10-26 16:58 GMT+02:00 Glenn Cavarlé : > > It is interesting only if performances are substantially improved. > Because it is also duplication, you duplicates in Bloc the OSWindow > conversion code without any added value. > No. You are not right here. Conversion

Re: [Pharo-dev] [Bloc] Hierarchy of BiEvent

2016-10-26 Thread Denis Kudriashov
2016-10-26 17:20 GMT+02:00 stepharo : > So then it means that OSWindowEvents are not useful. > Glenn said that OSEvent's has no specific state which needed for Bloc logic. And Bloc events define it: - OSEvents are too low level to be used directly in Bloc (nothing about >

Re: [Pharo-dev] [Bloc] Hierarchy of BiEvent

2016-10-26 Thread stepharo
Le 26/10/16 à 16:58, Glenn Cavarlé a écrit : Denis Kudriashov wrote I can't take it. Duplication is always bad. There is no reason for it. I agree, is always bad if there is no real reason to duplicate but i think there is a real argument for Bloc. OSWindow cannot be the only one backend for

Re: [Pharo-dev] [Bloc] Hierarchy of BiEvent

2016-10-26 Thread Glenn Cavarlé
Denis Kudriashov wrote > I can't take it. Duplication is always bad. There is no reason for it. I agree, is always bad if there is no real reason to duplicate but i think there is a real argument for Bloc. OSWindow cannot be the only one backend for Bloc so Bloc events cannot be based on

Re: [Pharo-dev] [Bloc] Hierarchy of BiEvent

2016-10-26 Thread stepharo
Le 26/10/16 à 15:35, Denis Kudriashov a écrit : Hi Glenn. Thank's for answer. 2016-10-26 13:29 GMT+02:00 Glenn Cavarlé >: More generally, i think that, to stay consistent, Bloc should not reuse directly (and everywhere)

Re: [Pharo-dev] [Bloc] Hierarchy of BiEvent

2016-10-26 Thread stepharo
More generally, i think that, to stay consistent, Bloc should not reuse directly (and everywhere) some part of underlying libraries even if sometime these parts seems to match directly. I do not understand why duplication is a good approach. I would prefer to enhance OSWindowEvents because

Re: [Pharo-dev] [Bloc] Hierarchy of BiEvent

2016-10-26 Thread Denis Kudriashov
Hi Glenn. Thank's for answer. 2016-10-26 13:29 GMT+02:00 Glenn Cavarlé : > More generally, i think that, to stay consistent, Bloc should not reuse > directly (and everywhere) some part of underlying libraries even if > sometime > these parts seems to match directly. > I

Re: [Pharo-dev] [Bloc] Hierarchy of BiEvent

2016-10-26 Thread Glenn Cavarlé
Hi Denis and thanks for looking at this. Yes there is a duplication of the event hierarchy between Bloc and OSWindow. We discussed about the fusion of OSWindow and Bloc events one year ago at INRIA. We made the choice to keep Bloc independent from OSWindow because, among other things: - Bloc

Re: [Pharo-dev] [Bloc] Hierarchy of BiEvent

2016-10-26 Thread Tudor Girba
Hi Denis, Thanks a lot for looking at this! I will let Glenn answer. Cheers, Doru > On Oct 26, 2016, at 11:02 AM, Denis Kudriashov wrote: > > Hi. > > I look at dev version of Bloc and have questions about events. > > There is duplicated hierarchy of OSEvent and