On Fri, Oct 2, 2009 at 18:30, Paul Giannaros <[email protected]> wrote: > > On 2 Oct 2009, at 23:53, Robert Kern <[email protected]> wrote: > >> On Fri, Oct 2, 2009 at 17:41, Hugo Parente Lima <[email protected]> >> wrote: >>> >>> On Friday 02 October 2009 19:23:52 Robert Kern wrote: >>>> >>>> On Fri, Oct 2, 2009 at 08:45, Matti Airas <[email protected]> >>>> wrote: >>>>> >>>>> With all of this in mind, I believe it would be now a perfect time to >>>>> begin planning any backwards-incompatible changes to PySide, as the >>>>> Python 3 transition would be the perfect instant to incorporate them. >>>>> So, >>>>> start discussing those PSEP ideas! ;-) >>>> >>>> The Python team *very* much discourages libraries from using the >>>> Python 3 transition in order to introduce backwards-incompatible >>>> changes. If all of our libraries also change out from under us, it >>>> makes transitioning to Python 3 too difficult. It will already be >>>> fairly challenging to port from PyQt to PySide and Python 2 to Python >>>> 3 separately. Please do not make it more difficult. >>>> >>> >>> IMO is better to add backwards-incompatible changes when the user already >>> have >>> to worry about some other backwards-incompatible changes than when the >>> user >>> don't need to worry about that (2.x series). So you can push all desired >>> changes at once, because will be impossible to do this later. >> >> That assumes that everyone will still make the change. If you provide >> enough of a barrier, then people simply won't migrate to the >> incompatible versions at all. > > I see your concern but I think this accounts for a tiny minority if people > if any. > >> >>> The Python developers just want to see everyone moving on to Python3, so >>> they >>> want to remove any barrier, but if they can put some sand-bags in the >>> barrier, >>> why negate the extension developers to do so? >> >> All you will do is encourage people to stay behind on Python 2.x. Now >> you will have two incompatible branches of PySide to maintain with two >> sets of documentation and a fractured community. No one benefits from >> this. > > I disagree that these actions only encourage people to stay with v2. When > one is in a porting or 'experimenting' mood it makes sense to go for the > whole new package in one fell swoop.
I don't think you are understanding my point. If you keep adding these incompatibilities on top of each other, you will kill the "experimenting" mood. In any case, this certainly does not accord with my experiences when porting. Reducing the number of changing components is always a primary goal for me when porting. If I can handle one component at a time and still keep the whole system working, it takes less work and I have much more assurance that my modifications are working. If I have to do two components at the same time, I don't have a working system until I finish both tasks. In the meantime, I am chasing down bugs from two different sources and not knowing which one is causing the problem. You can easilly end up rewriting the application rather than porting it. I won't try to convince you that every developer thinks this way; obviously, you don't. But please don't claim that developers like me don't exist. > Plus I believe that an update to the > frankly dated API should be welcomed at all cost. Perhaps the availability > of a more Pythonic API will encourage people to adopt PySide or move to > Python 3; it would certainly seem enticing to me. > > PySide needs to attract new users and does not have a vast swathe of legacy > apps depending on it to worry too much about breaking the API. If making those changes is really so important, make them for the Python 2.x version, too. Dividing your community into two incompatible parts is not a good way to grow that community. Just remember that a good portion of your new users will be PyQt users that do have legacy apps that they want to port. Many of them will also be stuck with Python 2.x for other reasons. -- Robert Kern "I have come to believe that the whole world is an enigma, a harmless enigma that is made terrible by our own mad attempt to interpret it as though it had an underlying truth." -- Umberto Eco _______________________________________________ PySide mailing list [email protected] http://lists.openbossa.org/listinfo/pyside
