Cedric, AFAIU the number of API is short, just leave the old in place and add a new one. Mark the old as deprecated. At least it won't crash due dynamic linker not finding the symbol
On Thu, Mar 13, 2014 at 10:45 PM, Cedric BAIL <cedric.b...@free.fr> wrote: > On Fri, Mar 14, 2014 at 5:26 AM, Davide Andreoli <d...@gurumeditation.it> > wrote: >> 2014-03-13 3:14 GMT+01:00 Cedric BAIL <cedric.b...@free.fr>: >>> There is an open review https://phab.enlightenment.org/D562 that raise >>> a good point. The current API is not correct and don't let Edje_Edit >>> change edje value accurately (lost of precision). Changing the type of >>> the parameter will be an ABI break, but the only application that >>> should use this API should already be impacted by the problem or there >>> is only one application that use it (and that's where the patch come >>> from). >>> >>> I am very close to accept this patch in. If anyone does have any >>> strong reason to say no to this, time to raise your voice. >>> >> >> Python bindings are using edje_edit, thus this patch are going to break the >> bindings >> in some way. I know that api is marked as unstable but we are using it for >> editje. > > Right, it will require a recompile of the python binding. Is it > something we can force on distribution or not (To do a sync update of > EFL + Python binding). > >> I really dunno well how much this will break the bindings but it seems to >> me quite >> crazy to break now. I strongly suggest to just add new api with the correct >> params, >> like: >> edje_edit_rel1(2)_offset_x(y)_int_set() >> and deprecate the wrong one. > > That could be the solution. > >> Edje_edit api is more like an internal api, so having one that is >> not-so-clean seems >> to me better that breaking stuff randomly. >> >> btw, I'm really curious to understand when this problem was introduced, my >> old >> edje_editor was using the full edje_edit api and (I'm quite sure) that >> error was >> not possible at the time... :/ > > It is a rounding error. So you need to want a specific value and have > that one not part of the possible range of float. You may have had it > and not noticed also. > -- > Cedric BAIL > > ------------------------------------------------------------------------------ > Learn Graph Databases - Download FREE O'Reilly Book > "Graph Databases" is the definitive new guide to graph databases and their > applications. Written by three acclaimed leaders in the field, > this first edition is now available. Download your free book today! > http://p.sf.net/sfu/13534_NeoTech > _______________________________________________ > enlightenment-devel mailing list > enlightenment-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/enlightenment-devel -- Gustavo Sverzut Barbieri -------------------------------------- Mobile: +55 (19) 99225-2202 Contact: http://www.gustavobarbieri.com.br/contact ------------------------------------------------------------------------------ Learn Graph Databases - Download FREE O'Reilly Book "Graph Databases" is the definitive new guide to graph databases and their applications. Written by three acclaimed leaders in the field, this first edition is now available. Download your free book today! http://p.sf.net/sfu/13534_NeoTech _______________________________________________ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
