[PD-dev] bug in double click event in a GUI in a GOP?

[Alexandre Torres Porres]

Oh well, I remembered that ceammc also implemented double click for their
knob, I was inspired by VCV's knob but when studying all knobs in Pd I saw
they also had the same idea. The ceammc external works in Pd inside GOPs,
so there is a way. The problem is that they use what seems to be a
very complicated workaround as they do not use the click function or the
framework/API provided by Pd... ceammc also is mostly C++, I didn't even
bother to understand it as my programming skills are way way limited. I got
the hang of coding in C for Pd externals and that's it, for rather simple
objects I may add. Just now trying to better deal with GUIs and tcl/tk... I
really hope I don't need to add a new layer that doesn't seem trivial at
all to me just so I can have a double click.

The thing is that it's impossible or can't happen, there are externals that
do this already, we just should also offer a builtin and simple solution in
Pd's framework.

I'm also adding this information on the github discussion.

cheers

Em sáb., 8 de abr. de 2023 às 10:22, Alexandre Torres Porres <
porres at gmail.com> escreveu:

> ok, so a bug it is
>
> I see iohanness mentioned that a fix is 'convoluted' and that miller just
> didn't implement this cause he hadn't figured it out.
>
> You can still click and edit number boxes and other thingies in Pd, that
> is type a new number/symbol/list. The thing is that with the recent
> addition of list boxes it was necessary to add this double click feature so
> it was easier to edit a single element in it, and thus this bug came up.
> Until then it seems no other object or external ever needed double clicks
> so it finally bit us.
>
> As I understand, being 'convoluted' means that the fix is not in the
> object (number/symbol/list box) itself, but under the hood deep inside the
> guts of Pd.
>
> Now, it seems no one ever attempted a real fix/implementation yet, huh?
>
> I haven't really checked deep inside so I'm missing the real challenge
> behind this, what makes it tricky to do and all, any details on this?
>
> It is not clear on the github discussion, so I still see that it could be
> fixed/implemented... I can add to the discussion on github that no
> external will work as well, which makes this more of a crucial problem that
> needs to be fixed for list box (mostly) and the new externals I am
> designing (I'm also planning on including 'double click' features in other
> GUIs than my new knob).
>
> cheers
>
> Em sáb., 8 de abr. de 2023 às 06:29, Antoine Rousseau <antoine at metalu.net>
> escreveu:
>
>> Hi Alex,
>>
>> This has already been reported in
>> https://github.com/pure-data/pure-data/issues/1661.
>> It turns out that it won't be easy to fix, even if it should be possible.
>>
>>
>>
>> Le sam. 8 avr. 2023 à 00:18, Alexandre Torres Porres <porres at gmail.com>
>> a écrit :
>>
>>> Hi, I am designing a knob GUI and it has a feature for double click
>>> where it goes to the mid position (or some other setting) when you do it.
>>> This is quite cool and I'm happy with it.
>>>
>>> Then I was testing it inside an abstraction, as I wanna use this to
>>> design modular inspired abstractions and to my frustration it doesn't work.
>>> Somehow the 'double click' event that we get from 'w_clickfn' in
>>> 't_widgetbehavior'...
>>>
>>> In Vanilla, double click in a GUI (when in run mode) is only possible
>>> for gui boxes (atom/symbol/list), which lets us select the contents of the
>>> box. I haven't looked at its code, but I can see it is also not possible to
>>> double click it in a GOP, so it might be the same thing.
>>>
>>> Well, is this intentional? Is this a bug? Can I ask to change this?
>>>
>>> I would really like to be able to write GUIs that make use of double
>>> clicking, not being able to do so seems like a very weird and arbitrary
>>> restriction.
>>>
>>> Thanks
>>> Cheers
>>> _______________________________________________
>>> Pd-dev mailing list
>>> Pd-dev at lists.iem.at
>>> https://lists.puredata.info/listinfo/pd-dev
>>>
>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.puredata.info/pipermail/pd-dev/attachments/20230408/ea304d71/attachment.htm>