Hey Miller et al.
Some of my data structures based patches exhibit issues, probably since 1b928e6. I haven't yet the opportunity to deeply investigate and I'll try to open issues proper soon. Since these days it sometimes takes a few days for me to sit down and look at things, I thought it might be better to communicate early that some things might still need fixing.
Some patches don't report mouse clicks while they do so in 0.55-2. Another issue is that colors of [drawnumber] elements are black and not in the color defined. Both those findings don't apply to all my patches, thus making debugging a bit more complex.
I have often two or more layers of elements above each other so that the lowest layer is optimized for mouse interaction and the top layer for appearance. Maybe ordering of layers plays a role here?
I will report with more details when I can provide them.
Thanks, Roman
On Sun, 2025-01-12 at 21:19 +0100, Roman Haefeli wrote:
Both those findings don't apply to all my patches, thus making debugging a bit more complex.
That's not true. They are consistent, thus easy to report. I was fooled by some patches using iemguis for click detection.
Roman
I'm happy at least that it's consistent :)
I fixed something that I _think_ was a bug - previously, when I clicked an array that had visible elements , every single element of the array reported the click. I changed the behavior to what I think was originally intended - only the element that is clicked on sends the "click" message to its "struct" object.
for instance in the patch here: msp.ucsd.edu:tmp/z3.pd
when I drag up or down on the hot-points of the small rectangles I get notifications from both that one element and from the array as a whole.
OTOH, if I click in the interior of the rectangle I get no action, since it's not on a hot point. (Perhaps there should be a way to catch that as well? That's a different functionality from a hot-point-click as reported in the "click" message.)
cheers Miller
On 1/12/25 10:33 PM, Roman Haefeli wrote:
On Sun, 2025-01-12 at 21:19 +0100, Roman Haefeli wrote:
Both those findings don't apply to all my patches, thus making debugging a bit more complex.
That's not true. They are consistent, thus easy to report. I was fooled by some patches using iemguis for click detection.
Roman
pd-dev@lists.iem.at - the Pd developers' mailinglist https://lists.iem.at/hyperkitty/list/pd-dev@lists.iem.at/message/HZMKJRJ33DK...
I just updated to the last version of Pd (64b windows) and tried it out, and funilly got different results:
- when I click on the array area, i get a report on all array elements. but when I click on an element, only that element is reported.
- when the hot-points of each elements are dragged, only the message from toplevel is sent. also when sliding through other elements while click-drag, no other messages are sent, should something come out when a new element is active?
- when in edit mode, I don't see the array bounds anymore. also it's not possible to select or move the array using the mouse, but it works with select all and keyboard keys.
Best, Joao
I'm happy at least that it's consistent :)
I fixed something that I _think_ was a bug - previously, when I clicked an array that had visible elements , every single element of the array reported the click. I changed the behavior to what I think was originally intended - only the element that is clicked on sends the "click" message to its "struct" object.
for instance in the patch here: msp.ucsd.edu:tmp/z3.pd
when I drag up or down on the hot-points of the small rectangles I get notifications from both that one element and from the array as a whole.
OTOH, if I click in the interior of the rectangle I get no action, since it's not on a hot point. (Perhaps there should be a way to catch that as well? That's a different functionality from a hot-point-click as reported in the "click" message.)
cheers Miller
On 1/12/25 10:33 PM, Roman Haefeli wrote:
On Sun, 2025-01-12 at 21:19 +0100, Roman Haefeli wrote:
Both those findings don't apply to all my patches, thus making debugging a bit more complex.
That's not true. They are consistent, thus easy to report. I was fooled by some patches using iemguis for click detection.
Roman
--- pd-dev@lists.iem.at - the Pd developers' mailinglist https://lists.iem.at/hyperkitty/list/pd-dev@lists.iem.at/message/HZMKJRJ33DK...
This is odd - it doesn't sound like either the behavior of the last release, nor of the latest continuous-integration compiled output, as far as I know, but of something in between that I might have uploaded before it was in its present state.
I'm getting notifications about clicking and changed-values, either for one element or for the whole scalar. Hopefully this is true on all platforms :)
Meanwhile there are various things that are unfortunate or wrong about the latest version - I'm going to work on teh "easy" stuff (wrong colors or font sizes) and then try to work back up to the main job.
cheers Miller
On 1/16/25 12:27 AM, João Pais wrote:
I just updated to the last version of Pd (64b windows) and tried it out, and funilly got different results:
- when I click on the array area, i get a report on all array
elements. but when I click on an element, only that element is reported.
- when the hot-points of each elements are dragged, only the message
from toplevel is sent. also when sliding through other elements while click-drag, no other messages are sent, should something come out when a new element is active?
- when in edit mode, I don't see the array bounds anymore. also it's
not possible to select or move the array using the mouse, but it works with select all and keyboard keys.
Best, Joao
I'm happy at least that it's consistent :)
I fixed something that I _think_ was a bug - previously, when I clicked an array that had visible elements , every single element of the array reported the click. I changed the behavior to what I think was originally intended - only the element that is clicked on sends the "click" message to its "struct" object.
for instance in the patch here: msp.ucsd.edu:tmp/z3.pd
when I drag up or down on the hot-points of the small rectangles I get notifications from both that one element and from the array as a whole.
OTOH, if I click in the interior of the rectangle I get no action, since it's not on a hot point. (Perhaps there should be a way to catch that as well? That's a different functionality from a hot-point-click as reported in the "click" message.)
cheers Miller
On 1/12/25 10:33 PM, Roman Haefeli wrote:
On Sun, 2025-01-12 at 21:19 +0100, Roman Haefeli wrote:
Both those findings don't apply to all my patches, thus making debugging a bit more complex.
That's not true. They are consistent, thus easy to report. I was fooled by some patches using iemguis for click detection.
Roman
--- pd-dev@lists.iem.at - the Pd developers' mailinglist https://urldefense.com/v3/__https://lists.iem.at/hyperkitty/list/pd-dev@list...
pd-dev@lists.iem.at - the Pd developers' mailinglist https://urldefense.com/v3/__https://lists.iem.at/hyperkitty/list/pd-dev@list...
On Wed, 2025-01-15 at 14:48 +0100, Miller Puckette via Pd-dev wrote:
I fixed something that I _think_ was a bug - previously, when I clicked an array that had visible elements , every single element of the array reported the click. I changed the behavior to what I think was originally intended - only the element that is clicked on sends the "click" message to its "struct" object.
OK. If I understand correctly, this is unrelated to my bug report since it is about data structure arrays. Up until 0.55-2, [filledpolygon] scalars do send a 'click' message to their [struct] object. The 'click' is detected anywhere inside the surrounding rectangle of the scalar, not just in the corners. However, in the current master clicking on a [filledpolygon] scalar doesn't emit a 'click' message anymore.
I reported this here: https://github.com/pure-data/pure-data/issues/2514
Roman