freenode/#clim - IRC Chatlog
Search
9:52:19
jackdaniel
can you suggest a viable, good looking alternative without full theming support for careful color palette?
9:54:16
loke
jackdaniel: I've experimented a bit. Just plain inversion looks better (imho) most of the time. (not perfect though, since it only affects text with default colours, so if you have, say, dark blue text it won't invert correctly)
9:54:59
loke
I have some ideas how to deal with that. I just wnated to know if there was some particular reason why it's the way it is, or if you'r eopen to conaging it
9:55:40
loke
Perhaps it can wait until we have some global configuration system in place, so that theming becomes viable.
9:55:47
jackdaniel
I can provide some historical background: former behavior was to change background, so for text with custom ink you couldn't see anything
9:56:39
loke
jackdaniel: Ah right. That makes sense. It's bacailly what I oberved too. I'm playing around with overriding text colours when drawing the highlight. It's questionable if it's worth the effort though.
9:57:25
jackdaniel
as I envision it for the future reference is to have half-transparent rectangle drawn on the selection
10:25:02
loke
jackdaniel: Another question: I'm building a basic HTML renderer (only handles stuff like bold/italics, etc). I want to use it in the clipboard demo when pasting rich text.
10:31:15
jackdaniel
I don't think it is, but I have no strong arguments against that, only the dependency complexity concern
10:33:23
jackdaniel
nb: we should think about stripping some dependencies putting some extensions out of direct mcclim dependency list (so they can be added separately)
10:46:07
loke
jackdaniel: Fair enough. It's not overly important. I'll just display the html in, well, html form :-)
11:24:33
loke
Can anyone take a look at this form? I sometimes get booleal logic wrong, and I just want to ake sure I read this right./
11:24:51
loke
https://github.com/mcclim/McCLIM/blob/clipboard-experiments/Core/clim-basic/stream-input.lisp#L220
11:25:31
loke
Am I right is asserting that the (and (gadgetp...) (gadget-active-p ...)) part is a noop?
11:26:04
loke
because whenever that one is true, the second half of the OR will also be true, correct?
11:30:01
jackdaniel
it says: if there is an event deliver if it is an active gadget, and if it is not an active gadget, deliver it nonethless unless it is a clim-stream-pane (and one of the events it hogs for its own processing)
11:32:18
jackdaniel
so to turn things around: don't deliver it when it is is clim-stream-pane and event is interesting to it (*unless* said clim-stream-pane is also a gadget, then deliver that)
11:36:00
loke
what I'm saying is that (not (and (typep sheet 'stream-pane) ...)) will always be true if sheet is a gadget
12:34:42
jackdaniel
as I imagine things that *should* be the case, that gadget may present some objects acceptably for instance
12:35:24
jackdaniel
then we could create reusable components which fit well into CLIM model (at least for native implementations)
13:34:27
loke
https://github.com/mcclim/McCLIM/blob/clipboard-experiments/Libraries/Drei/input-editor.lisp#L442
13:41:24
jackdaniel
I can think of something like (if-let ((name …) (format nil …) (vector …)), but I don't know
14:01:51
loke
jackdaniel: I followed it. I did an initial review when he first posted it... I agree. Significantly better.
14:09:58
loke
Open the drawing tests, then select the "bordered" category, and press arrow down about 5 or 6 times
14:43:50
loke
jackdaniel: oh wait. the problem is a redraw problem. I'll show a screenshot where it's even more visible:
14:58:56
loke
Interesting. It also happens in the arrows test, but only when scrolling horizontally:
15:02:44
loke
I can't say if the other redraw issue is related, since I don't know the root cause. It could be...
15:06:49
josrr
loke: "but only when scrolling horizontally", I think is happening vertically also, but, you can not tell because everything is gray
16:17:25
josrr
I think there's something wrong in the calculation of the sizes of the panes: https://www.rufina.link/archivos/d-t-sc-01.png
16:30:56
josrr
this stops the problem while scrolling lists: https://gist.github.com/josrr/5a689434b3654ebaaa05f83783635156
16:41:39
jackdaniel
and as of this problem with scrolling demos, please either you or loke make a ticket on the issue tracker. I need to go now