freenode/#clim - IRC Chatlog
Search
4:31:11
loke
beach: Thanks for your points yesterday. I'm now looking at implementing fast scrolling in a different way, that doesn't change the underlying transformation
4:32:10
loke
beach: In my thinking I boiled down your opinion to one thing, please let me know if I'm on the right track:
4:33:06
loke
beach: The number 1 issue seems to be that if I set transformation on a sheet to X, and then read that same transformation back, then you should get X back. Not, as in my case, kinda-sorta-X-but-rounded-a-bit
4:34:16
loke
The idea being that scrolling operations attempt to would ensure that the subpixel offset remains constant.
4:34:36
loke
That will not eliminate rounding errors, but will ensure that they are always resulting in the same error.
4:36:37
loke
beach: I came to that conclusion after thinking about cases where the forced rounding could cause issues, and came up with only one case: That case being an animation that slowing scrolled the screen by adding a small value to the transformation every frame: Like adding 0.4 pixels per fracme. With forced rounding, that would prevent movement since the 0.4 would be rounded to 0 every time it was updated.
4:37:28
loke
(I don't think this is something that would happen in practice, but it's nice to expore possibilities that would result in a theoretically “correct” behaviour)
4:38:15
loke
Anyway, I spent a day thinking about it, and I just now started to actually experiment with code, so it'll take another day or so before I have anyting to report.
4:38:56
loke
slyrus1: didn't have any time in front of Emacs yesterday so I didn't look at your code yet.
4:41:01
loke
beach: Oh right, I almost forgot: Tell me: Can Second Climacs handle non-text objects in its buffers?
4:42:38
beach
Yes, it can. I but I don't have any ways of inserting them, and I don't know how to display them (yet). :)
4:43:56
loke
beach: What I want to do is to build a “notebook” that can be used to draft scientific papers. When experimenting in Maxima, you'd just click on an equation to copy it into the notebook (which would be a (second climacs?) editor buffer).
4:44:30
loke
Id' want those to be displayed in maths form of course. Then you can export the notebook as LaTeX.
4:46:59
loke
beach: Ideally, I'd like a way to click on the visual form of the equation, which then gets replaced with the maxima-text-form (the kind you type, like “1/x^2”). And then when you're done, it filps back and renders it in mathematical form again.
4:48:42
beach
I haven't thought about it much. My initial thinking would be to include the entire formula as a single item in the buffer as an instance of some appropriate class. Inserting it as text would be trickier, because you would then have to somehow identify text snippets that should be rendered differently.
4:49:51
loke
beach: Correct. The text form would only exist while you edit the text. The moment you move out of the formula it rrverts back. So I presume at all times it would still be stored as a formula (implemented by the class ‘maxima-native-expr’ in CLimaxima at the moment)
4:50:24
loke
The sextual form of the formula is only there since it makes editing it maually easier (or shall I say, ‘possible’)
4:51:34
loke
I think the equation-editing-text should be separately highlighted (or even displayed in a popup of some kind) to make it obvious it's not part of the actual text per se.