libera/#shirakumo - IRC Chatlog

14:12:36 larsen could someone double check a thing for me? Does the screenshot functionality in trial-examples work for you?

14:18:24 Colleen <Shinmera> it did last time I tried, which was many months ago :)

14:20:55 larsen I just updated the shirakumo distribution (2024.09.30-07.37.54), and it also didn't work for me with the version I had before (which I think it was the previous, because I update-all-dists frequently)

14:22:06 Colleen <Shinmera> how does it "not work"?

14:23:48 larsen https://gist.github.com/larsen/aaccdd94418c3c52802542f297c45638

14:24:26 larsen but there must be more, wait a minute

14:25:02 Colleen <Shinmera> :thunk:

14:26:40 larsen holy shit, last failure left a mess of a situation https://gist.github.com/larsen/c466800db1468d9ddcc420b35d0f518f

14:27:41 Colleen <Shinmera> yeah that can happen

14:28:30 larsen also, at the moment it became very slow to launch a new SBCL + Swank and to connect. Not sure what's going on

14:29:58 Colleen <Shinmera> just X things

14:30:09 larsen sorry. ignore me for now. when the sharks will be gone I will try again taking a screenshot in trial-examples :)

14:30:46 Colleen <Shinmera> thanks for alerting me to the issue

14:31:07 Colleen <Shinmera> I'm busy editing my own skull is all

14:31:22 larsen ouch

14:31:49 Colleen <Shinmera> It's very painful indeed, blender keeps lagging

16:15:17 tumdum Is the copyright situation clear for such medical scans?

16:16:11 Colleen <Shinmera> clear in what way

16:16:51 Colleen <Shinmera> (and why do you care)

16:17:48 tumdum I'm just wondering. Maybe medical facilities that perform those own rights?

16:18:25 Colleen <Shinmera> there's a greater chance of my entire body quantum tunnelling across the earth than a hospital prosecuting me for using a mesh I derived myself from a scan of my own skull

16:18:53 tumdum I do have quite a few of those myself :D

16:19:44 tumdum oh, for sure it's unlikely. I'm just wondering what is theoretically possible

16:20:12 Colleen <Shinmera> it's possible to sue anyone over anything. programmers love to get hung up on licenses and other texts as if they were somehow absolute

16:20:35 Colleen <Shinmera> the point is it's so unlikely that it's silly to wonder about it

16:22:40 Colleen <Shinmera> anyway, I wasted an entire day trying to simplify this scan into something usable with blender and failed

16:22:50 Colleen <Shinmera> dunno if I'll try again tomorrow

16:26:41 tumdum This sounds like it could be useful: https://github.com/ranahanocka/Point2Mesh/

16:26:56 Colleen <Shinmera> I already have a mesh. It's the cleanup that's the issue

16:28:07 Colleen <Shinmera> And if there's one thing I have even less patience for than blender it's figuring out some weirdo python NL thingamajig

16:28:13 Colleen <Shinmera> *ML

16:31:31 Colleen <selwyn> why are you doing this

16:31:47 Colleen <Shinmera> so I can have my own skull bounce around in my shitty physics engine

16:32:00 Colleen <Shinmera> as a metaphor for my own punishment and hubris

16:32:30 Colleen <selwyn> i hope the skull has fun

17:37:23 zyd thats a pretty good reason

17:37:51 Colleen <Shinmera> my actual reason is "I got the scan so why the fuck not"

17:49:35 scymtym i'm most astonished that neither technical nor legal obstacles prevented you from getting access to this data

17:49:52 Colleen <Shinmera> I just asked em

17:49:59 Colleen <Shinmera> and they were like "sure why not"

17:50:18 Colleen <Shinmera> and then there's an open source tool to extract a mesh from the scan data

17:54:26 scymtym that's pretty cool. maybe my preconceptions of medical IT are just wrong. one of my recent interviews was also pretty shocking in that regard

17:54:38 Colleen <Shinmera> heh

17:55:39 Colleen <Shinmera> They also did a scan of the surface of my skin and of my teeth and gums

17:55:43 Colleen <Shinmera> but I don't really want that data

19:36:30 larsen is there a UI component, in Alloy, that visualizes an image? (context: I am using a UI to show some information (player's score, for example) on top of the rendered view. I also want to represent player's lives with the usual miniatures of the spaceship. At the moment I am using shader-entities placed by hand next to the score, but this is wrong and works just because the camera does not move yet)

19:36:45 Colleen <Shinmera> larsen: fixed the screenshot issue.

19:37:07 larsen Shinmera: great, thanks!

19:37:53 Colleen <Shinmera> Kandria's UI has a couple of image parts, hold on

19:39:28 Colleen <Shinmera> https://github.com/Shirakumo/kandria/blob/master/ui/credits.lisp#L32-L38

19:39:41 larsen thanks, I will study that

19:39:55 Colleen <Shinmera> the alloy:icon component isn't a direct-value-component for some reason, even though it probably should be

19:40:30 Colleen <Shinmera> anyway, the actual value needs to be a simple:image, which you can get via simple:request-image

19:41:21 Colleen <Shinmera> I... *think* it can also be a direct trial:texture?

19:41:31 Colleen <Shinmera> Don't remember, but pretty sure

19:42:01 larsen I am still wrapping my head around realizations, updates, and to be honest the entire UI system. I will try a trial:texture too (which would be very convenient)

19:42:31 Colleen <Shinmera> yeah it's a bit involved.

19:42:40 larsen as usual, thanks for the pointers

19:44:18 Colleen <Shinmera> the advantage of the realization/data stuff is that alloy can be "automatically" informed when the UI needs to update its display

19:45:05 Colleen <Shinmera> eg in kandria the player is an alloy:observable-object, and the HUD healthbar just does (alloy:realize (slot-value player 'health) ...)

19:45:21 Colleen <Shinmera> it then notices on its own when it needs to change

19:47:57 larsen that part is very cool. [I have ... (alloy:represent (score player) ...) but I think we're talking about the same mechanism]

19:48:20 Colleen <Shinmera> sure

19:49:01 Colleen <Shinmera> I guess these days it's not as cool what with JS frameworks doing all that and getting it "for free" in immediate toolkits that are all the rage

19:50:34 larsen the specific problems I am having at the moment are: just understanding what I am doing when I work simultaneously on define-update's and define-realization's to change cosmetic aspects of the UI. Plus, for some reason I have a riotous panel that disappears after going back and forth two scenes

19:51:17 larsen but, I suppose I'll figure that out, or I'll reach a point where I can at least ask a coherent question

19:51:31 Colleen <Shinmera> define-realization only applies when the component is first constructed into shapes. after that it's all define-update

19:51:59 Colleen <Shinmera> so define-realization says what all the shapes are that make up the visual representation of a component

19:52:28 Colleen <Shinmera> and define-update says what to change the shapes' properties to when it needs to ... update, for whatever reason (value change, size change)

19:53:30 larsen I see

19:55:04 Colleen <Shinmera> usually frameworks kinda drop the "retained" pretense at the drawing stage and have a "painter" object that just rasterises stuff

19:55:31 Colleen <Shinmera> I wanted to have a more declarative way of defining the visuals and also keep opportunities for caching

19:56:00 Colleen <Shinmera> Though the latter hasn't really come true at all and the former could also just expand to an immediate drawing method

19:56:13 Colleen <Shinmera> so in some sense it may be a failure of an idea

19:56:19 Colleen <Shinmera> (there's a few of those in alloy)