!xiMNLXQQvmUKVgTwCc:matrix.org

VL-Language

29 Members
language design for VL3 Servers

Load older messages


SenderMessageTime
1 Jul 2020
@Aion:matrix.orgDottore * instead of unrolling everything along Y to make it divisible by lines09:37:27
@velcrome:matrix.orgvelcromegregsn asked to make a new [Proposal] for the region-slash-frame-slash-codeblock idea09:38:36
@velcrome:matrix.orgvelcromeI guess our current discussion shows that there is really a need :)09:39:26
@nissidis:matrix.orgc nisidisahh I just commented over this on the issue in VL-Language, I strongly agree with the suspicion that this will force the user (us) to change the way we patching and working mainly on the vertical axis. I was wondering, if layering (like z-index which is used in loops) would be a possible solution , like giving one more pseudo-dimension to our current layout.10:00:18
@nissidis:matrix.orgc nisidisVL-Lang-Layer-01.jpg
Download VL-Lang-Layer-01.jpg
10:10:07
@nissidis:matrix.orgc nisidisOr by defining "Layer" as a property and in extend as execution order 10:10:50
@tebjan:matrix.orgtebjanI've often though about 3d patches... It would offer a wide range of new possibilities. But I think it could also become a bit too much.10:16:41
@termnml:matrix.orgtermnml

I've often though about 3d patches... It would offer a wide range of new possibilities. But I think it could also become a bit too much.

This will be Delta in the future?

10:17:51
@nissidis:matrix.orgc nisidis sure that would be really catchy but it is well known that humans cant observe more than two spatial dimensions simultanesly :P so 3rd dimension could be the light gray you introduced for loops or even a totally different color, what I would like in any case would be to switch between these "layers" or to isolate them as I would do in any image or 3d manipulating app 10:24:09
@nissidis:matrix.orgc nisidisit*10:24:35
@velcrome:matrix.orgvelcromevr patching anyone?10:35:38
@velcrome:matrix.orgvelcromecan't be much different from solving wall connection puzzles in Alyx :D10:36:29
@nissidis:matrix.orgc nisidisRedacted or Malformed Event10:36:42
@nissidis:matrix.orgc nisidisI want to see you trying to create a link between two nodes (boxes) with a straight line 10:37:12
@nissidis:matrix.orgc nisidis:P10:37:23
@velcrome:matrix.orgvelcromeimage.png
Download image.png
10:37:25
@nissidis:matrix.orgc nisidis
In reply to @tebjan:matrix.org
I've often though about 3d patches... It would offer a wide range of new possibilities. But I think it could also become a bit too much.
* sure that would be really catchy but it is well known that humans cant observe more than two spatial dimensions simultanesly :P so 3rd dimension could be the light gray you introduced for loops or even a totally different color, what I would like in any case would be to switch between these "layers" or to isolate them as I would do in any image or 3d manipulating app
10:39:13
@Aion:matrix.orgDottoreexecution order is a timeline. time dimension is involved. we need more dimensions!!11:03:56
@Hayden_A:matrix.orgHayden_AI wonder, with execution order, if we only need to give one one line of execution priority over others. I also wonder if the order could be selected at pin level. If you have a node and you want to set one line of execution as primary, you might select this from the nodes output pin. Also you could choose an input pin on the outside of a node that should be the primary one. edit: spelling and detail12:17:04
@Hayden_A:matrix.orgHayden_A * I wonder, with execution order, if we only need to give one one line of execution priority over others. I also wonder if the order could be selected at pin level. If you have a node and you want to set one line of execution as primary, you might select this from the nodes output pin. Also you could choose an input pin on the outside of a node that should be the primary one. edit: spelling and detail12:18:25
@velcrome:matrix.orgvelcromeI am a little worried that hiding priorities in pins or configs might be a source of endless frustration when debugging12:45:08
@Hayden_A:matrix.orgHayden_Awhat if they had some colour?17:51:15
@Hayden_A:matrix.orgHayden_Abut i understand, you want to know which route the execution is taking17:51:51
@Hayden_A:matrix.orgHayden_Amaybe something like the error system, where the tracer was able to show the links that were executed, but maybe as a mode for looking at the patch17:54:37
7 Jul 2020
@tegan336:matrix.orgtegan336 joined the room.16:35:13
@tegan336:matrix.orgtegan336 left the room.22:25:35
14 Jul 2020
@Hayden_A:matrix.orgHayden_A
Als Antwort auf@Hayden_A:matrix.org
hat ein Bild gesendet.
Regarding this delegate approach : I wouldn't use the CreateDefault system for that. Doing so would set a default for all delegates of that type. But you want something different : you want to initialize the functionality field in this particular patch, which you could do on create. Just write into that pad on create
08:11:42
@Hayden_A:matrix.orgHayden_AFor small interfaces with very few operations to implement this delegate approach can also be quite powerful. 08:12:20
@Hayden_A:matrix.orgHayden_AIn your patch you accept the delegate from outside. This should result in a process node that can be created as region. As a user of that thing you now can link into the region and implement an interface anonymously and this by using closures. Edit: links into the region capture the value. This is what I referred to as a closure. 08:12:54
@Hayden_A:matrix.orgHayden_AIn skia eg many primitives are patched that way. They don't implement the layer themselves. A region called renderingNode or so implements the interface and already comes with a bit of intelligence. The missing part is patched in the region 08:13:17

Show newer messages


Back to Room List