Graphing a HAL configuration

I have always been fascinated by graphs and maps

Here's another tool for Hal visualization
Its a graph editor,  it is very preliminary



Its called Nodz and is at


github.com/LeGoffLoic/Nodz/tree/master

I modified it to:
pan zoom scroll and aded net names to the beziers
I changed some transparencies to allow the nets to be visible thru the comps
( rather than code the beziers to avoid comps )
I hard coded a few comps into the demo
I enabled save and load

There is a lot to be done to make this as useful as RockHopper and HalPyGraph
but it looks promising

I am not versed in Python, so the aliasing and choosing the
correct flavours of QtGui QtCore abd QtWidgets was hard-won.

The next challenge for me is to make unique names for each 'plug' (net names)
( another dict to learn, a plug is a source pin sorta )

So, many simple hacks have been done but the
parse hal, output editable graph, save graph and new hal
is a long way off.

I invite any wanting to, to get my source. ( its just 4 text files so far)

thanks
Tomp 06nov2021 20:16 Thailand

Sweet!
Thank you.

interesting - it could be embedded into one on the qtvcp screens.
I would suggest setting it up similar to halshow where you can pick a component for a list then it will visualize the connections related to that component.

Showing the whole configuration in one visualisation seems like it would be awkward.

cmorley wrote:
interesting - it could be embedded into one on the qtvcp screens.
   I have not thought about embedding it, possible though.
   It's meant to help explain  text based netlists

I would suggest setting it up similar to halshow where you can pick a component for a list then it will visualize the connections related to that component.
  Thats possible, a palette could show comps,
  but again the intent is to show hal relationships rather than indvidual comp pin/params
  But users can do as they wish ( i did similar in geda long ago )

Showing the whole configuration in one visualisation seems like it would be awkward.
  Agreed, a basic problem in visualization is seeing the part you are interested in  and how that relatess to the whole.
   I think thats where the user can re-organize/subdivide logical groups  that fit on a 'page' ( old Forth concept )
   yet still scroll to the sorurces or destinations of that logical group. And to zoom out to see how the grouping
   relates to the whole.  I think it is an 'explainer'.

  There's a great video by Lyft Architects about how complex diagrams can be viewd /edited
  Its on Vimeo, but you dont really have to join to see it
     vimeo.com/72447811
  thats the bleeding edge bladerunner extreme tomp
 

I might not have been clear with my thought.
I meant that it would show only the pins, connections and components related to the selected component.
Of course that could still be a lot of information..

hmm, like some kind of markup in the .hal?
could be, but i am a long way from that.
Thx for the idea about limiting the scene/graph
tomp