VisPy User Survey

[dho...@gmail.com]

Hi everyone,

The VisPy project wants to know more about our users. We've put together a little survey: https://forms.gle/2BrqGZN3e9EWqPvG6

The results of this survey will have a major impact on the future of VisPy so please take a couple minutes to help us out.

Thanks,

Dave

[dho...@gmail.com]

Hi everyone,

We've started reviewing the survey results and realized we'd like some more information for some of the answers. So far there have been 5 people who answered the question "How much do you rely on OpenGL compared to more recent GPU technologies such as Vulkan or Metal?" with "I specifically need OpenGL", but none of them left any contact information or described how they use VisPy in their own applications. If you were one of these people and are comfortable doing so, could you please reach out to me (publicly or privately) and provide some more details on your use case? Thanks everyone.

Dave

[jim.v...@gmail.com]

I can't remember how I answered the question. I have one application for which I wrote an OpenGL shader (for height-based color-coding of a 3D mesh rendering of a real-time camera image consisting of bright spots on a dark background). I couldn't figure out a performant way to do it otherwise. I have no particular fondness for OpenGL (to put it mildly), I just don't know anything about Vulkan or how it would integrate with VisPy.

[David Hoese]

Hi Jim,

> I just don't know anything about Vulkan or how it would integrate with
> VisPy.

No problem. Neither do we. :)

Ok, that's not true, we have ideas and some of us have more Vulkan
experience than others. Some of the responses we got made me realize we
should have worded the question differently. Vulkan still uses shaders,
but uses a newer version of the GLSL language than what most of VisPy's
existing shaders are in. If/when VisPy adopts a Vulkan "graphics
backend" as an option, it would likely be available as a separate set of
Visuals that use Vulkan (datoviz) instead of "gloo". Gloo would still be
an option, but any use of the other backends would likely require some
minimal rewriting of custom shaders.

The main goal of the question I mentioned was to find out if there is
some use case or reason that we are missing for why people might *need*
OpenGL over some of the newer technologies/libraries. For example, a
hardware limitation or something that is possible in OpenGL that somehow
isn't possible in a newer technology. Of course, in hindsight, we should
have just allowed for a short answer for an explanation.

Thanks for responding.

Dave

> --
> You received this message because you are subscribed to a topic in the
> Google Groups "vispy" group.
> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/vispy/kYn93OGq_UM/unsubscribe
> <https://groups.google.com/d/topic/vispy/kYn93OGq_UM/unsubscribe>.
> To unsubscribe from this group and all its topics, send an email to
> vispy+un...@googlegroups.com
> <mailto:vispy+un...@googlegroups.com>.
> To view this discussion on the web, visit
> https://groups.google.com/d/msgid/vispy/eb591ded-0836-4089-8d26-1604ed42d932n%40googlegroups.com
> <https://groups.google.com/d/msgid/vispy/eb591ded-0836-4089-8d26-1604ed42d932n%40googlegroups.com?utm_medium=email&utm_source=footer>.

[Cyrille Rossant]

In principle it should not take more than a couple of minutes to port an OpenGL ES shader (VisPy) to a recent version of GLSL (4.50 for example) that is supported by Vulkan. I even wonder if there isn't already a GLSL->SPIR-V compiler that supports OpenGL ES GLSL, in which case the existing shaders could be reused as-is.

Cyrille

You received this message because you are subscribed to the Google Groups "vispy" group.
To unsubscribe from this group and stop receiving emails from it, send an email to vispy+un...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/vispy/5a4ff3e3-0d39-0459-f2bf-ea3cf0fbf23b%40gmail.com.