Cannot run the tests about jdotnet

[HyperSphereStudio]

There are issues with the released version 3.0.

I am currently fairly busy with college but I will do my best to patch them. Thanks for the log, I will check this out when I have time.

I am currently trying to build an experimental julia runtime based on .NET.
This project would provide full native interopability with .NET and julia and would deprecate JuliadotNet. I will continue to maintain JuliaDotNet for now but keep in mind I plan to deprecate this once I have a working julia runtime .NET prototype.

[reyntjesr]

Thank for the quick reply.
I'm really looking forward to the new project.

[damian-666]

if i can actually run code like this https://github.com/weymouth/WaterLily.jl

i would probably use it.. and see about doing the viusalization using a game engine like monogame or mabye Silk.. . his other version is in protobuld.. which is easy to port C# but doesn't not include any of his performance stuff... even if its a many months time, im just curious how likely that is.. i haven't evaluated the code yet, ran pretty fast before but on net 6 might be blazing and with his newest code.

[BambOoxX]

if i can actually run code like this https://github.com/weymouth/WaterLily.jl

i would probably use it.. and see about doing the viusalization using a game engine like monogame or mabye Silk.. . his other version is in protobuld.. which is easy to port C# but doesn't not include any of his performance stuff... even if its a many months time, im just curious how likely that is.. i haven't evaluated the code yet, ran pretty fast before but on net 6 might be blazing and with his newest code.

If you only intend to plot 2D stuff, try ScottPlot

[damian-666]

thanks, Scott Plot looks ok, but im needing to display a visual flow, like an advected density field to show a fluid represented as an 2d image, at real time interactive rates, many of his Weymouth simulations are real time interactive. so i need it to be deployable on mobile and pc and that one Scott Plot looks to high level and wpf based.. https://github.com/weymouth/WaterLily.jl/blob/master/examples/julia.gif i hope not TLDR but I can't overstate how important i thnk this project is. So, Weymouth is a leading researcher solving problems for the US Navy ( you won't find on open source)... hes movng to julia because he has fortran MPI code, tensor flow on cloud in C and python scripts, to implement for actually practical and cheap Machine learning to help solve nonlinear ODEs faster in real time , approximately linearizing very chaotic and complex phenomena such as the wakes of boats, visible to satellites, in almost real time as opposed to weeks, This is for opimizing swimming robots , foils and understanding fishes , to swim wiht least effort while generating turbulence in 3d which has never been accurately modelled. AS for and me, im merely making a 2d game with surfers and fishes and breaking waves so that kids can play games, encourage sports, but learn how thngs really move not like Mario Bros. so i i aim to run the physics at up to 1000 fps if i can load all the cores and hopefully SIMD lanes, on background threads, then I draw the last good frame of visible data using the producer/consumer pattern, so, when i get a refresh call usually at 60hz...at which pont i draw the last updated physics image without needing a lock, since its building a the next frame in another bufer at the same time . his old java code is pretty fast and interactive, easier to port to C#, but the new julia one is much more advanced.. its using quite alot of Julia math and standard libraries.. for performance. with ..net i can alos use generalize shaders at least for upscaling wth monogame so ill stick with mongame for now. Silk .net is making huge strides in generalized scientific visualization and gpgpu will works hopeful wth one code source , and easy to deploy... with 500 people on it i think the progress will outpace others. Since microsoft sells cloud services and use ARM.. I think they dont care so much about locking oto windows , certain hardware, or anything anymore, so we benefit from this new model. *But i can't understate the importance of this or a similar project...**i* think of this project because of how much effort Microsoft is putting in Net6 and now 7 has net intrinsics. possible google are openng sources too but they dont have a sort of IL as far as I know.. What this will do is justify the 20 year old unfullfilled promse of JIT compiling as the intrinsics on a particular device are not known A priori ( to use a trendy AI term from Kant) to deployment. SIM vectorization is now hideous code.. must be done for each type of processor, hopefully generalized and easier in Net 7. I don't know who to talk to at microsoft but tooling is another way they make money but JetBrains, a direct competitor has sponsored silk and its all under MIT licences.. Ther is very little squabbling and its alll transparent and visible in the vautls f anyone has worries about MSFT being a sort of evil /greedy / force or foottrap, ( anymore) .. i wont use hs code until mabye next year because its too slow and too exact for my needs so i invent first order chimera methods for now. but id love to just script it wiht Rosylin and buld it with and deploy it wht Net7 wiht fast math libraries (either Julia or MathNet) that use JIT and simd.

…

On Mon, Sep 26, 2022 at 4:16 AM BambOoxX ***@***.***> wrote: if i can actually run code like this https://github.com/weymouth/WaterLily.jl i would probably use it.. and see about doing the viusalization using a game engine like monogame or mabye Silk.. . his other version is in protobuld.. which is easy to port C# but doesn't not include any of his performance stuff... even if its a many months time, im just curious how likely that is.. i haven't evaluated the code yet, ran pretty fast before but on net 6 might be blazing and with his newest code. If you only intend to plot 2D stuff, try ScottPlot <https://github.com/ScottPlot/ScottPlot> — Reply to this email directly, view it on GitHub <#16 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AD74XGKADMPHDDUG2KJJEETWAFZYHANCNFSM6AAAAAAQIR5TVQ> . You are receiving this because you commented.Message ID: ***@***.***>

[HyperSphereStudio]

I have been working on this project this past winter break. Check out the new readme and feel free to ask if you have any questions. A lot of my energy has been going into developing my Spinor Prog Lang. (Which is actually using a lot of the ideas from this project). But I will be happy to answer them! This newest version does not have a lot of the features of the old version (V3) but I don't think anyone used them anyway. It was a nightmare it maintain.