Repository navigation
Replies: 4 comments
|
I have started doing something in this direction a while ago, but got nothing nearly production ready out of it. Maybe over the next months I will find some time to push it towards something usable, but don't get your hopes to high for this. From my experience so far there is actually very little gain out of a symbolic representation for the elements, because the Ferrite codes are already super flexible and concise. |
|
I am not sure researchers care much about solving simple text book problems in a very compact way. And for actual hard problems I don't think it is very easy to provide flexible enough high level syntax that is actually helpful. In some ways, Ferrite is already too high level. There are many real problems that are hard to implement in Ferrite because we don't have the features available and it isn't really possible to hook into various parts in a convenient way to implement them yourself (say dof distribution of a field living only on a facet). |
|
If you are interested, I shared my thoughts about this back at FerriteCon2023 https://www.youtube.com/watch?v=mxtwoDtqrDo My conclusions from then pretty much stand today and I think align with @KristofferC and @termi-official above. |
|
Than you all for the detailed explanation. |
Uh oh!
There was an error while loading. Please reload this page.
Dear Ferrite Developers,
Is there any possibility for future to have high-level finite element programming using Ferrite?
I believe with that Ferrite will be "one-stop-shop" for researchers.
What do you think?
Thanks in advance.
@KristofferC
@fredrikekre
@KristofferC
@Kevin-Mattheus-Moerman
All reactions