The current processing page is a very early sketch (from fist version) and only quick look at all the running jobs. There was intentionally not much thought put into this, because it was all at change back then and this part of the app was simply not that important. But this should change, obviously in the future.
There are some issues already regarding this: #363, #364.
Besides the bad look & feel, the page also polls periodically the whole state of the job queue (including logs). This is not efficient and could be done much better, i.e. with websockets. But elm dropped its support for websockets in 0.19, which must now be implemented using javascript. This is a problem, as I won't maintain a non-trivial piece of javascript in my spare time :-). So I'd still go for polling then, or someone else volunteers to do it. However, I think when there are multiple endpoints that make the payload of the one endpoint smaller that is used for polling, this is ok.
I don't have a vision yet on how it should look like; I'm open for suggestions.
The current processing page is a very early sketch (from fist version) and only quick look at all the running jobs. There was intentionally not much thought put into this, because it was all at change back then and this part of the app was simply not that important. But this should change, obviously in the future.
There are some issues already regarding this: #363, #364.
Besides the bad look & feel, the page also polls periodically the whole state of the job queue (including logs). This is not efficient and could be done much better, i.e. with websockets. But elm dropped its support for websockets in 0.19, which must now be implemented using javascript. This is a problem, as I won't maintain a non-trivial piece of javascript in my spare time :-). So I'd still go for polling then, or someone else volunteers to do it. However, I think when there are multiple endpoints that make the payload of the one endpoint smaller that is used for polling, this is ok.
I don't have a vision yet on how it should look like; I'm open for suggestions.