Fixfeat temp dir device name #38
Conversation
|
Interestingly, in this PR, I get the issues fixed in #36, as well. Didn't see them earlier... |
too lazy to rebase
Ah probably forgot to remove it from this branch before submitting |
|
Interesting combination of errors: As the files where not deleted from /tmp (fixed in this PR), the kill didn't actually delete the front pages and folder, thus the rear page scan could progress as intended. Only after this fix, the other error showed up. Sometimes, two wrongs do make a right... ;) Scans work via flatbed & buttons, ADF & web. Should be sufficient testing. ready to merge. @pedropombeiro : Would you do the honors? |
There was a problem hiding this comment.
Thanks for testing this @PhilippMundhenk
Side note: If it weren't for losing the ability to easily override the scripts, I'd be tempted to rewrite the scripts in something like Go, to make the logic easier to reason about, change, and test. IMHO Bash is too complicated for scripts like this (I understand this is how Brother recommends doing things with brscan4).
|
Indeed. This project has grown. The scripts used to be a hand full of lines and easy to handle. Supporting all kinds of new features as we do now really is not great in this setting. I was thinking of rewriting this, but frankly, I'm lacking time and motivation right now. I would likely use python though, which keeps the ability to override and easy adaptions, while making things much more handelable. The scripts would then remain empty shells calling (potentially a single) python script, handling all features much more easy to maintain. |
|
Yeah, Python would be a good choice. I might take a stab at it at some point. |
@pedropombeiro: This is the PR to get this to master, after manual testing.