What is the current state of LIOS?

[Linux for blind general discussion]

Hi Ratislav,

All I know is that the last commit dates back Jun 27, 2021 and it was a merge of
a PR: https://github.com/Nalin-x-Linux/lios-3

Maybe try to file an issue?

The are 16 forks (including mine) , none with enhancements or bug fixes.

Cheers,
Didier
--
Didier Spaier
Slint maintainer

Le 28/07/2022 à 21:53, Linux for blind general discussion a écrit :
> Hello list,
> 
> today, I finally decided to get my scanner working, and thus, try out
> LIOS as well.
> 
> 
> Unfortunately, this turned out to be a little more troublesome one would
> expect.
> 
> 
> First of all, LIOS was unable to parse the information of my network
> scanner, there already seems to be an issue about this on the project's
> GitHub, if the causes are identical.
> 
> 
> This didn't really surprise me, there are many scanners and I'm not sure
> how precisely are they following the standards, so okay, I dived into
> the code and fixed dthe parsing on the place where it broke.
> 
> 
> But the other issues I found were far more stunning.
> 
> 
> First of all, the update scanners list function doesn't have scoped
> try/except blocks, but rather one big, that basically covers all the
> function's code.
> 
> 
> What this means in practice is, that if LIOS fails to open a scanner
> (like it did with mine), instead of proceeding to others, the function
> simply aborts and that's that.
> 
> 
> My scanner is represented by approx. 3 or 4 devices, I get one in the
> list, since the opening of the second-one crashes and the others are
> thus not loaded.
> 
> 
> I'm not sure whether there is any difference between them in this
> particular scenario, but if their order was either different or I had
> multiple connected scanners, the devices evaluated after the
> problematic-one would be simply unreachable.
> 
> 
> And there is another issue, that's totally cryptic to me.
> 
> 
> When I choose to scan an image with the scanners already loaded, instead
> of simply using the selected-one, LIOS updates the list first (what
> takes quite a some time), and just afterwards makes the scan.
> 
> 
> That's just obviously wrong, so I checked the code again, and this is
> what I found in the scan methods:
> 
> 
>          if(not self.is_updating_scanner_list or
> len(self.scanner_objects) == 0):
>              self.update_scanner_list();
> 
> Translated to english, this says: if the scanner list is not being
> updated or there are no scanners, update the list.
> 
> 
> This kind of shocked me. You don't even need to be a programmer to know
> it's wrong, the correct formulation would be: if the list of scanners is
> not being updated and there are no scanners, update the list.
> 
> 
> i.e. you need to change one word. A trivial, elementar mistake.
> 
> I'm not sure who did it, I kind of can't believe someone who coded a
> program of this complexity would, perhaps it came with a PR.
> 
> 
> What bothers me more is why didn't it get fixed, it's not something
> you're going to overlook as scanning is the main function of the
> program, and dsince it's so easy to fix, I would expect it to be already
> repaired.
> 
> 
> So, the question is, does anyone have any information on what's the
> current status of the development?
> 
> 
> Is it still active, are PRs accepted or, didd anyone create a fork if not?
> 
> 
> Thanks!
> 
> 
> Best regards
> 
> 
> Rastislav