deps: replace python-lzo by lzallright.

[qkaiser]

We propose to move away from python-lzo because it doesn't have a wheel, so you still need a C compiler and lzo sources.

This commit replace python-lzo with lzallright, a Python binding for lzokay, a Rust based LZO compression/decompression implementation.

This is a move we already did for Jefferson (see https://github.com/onekey-sec/jefferson/pull/14/files) and unblob (see https://github.com/onekey-sec/unblob/pull/573).

The only difference is that ubi-reader will be easier to install :)

[AT0myks]

I'm also waiting for python-lzo wheels. Windows ones might be coming soon (I just pinged the author right before seeing this issue), but yes it seems like lzallright is just a better replacement, though it's Python 3.8+ while ubi_reader is still compatible with Python 2. Maybe once support is dropped the switch can be made?

[qkaiser]

I don't think it makes sense to maintain backward compatibility with Python 2. Version 2.7 is no longer supported since January 2020, and it actually becomes a liability to support Python 2 since security fixes are no longer applied for it.

If you take a look at the download stats, Python 2.7 users represent 3% of the user base on average (see https://pypistats.org/packages/ubi-reader). Of those 3%, I'm pretty sure the vast majority are on a system that already supports Python 3.8 (released 2019).

[AT0myks]

I'm aware of that and I agree with you. I was simply pointing out the compatibility conflict. I would also like Python 2 support to be dropped so that the project can evolve but that decision isn't up to me. I think the author plans on doing that and maybe what you propose will encourage the change?

[qkaiser]

Hi everyone ! I suggest we continue the discussion here: https://github.com/onekey-sec/ubi_reader/discussions/74