(2025-02-03) A message to copycats
----------------------------------
I didn't even suppose that I'd have to write anything like this at any point
in time. Yet here we are. This is a post for a specific parts of my
audience, as that part recently got surprisingly large.

As you know pretty well by now, I am a strong proponent of open source and
*absolute* software freedom. That's why I usually release my own code into
public domain. That's why I totally do not object to usage of my ideas in
the others' projects. However, you know, sometimes it really hurts to see
how the projects employing those ideas are... mediocre at best. As soon as I
abandoned KaiOS-related development, I hoped that practice would cease. Alas.

For instance, tell me please, what kind of nonsense is ([1]) this? The author
even mentions several posts of my phlog as the reference points. He
introduces the final step that I haven't even published yet (although it had
been found back in the late December, and I will prove this when my own tool
repo goes public on Mar 1 2025). Yet the logic he uses in the explanation
("Technicalities" section) is not completely wrong but contains several
redundant steps. On top of that, he states that the method "will not work if
you want to change the phone's imei number to something other than what it
had originally", which is not true. In fact, the step that I had been
missing but he added makes it work on any IMEI combination (without it, as
you remember, both IMEIs were zeroed out, and there only was a way to bypass
this on the 6/6 Pro). You'll see it for yourselves when my tool is
published. As such, he has written twice more code than necessary, and
presented all this in the most uncomfortable way to use.

But this is not even the most egregious example. At least it's open-source
and honest, mentioning all credits below. There was, however, some other guy
who created a Faildows-only Tkinter-based application in Python that
couldn't even work without a ton of external components in the same folder,
and a part of the logic was baked into this application but a part of it was
inside an external script. Well, I managed to decompile it. It's a total
mess. Even more extra steps, even more confusion, and the level of GUI is
like... a third-grade student could create a better one. I wouldn't be
surprised if the GUI part was AI-generated. And the author decided to remain
completely anonymous (maybe to avoid public humiliation). The only thing I
found as the hint to his origin was the docstrings. Which, in case you
didn't know, remain intact in the compiled .pyc bytecode files, unlike the
"normal" single-line Python comments. This, AFAIK, was a deliberate design
decision for Pydoc to be able to generate documentation from compiled
modules as well. So, I saw those docstrings and they were written in
Turkish. I'm well aware that Turkey has a big problem with imported
cellphones, but come on, you don't have to be that sloppy.

The main question is: what drives all those people? If they don't seem to
have a basic understanding of what's going on and just paste whatever they
found into their haphazardly put together utilities, then why publish them
at all? If you require root access anyway, might as well make a decent
autonomous tool to use on the device itself, or at least something that's
pleasant to use and optimized on a fundamental level. It's as if there is no
mid-space between the amateur script kiddies and "flashing box" thin air
sellers. No, unlike the "boxers", I respect every effort to spread the
knowledge, but this knowledge has to be based. If you don't understand what
you're doing, then you're not really helping anyone in the long run. What
will you do when/if they change the baseband in Pixel 10? Yeah, you'll wait
for someone to research it again.

I had started all this in an attempt to prove that the GrapheneOS devs were
totally wrong, and now I have everything at my disposal to prove it.
Whatever y'all do with the results of this proof is none of my concern. The
spring is coming. Mark your calendars.

--- Luxferre ---

[1]: https://github.com/bitdomo/restore_imei