On Sun, Feb 20, 2022 at 1:08 PM Jon "maddog" Hall via lpi-examdev <
[email protected]> wrote:

> /*
> I was just pointing out that some (e.g., LZ4/p7z) have program that are
> getting blacklisted by the US, which does affect some American allies as
> well, and definitely American companies operating in Europe (and
> elsewhere).
> */
>
> I guess I missed something.   The US is blacklisting compression
> techniques?  Or are they blacklisting encryption techniques?  The latter I
> could *almost* understand, but the former?
>

And that's something I brought up from the get-go, but has been lost in the
various responses (along with other things).**  ;)

You are correct, it's only implements, not the algorithm, and some newer
programs are available ... while some programs don't various aspects other
than block -- e.g., p7z is becoming rather proflierated, and quite
troubling.  So ... do we 'step back' and do what some suggested ... and
remove most or even all compression programs, and just leave the general
concepts?  Or do we need them to be aware of  the lot?

After this discussion, I'm almost at the point that only LZ77-gzip should
be covered for everything from zImage to commands, with the note that
others are available.  DEFLATE is being commonly used in many algorithms
now, as well as compression and/or archivers, with more and more focused on
speed and program size (for embedded loaders).

>
> In any case, if you are going to remove specific packages of compression
> techniques (fair enough), you should at least talk about the differences
> between lossy vs lossless techniques and where they might be used and how
> they might or might not be useful in combination with archiving tools
> (lossy compression of archiving tool output is a recipe for disaster).
>
> These were all lossless algorithms and compression libraries/programs.

- bjs

**P.S.  As always ... I try to avoid giving opinions (even if I call them
at times), but try to analyze comments by others, first, and where there
can be valid considerations as well as risks.  It's up to others to
remember the context I make things in -- e.g., the Ansible comment went way
off where I even intended, and I never asserted only Ansible.  Although in
your case Maddog, it looks like you're reading the digest, which won't have
it all.
_______________________________________________
lpi-examdev mailing list
[email protected]
https://list.lpi.org/mailman/listinfo/lpi-examdev

Reply via email to