[Replicant] Is the i9300 still the best phone for running Replicant? (and a few other questions)

[Sophie Hamilton]

>> (Yes, I realise the irony/hypocrisy about how I distrust binary blobs
>> in the firmware on my phone and yet I'm using a pre-built version of
>> Replicant. I'd like to be able to compile my own version of Replicant
>> in the future but so far haven't been able to look deeply into it.
>> This will probably change from now.)
> [...]
> In addition the builds are probably not reproducible, so if one day
> there is work on that, and that everything is built from source, then
> we could have way more trust in the binaries.
>
> Though in that case, it would also shift the trust on the distribution
> that is being used to build Replicant (here Trisquel 9 for Replicant
> 11).

I personally use Gentoo, and while it does have ebuilds that let you
install non-free software, I feel more confident when the free
software that *is* installed on here is compiled from scratch, and I
can read how.

>> 1. What is the best phone that can run Replicant right now? Is it
>> still the i9300?
> Between the supported devices, yes. It might change in the future
> though but for now work is being done on that phone for Replicant 11.
>
> Though as doak pointed out, it's doesn't support networks more recent
> than 3G, and this is a serious issue in several areas of the world.

I actually did not realise this. That does sound like a pretty serious issue.

Talking about this is actually reminding me that I'm also having
another issue with my phone, which I assume is another byproduct of
being on Replicant 4.2 - I have issues with calls from and to just
about any other mobile phone nowadays. (In the UK, where I live.) I
can call landlines and receive calls from landlines with no problem,
but whenever I call or am called by a mobile, generally one of two
things will happen:

1) Commonly, both sides will hear nothing.
2) Less commonly, audio on my side from the other person will be
heavily distorted (but speech can *barely* be made out) but I am heard
with no problem on the other side.

I'm assuming this is some kind of codec issue - maybe 4.2 doesn't
include newer codecs and whether or not 1) or 2) happens might depend
on whether the other side can fall back to older ones? - but that's a
complete guess on my part. However, it used to be that calls would
work fine, then some calls would fail in one of these two ways, and
now pretty much every mobile call will fail.

Is my diagnosis of the problem correct, or is there something else
going on here?

> The huge downside of that approach is that new modems would also need to
> be manufactured, compatibility with the modem connector would also
> need to be assured over time (so the device would still be obsolete
> when no new modem are manufactured for older connectors), etc).
>
> And that kind of design also makes the devices more expensive and so
> they don't spread that much in devices.

I've heard of phones like the PinePhone and the Librem, but know
little about them. Are there plans to run Replicant on phone like
these?

>> 2. I'm aware that there is work on getting the Lima graphics driver
>> working in the dev versions of Replicant, which as I understand it
>> would speed up graphics on the i9300 considerably.
> That part is already merged and it seems to work fine.

I'm glad to hear this - I imagine it'll make things a lot nicer to use.

>> Is there an estimate for when this might appear in a maintained version
>> of
>> Replicant?
> It's really hard to estimate, especially because I'm trying to do it
> right in a way that is somehow maintainable instead of rushing to get
> something that somehow works but that is unmaintainable and that would
> need to be redone from scratch.

Yeah, that makes sense. I'd like to help if I can.

>> 3. Talking of maintained vs. dev versions... as I said, I use my
>> phone as a daily driver and this would be my only phone. With that in
>> mind, I'm assuming it would be wise to only stick to maintained
>> versions of Replicant rather than trying dev versions.
>
>> Is this correct or am I making assumptions that are not necessarily
>> true?
> For Replicant 11, it's not ready yet to be usable daily.
>
> For Replicant 6, the 0004 RC5 probably works better than the Replicant 6
> 0003.

Thank you. I'll try giving these a go again if I can.

 - Sophie.