The Librem 5 Smartphone Software Made More Progress In May But Still No Hardware Signs

[phoronix]

Phoronix: The Librem 5 Smartphone Software Made More Progress In May But Still No Hardware Signs

Purism just published a monthly summary of their activities pertaining to the Librem 5 smartphone this month. They continue working on their software stack with the Librem 5 developer kit but there still is no sign of their production hardware design yet or if they'll be able to ship next quarter as planned...

http://www.phoronix.com/scan.php?page=news_item&px=Librem-5-May

[drohm]

Is it just me or does Michael really not like/trust Purism?

[Guest]

Is it just me or does Michael really not like/trust Purism?

The article looked dry to me, it's normal to begin to doubt things when issues keep popping up.

[LoveRPi]

It takes at least a LTS Linux releases to get an upstream kernel in shape for a platform. If the hardware design is dependent on that, then it takes at least 2 years.

Librem is aggressive on the timeline front and facing all kinds of stack issues. Hardware is easy, software takes crazy amounts of time for the noble effort of upstream.

Sunxi community for Allwinner took 5 years to get where it is. Libre Computer's first Amlogic boards took 2 years to finally get into shape. Rockchip platforms are finally getting the first usable upstream with Linux 5.3 with u-boot nowhere in sight.

It never moves as quickly or as easily or as cheaply as people think. Your hardware cost is maybe 1/10th the total effort cost but when completed, it's a beautiful stable creation.

[andyprough]

PinePhone seems to be moving along nicely. Second set of development devices are out to devs I believe. For $150 and nearly all open source firmware/software and physical kill switches, this is going to be one to watch.

Here's a picture from last week of SailfishOS lock screen using open source Lima driver, Mesa and kernel 5.1. on a PinePhone dev kit: https://pbs.twimg.com/media/D7WKEbTXkAAu7EL.jpg
And a week and half old video of LuneOS (WebOS community fork) on Pinephone devkit, with patched Mesa/Lima for QtWebEngine on PinePhone: https://www.invidio.us/watch?v=3SyqbI24qu0

[tildearrow]

That on-screen keyboard is too small... Is this supposed to be a Palm/Nintendo DS-like device (keys too small you need a stylus to type) or a modern smartphone?

[caligula]

Sunxi community for Allwinner took 5 years to get where it is. Libre Computer's first Amlogic boards took 2 years to finally get into shape. Rockchip platforms are finally getting the first usable upstream with Linux 5.3 with u-boot nowhere in sight.

Do you how many paid work hours is the effort equivalent to? I suppose the previous hardware had to be reverse engineered without any documentation.

[starshipeleven]

That on-screen keyboard is too small... Is this supposed to be a Palm/Nintendo DS-like device (keys too small you need a stylus to type) or a modern smartphone?

that's a bit smaller than normal, and I don't like the small size of the text on the buttons, but a modern touchscreen can handle it. I mean smartphones with smaller screens existed (pepperidge farm remembers) and their keyboard wasn't that big either.

With decent capacitive touchscreens (hard surface, does not bend when touch, you don't need to apply force to operate them) the touch event is sent to the center of the pressed surface, it's not like a physical keyboard where if you push more than one key... you are pressing more than one key.

Nintendo DS and Palm had resistive touchscreens (plastic surface, does bend when you press it), resistive touchscreens are very rugged but always sucked at precision.
Even with industrial control panels (that have to be rugged) when you need anything resembling fine control you need a stylus (or a stick), and therefore buttons and keyboards are comically oversized.

Also, do note that capacitive touchscreens have bigass "stylus" devices that aren't pointy at all.

That's without going into the powered stylus that senses the pressure you apply to it (which is mostly for artists).

[LoveRPi]

Do you how many paid work hours is the effort equivalent to? I suppose the previous hardware had to be reverse engineered without any documentation.

For a single platform probably 10000 man hours to get everything outside of the GPU upstream.

[Michael]

Is it just me or does Michael really not like/trust Purism?

More information will be exposed in time.