4 ms·

In my opinion the biggest mistake Purism made is to try to build their OS from the ground up, when acceptable and proven alternatives existed already. This has

In my opinion the biggest mistake Purism made is to try to build their OS from the ground up, when acceptable and proven alternatives existed already. This has repercussions even now, when users are justifiably annoyed at a mass produced phone with a non functional camera app. The Sailfish OS and Mer communities have had a working plumbing and porcelain stack for years (granted Sailfish's UI is not fully opened source) but Purism insisted on bootstrapping their own instead of building on those. I feel like they could have forced Jolla's hand to open source the last bits of their proprietary UI and everybody would have been happier.

fsflover 6y ago
Here is why they chose this path: https://amosbbatto.wordpress.com/2020/08/05/advantages-of-phosh/ https://amosbbatto.wordpress.com/2020/08/05/advantages-of-ph.... See also: https://source.puri.sm/Librem5/community-wiki/-/wikis/Frequently-Asked-Questions#39-why-did-purism-decide-to-create-the-new-phosh-interface https://source.puri.sm/Librem5/community-wiki/-/wikis/Freque....
- mariusor 6y ago
  Yes, I can understand the principial reasons of why they made the choice. But it's still a bad one form a pragmatic point of view.
seba_dos1 6y ago
It's the camera kernel driver that's not complete yet, not the app. There are several apps that should already work once the driver is there. Regarding the whole stack, I recommended reading https://puri.sm/posts/what-is-mobile-pureos/ https://puri.sm/posts/what-is-mobile-pureos/
- mariusor 6y ago
  > It's the camera kernel driver that's not complete yet, not the app. I was sure I saw in a different discussion that it's the other way around. Maybe you have a link ?
  - seba_dos1 6y ago
    You can see some bits from the ongoing progress at https://source.puri.sm/Librem5/linux-next/-/issues/44 https://source.puri.sm/Librem5/linux-next/-/issues/44 Once it works it will be just a regular V4L2 device on the system, there are plenty of GNU/Linux apps that can consume it and there are now even mobile-focused ones like https://git.sr.ht/~martijnbraam/megapixels/ https://git.sr.ht/~martijnbraam/megapixels/ and https://gitlab.gnome.org/jwestman/camera https://gitlab.gnome.org/jwestman/camera

fsflover 6y ago

Here is why they chose this path: https://amosbbatto.wordpress.com/2020/08/05/advantages-of-phosh/ https://amosbbatto.wordpress.com/2020/08/05/advantages-of-ph.... See also: https://source.puri.sm/Librem5/community-wiki/-/wikis/Frequently-Asked-Questions#39-why-did-purism-decide-to-create-the-new-phosh-interface https://source.puri.sm/Librem5/community-wiki/-/wikis/Freque....

mariusor 6y ago
Yes, I can understand the principial reasons of why they made the choice. But it's still a bad one form a pragmatic point of view.

seba_dos1 6y ago

It's the camera kernel driver that's not complete yet, not the app. There are several apps that should already work once the driver is there. Regarding the whole stack, I recommended reading https://puri.sm/posts/what-is-mobile-pureos/ https://puri.sm/posts/what-is-mobile-pureos/

mariusor 6y ago
> It's the camera kernel driver that's not complete yet, not the app. I was sure I saw in a different discussion that it's the other way around. Maybe you have a link ?
- seba_dos1 6y ago
  You can see some bits from the ongoing progress at https://source.puri.sm/Librem5/linux-next/-/issues/44 https://source.puri.sm/Librem5/linux-next/-/issues/44 Once it works it will be just a regular V4L2 device on the system, there are plenty of GNU/Linux apps that can consume it and there are now even mobile-focused ones like https://git.sr.ht/~martijnbraam/megapixels/ https://git.sr.ht/~martijnbraam/megapixels/ and https://gitlab.gnome.org/jwestman/camera https://gitlab.gnome.org/jwestman/camera