Quote:
Originally Posted by dottedmag
OK, we're dropping all readers except V3 right now (while it sounds like "We're dropping all PC except Dell ones").
|
Especially that comparision highlights what is different, because especially on the PC there is not the "one project that does it all". There is a project that does a kernel for all kinds of PCs. There is a project that does User Interfaces. There is a seperate project that just links all this things into a distro and so on.
Quote:
Oh, wait! All we develop is portable source code and Wenjie and Ondřej just need to build several drivers in kernel... So even if we drop everything except V3, they will complete their GSoC tasks and we'll get the support for Cybook (+several similar devices) and Sony PRS-505. Damn! Our project will be doomed immediately 
I assume you're not familir with how things work in the field: the idea that every device is unique and requires thorough and hard work is a myth: all the boards are quite standard ARM devices and userspace is just the same for most Linux devices all around the world.
|
I just observe several years various OS projects that worked well and ones that didn't. And IMHO approperiate focus is a key issue. Take for example worldforge, they concentrated on everything and ruled nobody out, as "hey this isnīt our project" now over several years they still didnīt manage to come out with anything significant. I consider for example if they said people to come with "their stuff"" but wanted to be part of worldforge, if instead they said, okay but our project does only this and that, they'd be far more succesfull, sure some people would have just done nothing, but other would have changed their work to fit the project, so there were at least some people pulling at the same rope.
In your case, now okay so the e-ink devices are very similar. Making one linux kernel variation for all kind of e-ink readers is a good and complete project on its own. There is no need to intertwine it with a specific user OS. This should be another project, maybe it are the same people working on it, maybe they arent. But one key task should be one project. Maybe make the projects work with each other. But both should be seperate projects, possible seperate downloads/installment, possibly maintained by different people. Thats my criticism to put it all into the one big project.