Palm prepping a linux OS?

According to some recent rumors palm could be preparing a linux OS completely seperate from Access’s ALP. This of course would explain why we haven’t seen any proof that palm is looking into ALP. However the rumors say this has been in the works for “years”. This doesn’t quite seem believable unless this is actually a aborted project from years past that’s been “renewed” in light of the up and coming ALP which does not belong to Palm. What they need about now is certainly a completely new OS and this if done correctly will certainly “revive” them. Of course depending on their integration with Palm OS I have serious questions about Access’s possible disagreement over this use which will essentially destroy ALP if true. Only time will tell, however I can say the only group that contacted two of us over at #shadowmite was the old PalmSource (now Access, and thus not this “mystery” linux group…)

http://www.engadget.com/2007/04/10/palm-prepping-its-own-linux-based-os/

EDIT: I guess I should state, this is NOT a rumor, it’s a announced fact. Although I still think this is a legal can of worms. If this is what I think it is, A BeOS relative, Access might just take issue with this.

This entry was written by Shadowmite , posted on Tuesday April 10 2007at 04:04 pm , filed under Cell Phone Related, News, PalmOS . Bookmark the permalink . Post a comment below or leave a trackback: Trackback URL.

2 Responses to “Palm prepping a linux OS?”

  • Spygoose Loves Palm says:

    Wow, it feels like 2004 all over again. When this fails, I wonder if they’ll go back to Cobalt!

  • ex-Be guy says:

    Most of the ex-Be folks went to the PalmSource side of the company after the split. I do not know who got the source, but my knowledge is it has been shelved since that time. (My understanding at the time was that PalmSource got both the BeOS source and the majority of the ex-Be developers. — I may have been wrong.)

    As BeOS was not at all based on Linux (or BSD for that matter) this is likely *not* a BeOS relative. The BeOS APIs were unique and depended upon kernel features not available in Linux.

    In fact, standard POSIX programming in BeOS performed very poorly. BeOS had native light-weight threads using their own API. IIRC, PosixThreads were only available as third-party add-ons. Also note that forking was *slow*.

    Not to mention that the Be sources no longer have features that are all that compelling. It has been shelved for so long that even the things that were cutting edge for the time are old news now. (This doesn’t even take in to account licensing issues with commercial products they integrated.)

    Their GUI APIs had a messaging architecture where there is no global appliction-level messaging loop — instead each window had a separate thread, each with a separate messaging loop. This is unlike all GUI environments in common use today.

    You also have to remember that BeOS didn’t have support for things like firewire/1394, bluetooth, USB 2.0, etc. The architecture also didn’t support plugging in kernel components from other systems. (Like, say BSD-derived systems.) This is why they wrote their own networking stack (twice).

    Palm would be better off doing their own Linux-based distribution. At this point the majority of the features they will want are already publicly available, and it is a matter of gathering them up, adding their own hardware support, and the glue to make it a cohesive platform. (It is work, but relatively straight-forward and much less than basing anything off of the BeOS sources.)

    Maintaining Garnet support may be an issue, but with Linux’s misc. binary support all they need to do is create the appropriate run-time environment / interpreter and they’ll be fine.