MobileRead Forums - View Single Post

hacker · 12-08-2004, 11:23 AM

I think the best approach towards leveraging the mindshare already built up in the Free Software community around Linux and Linux-based solutions is a huge benefit to Palmsource, since they can't keep up with their svelte staff, innovate, and beat the competition. They simply don't have the manpower or dollars to make it happen.

With Palmsource profits slipping due to lackluster updates in their OS (and the recent OS-specific bungles), they need to make some pretty major leaps in their OS functionality, but how can they do that when they can't pay more people to work on it for them? Enter the good-will fostered by the Free Software community.

That being said, I think providing the upper-level "glue" for the OS on top of the Linux kernel (not the Linux "OS", remember, they use a strictly-licensed KADAK kernel on Palm handheld devices), is a huge win/win for them. Provide a useful toolkit + an SDK as they do now for their Windows development, and they have a good deal of support waiting to help them.

...if they don't piss everybody off. Palm* (source/One/Inc.) has recanted on a pretty strong statement that they were going 100% Windows a couple of years ago, if this proves to be true.

They also immediately gain support for running PalmOS on any architecture, not just m68k or ARM, because Linux runs on 32 separate chip architectures. If its logically separated properly (and with a BeOS core, I'd have to imagine that it is), then porting to a new arch is simple, because you don't have to worry about the underlying kernel bits.

For me, of course, this is very significant, because my code is THE toolkit, libraries, and userland tools used to communicate with Palm handheld devices under Linux and UNIX platforms. Nothing else even comes close (in fact, several commercial companies use our code in their proprietary products, including one prominent one using it on MacOSX... MarkSpace, with their MissingSync product).

I just hope they do things the right way, by working with us and not against us.

Update: I should also mention, after some people have emailed me privately with their concerns, that just because something runs on Linux, doesn't mean the source has to be provided. Palmsource is not "open sourcing" PalmOS with this venture. They are simply considering running PalmOS (and perhaps only the GUI-specific bits at that), on top of a Linux kernel and libraries. Until they provide more detail, we can only guess how deep that goes.

12-08-2004, 11:23 AM	#4
hacker Technology Mercenary Posts: 617 Karma: 2561 Join Date: Feb 2003 Location: East Lyme, CT Device: Direct Neural Implant	I think the best approach towards leveraging the mindshare already built up in the Free Software community around Linux and Linux-based solutions is a huge benefit to Palmsource, since they can't keep up with their svelte staff, innovate, and beat the competition. They simply don't have the manpower or dollars to make it happen. With Palmsource profits slipping due to lackluster updates in their OS (and the recent OS-specific bungles), they need to make some pretty major leaps in their OS functionality, but how can they do that when they can't pay more people to work on it for them? Enter the good-will fostered by the Free Software community. That being said, I think providing the upper-level "glue" for the OS on top of the Linux kernel (not the Linux "OS", remember, they use a strictly-licensed KADAK kernel on Palm handheld devices), is a huge win/win for them. Provide a useful toolkit + an SDK as they do now for their Windows development, and they have a good deal of support waiting to help them. ...if they don't piss everybody off. Palm* (source/One/Inc.) has recanted on a pretty strong statement that they were going 100% Windows a couple of years ago, if this proves to be true. They also immediately gain support for running PalmOS on any architecture, not just m68k or ARM, because Linux runs on 32 separate chip architectures. If its logically separated properly (and with a BeOS core, I'd have to imagine that it is), then porting to a new arch is simple, because you don't have to worry about the underlying kernel bits. For me, of course, this is very significant, because my code is THE toolkit, libraries, and userland tools used to communicate with Palm handheld devices under Linux and UNIX platforms. Nothing else even comes close (in fact, several commercial companies use our code in their proprietary products, including one prominent one using it on MacOSX... MarkSpace, with their MissingSync product). I just hope they do things the right way, by working with us and not against us. Update: I should also mention, after some people have emailed me privately with their concerns, that just because something runs on Linux, doesn't mean the source has to be provided. Palmsource is not "open sourcing" PalmOS with this venture. They are simply considering running PalmOS (and perhaps only the GUI-specific bits at that), on top of a Linux kernel and libraries. Until they provide more detail, we can only guess how deep that goes.