Clarifications from David "Lefty" Schlesinger, Director, Core Tools and Technologie
I'd say that WR's calling us a "client" would be an mischaracterization. We discuss a number of things with WR [Wind River], we're doing ongoing evaluations of Workbench and some of the tools associated with it, and we work together on things like the OSDL's Mobile Linux Initiative and the Device Software Development Platform (DSDP) for Eclipse, but ALP is dependent on neither WR's kernel nor their toolchain.

We want the kernel additions and user-space components of ALP to be usable with WR's kernel (within reason as far as version, patchsets, etc., go), as well as with, say, Intel's kernel, TI's kernel, MontaVista's kernel, etc. (within a similar degree of reason).

This "kernel agnosticity" is a significant part of the reason that we put together a distribution for use internally and by our licensees on specific reference boards. Our own requirements for the kernel "ecosystem" might not mesh well with the road map of an otherwise-perfectly-good existing embedded distribution from WR or others, and we didn't want to be limited by someone else's schedule.

It's not our direction to be an "embedded distribution vendor" -- we market a platform intended for some fairly targeted purposes and our primary "value add" isn't in the kernel per se (although we do work in that space to support that "value add"). Of course, we're perfectly happy if a licensee wishes to work with the kernel ecosystem we provide.

Wind River has an interesting set of tools; we're actively looking into the possibilitiy of using some of them in our own development, and I wouldn't be surprised if they'd be useful to some of our licensees, another reason for us to be interested in working with -- while not being dependent on -- WR. But we're similarly working with Code Sourcery's EABI-compliant versions of the gcc tool chain, so as to be able to coexist with code compiled with proprietary toolchains like ARM's RVDS suite.

The other major inaccuracy actually seems to have been Linux Devices' rather than yours: Maureen O'Connell is _our_ PR representative, not Wind River's. Your "confirmation" of our relationship with WR represented a pretty significant departure from reality, however, and I wanted to correct that.

Wrap up
David also pointed us to the same press release quote listed above in Maureen's letter and the same clarification on the user-interface issues related to MAX and the Palm OS. Be sure to read the full release for more details. It's this kind of dialog between platform vendors and user community that may give the new PalmSource/ACCESS venture a good chance to create a viable product.

We'll continue to bring you more coverage on this over time.

Dr. Humayun Bakht completed his Ph.D at the School of Computing and Mathematics in Liverpool at John Moores University. Prior to his current course of studies, he has successfully completed his MSc Software Engineering and BSc Electronics Engineering from University of Liverpool UK and Sir Syed University of Engineering and Technology in Karachi, Pakistan. Humayun's research interest is mobile ad-hoc networks, in particular, the development of a new routing algorithm for this type of networking. He can be reached at humayunbakht@yahoo.co.uk and you can visit his Web site at http://www.geocities.com/humayunbakht/.