Yes, but marketing promises and good intentions aren't the same as vetted licenses -- in point of fact, many of the closed components simply are encumbered with external IP, regardless of what was claimed. Therefore, I think Maemo won't ever be completely open-sourced. I don't say it shouldn't have been, but the past is set.

(BTW, on reflection, "meant" was a poor word choice there, as it does imply I'm considering motives and such, not actual licenses.)

True -- and I don't deny it can be done for Maemo, merely that because of the prohibitive expense, it won't be. It's worth noting the relative resources available for the Symbian team vs. the Maemo team. And even so, it took them almost a year, and two evolutionary versions -- the open-source product at the end (Symbian^3) is not a simple relicense of the initial codebase (Symbian^1, or S60v5) -- if anything, what I know of the Symbian Platform story seems supportive of my assertions that it's a massive undertaking, and that gradual progress with each release is acceptable.

Well, I'm afraid the arguments, coming as they are from multiple people, don't always line up nicely. If I understand it, your perspective is that "They said it would be open; they should do what it takes to fulfill that promise."?

If so, that's well and good for trying to persuade them that they should, but I'd suggest that expecting a corporation to keep a promise made several years ago is not an accurate way to predict what will happen. Fundamentally, I think ROI is the way to try to predict their actions -- you are expecting them to invest significant expense in making some current version of Maemo open-source, and the only reason they actually will do that is an expected return, whether that takes the form of consumer trust (because they're seen to be keeping their word), increased sales, or (if the promise of open-source were legally binding) simply avoiding a lawsuit.


Problem is, I agree fully with that.

"Abandoning"? To me, asymptotically approaching is the exact opposite, and that's what's going on to date. As long as they keep being more open with each rewrite, I think it's undeniably a good thing, and IMO a better thing than seeing an old version completely opened at the expense of future development.

Then again, I guess I should say I never came away from any of their advertising with the notion of any real promise to be completely open, as you apparently did. Maybe my corporate-mouthpieces-are-always-lying filter was tuned too high and blocked it, maybe it's just because I didn't really follow IT news that closely till I got my N800 (shortly after the N810 launch leak) and all the solid promises were before that; I don't know. But as I see it, the promises really don't mean much at this point. If they were made, and are broken, then they obviously weren't worth the $MEDIA they were $PUBLISHed on, and if they weren't then they're also of no effect.

I guess I never saw Maemo or open source in general as some alternative to upgrading. After all, I don't run Linux 2.4 anymore -- how would even a completely open system the eventual need to upgrade to support new hardware and new applications?

To me, the principal benefit of an open-source OS is the hackability, the ability to make it do as I wish now, not the expectation that I'll be able to run it on any new hardware, or that it'll remain useful on current hardware forever. To this end, some components matter much more than others, and it's perfectly acceptable to me that some parts (ones that I can simply skip using) may never be open.

Where it ends? The key measure at which I'd consider it adequately open, and stop seeing a need for improvement from Nokia in the next version, is when a completely open-source image can be meaningfully flashed and run (yeah, that's somewhat vague, but I expect e.g. battery charging, wireless, cellular, etc. to all work.) So, while it's quite nice that Modest, MicroB, etc. are open-source, I could deal with them being closed by using sylpheed, Fennec or Tear, and the like. The platform is only decent, with a bunch of closed nuggets here and there, but improving.Some things Nokia keeps closed, like BME, and (IDK for sure) probably some radio-related stuff, for security purposes -- to make sure you can't trivially damage hardware or violate FCC regs. While I understand their reasoning, I strongly disagree with it; these bits will need to be reverse-engineered and open equivalents made by the community to ever make it a completely free platform.