OSNews

Well, it's an interesting thing, comparing the reasoning of BeOS users for waiting for Haiku versus committing to Zeta.

1. The best indications I see indicate YellowTab doesn't have the source to the system, or at least not the complete source.

2. As imperfect as Haiku is (after all, it's pre-alpha stage, and makes no claims otherwise) the code is 100% available and will live on somewhere as long as there's a single developer with interest in it.

3. There is enough market demand for an alternative to Windows/OS X (the two commercial powerhouses on their platforms) and Linux/Unix variants that is unified, easy to use, doesn't require huge hardware resources, and is sufficient to everyday tasks.

4. Some of those other Open Source OS's have incorporated massive amounts of source code from various sources that are incompatible with the license of Haiku, which might reduce their viability in the future (if there is one) of those other OS's becoming commercially attractive to commercial software developers, due to the taint of GPL. Haiku has taken a different route than Linux explicitly due to the tainting of GPL on the system itself: they WANT people to use and abuse Haiku without being whipped senselessly by the GPL should they wish to make their own custom versions, while still having fully opened source.

5. Who knows how long it will take YellowTab to release something more than a Release Candidate (multiple revisions of alpha/beta versions of post-R5.03 BeOS) with their manpower (or lack thereof) management and attention to detail (investigate the release candidates for evidence to point to that: I don't need to say anything here, as the evidence screams for itself) and their ability to modify the kernel itself, and the various kits?

6. Who knows how long Bernd can keep the money coming in sufficient to cover the costs, regardless of how good/bad/marketable Zeta is?

7. There are serious questions regarding the details of the legality of what YellowTab has. What will this do, and will it become a real issue for business, or is it already too late?

8. As Axel has pointed out, it takes real time to do real solid development, as that requires a lot of testing and attention to details. For the first version of Haiku, the regression tests must incorporate (to a fairly large degree) the testing for the quirks of BeOS 5.03 lest it break the existing binaries and source code of existing applications. In effect, that adds a large process of discovery in the development, as it requires actually testing TWO different operating systems. So, while the API is mostly documented/engineered for syntax, the semantics make it an interesting target to replicate. If they didn't care about semantic compatibility, the task would be much easier, as they could implement something that seems reasonable based on the general concept, and pronounce it done.

So, will Haiku or Zeta or any other BeOS compatible OS (source or binary) be viable, if they are still at all viable? Time wounds all heels, I guess. If so, I may end up being wounded to some degree, as I'm working on an IDE for them... If nothing else, it's something I can put on my resume, and enjoy developing applications with it, even if the OS it is targeted for has major flaws that make me cringe. Regardless of what BeOS zealots may say about the design beauty or implementation of BeOS, objective evidence demonstrates that Be didn't get some major issues fixed in terms of reliability/stability, even with all that manpower and time: the need to pay the bills forced their hand in sending it to customers before many gaping holes were sealed. Nonetheless, it isn't a useless OS, and it isn't beyond redemption.

And no, I'm not an Free Open Source Software zealot, and I'm not against commercial software making money. However, of the two choices I presented of Zeta versus Haiku, Zeta may be a short-term stopgap measure for those that actually have hardware it will work on, and if YellowTab lasts long enough, they may become a Haiku distribution maker as they transition over to the work done by the Haiku team. The current Zeta source code issue makes it very unlikely (if the indications are true) that they will be able to continue Zeta beyond a certain point without a heavy infusion of Haiku blood, sweat and tears.