Sun's StarOffice Release: Not Open Source

Bruce Perens wrote in to send us a link to an article he has written about Sun's StarOffice Release: Is It Really What You Think? It discusses the implications for Linux, and the fact that it technically isn't going to be released under a true Open Source license. Worth a read.

Comments

Few idle wonderings

[Pascal Q. Porcupine]

First of all, whatever happened to the Sun/AOL thing? Did that go through, or what? Given the lukewarm response Mozilla has led to, I don't think AOL would really want to deal with having "their" source code in the public eye.

Secondly, I can understand where Sun's license is coming from, and really in practicality it's not really any different than most opensource projects, except for the notable difference that only Sun can distribute the source. I mean, most opensource projects just have a single code maintainer and a single codebase when it comes down to it (though most projects using CVS these days does kind of distribute that load). However, this is disregarding the fact that in most opensource projects, the person(s) running that project usually care enough to make sure that it keeps with the freedom that opensource brings, and for the ones which aren't, someone can always fork development and do their own thang, preferrably better.

My main concern with the Sun license is that they may just sit on their asses and never incorporate any changes. At least with the GPL, should Linus suddenly get the stupids, someone else could fork it and release their own source tree. Also, hasn't this technically already happened with the various embedded projects, etc.? Though StarOffice doesn't immediately seem to be the kind of thing that'd need specialization-based forking, it probably could be. StarOffice on a wearable computer, for example, would need to have a completely different interface, which means either hoping the Sun folks would want to maintain multiple source trees (unlikely) or forking the source (not possible right now).

Whatever the case, at least this gives people many more choices. If anything, at least people can feel free to use Linux on any architecture and compile it for it, rather than having to use a supported architecture. What good does StarOffice binary releases for IA32, PPC, RS6k, MIPS, and StrongARM do if you need to run it on a simulated Merced?
---
"'Is not a quine' is not a quine" is a quine.

What concerns me (if you care)

[Booker]

I'm concerned by all the comments on this story saying "Well, it's free, and that's good enough for me. Who cares if it's open source."

On a case-by-case basis, this works just fine. Take yer free stuff and be happy. But when you look at the big picture, it falls apart. I don't think I need to go into specifics on this - just think about this scenario:

Torvalds Software, Ltd. releases the Linux Kernel (c)(tm)(r)($) for free, but with restrictions on it's commercial use and redistribution. Xfree86 Partners, Inc. releases their version of the X-Windows system, but you can't make any modifications to it. Gnu Gurus, LLC has a whole nifty suite of Unix-compatible tools, and their down-rev versions are always available in binary form, for free download. These three things together are a good foundation for a "free" Unix system.

But - given that foundation, do you think Linux would be anywhere near where it is today? Would anyone be developing any software for this thing, closely held by some companies, under licenses which may someday change radically?

To me, accepting "Free Beer" software is akin to saying "That's enough - we can stop moving forward now" because if enough of the "foundations" of Linux become less-than free, I really believe that the whole thing will slowly fall apart in the future. I think that "free beer" software, much more so than traditional commercial software, contributes to this complacency.

Sometimes, if there's no alternative and I have a task I must accomplish, I'll use non-free software, but I try to avoid it for the reasons above.