[conspire] Christian draws out a Sun Microsystems guy
rick at linuxmafia.com
Fri Apr 29 18:08:18 PDT 2005
Quoting Christian Einfeldt (einfeldt at earthlink.net):
> Hi Rick, thanks for mentioning this interview here.
No problem. It was interesting.
> Ultimately, I believe that open source is as unstoppable as a rising
> tide. I believe that we will see gcj and other open solutions to
> the Java problem.
Red Hat is giving that project massive amounts of help. Sun could
certainly, at bare minimum, commit to making sure that OO.o 2.0
functions fully on release versions of gcj/Classpath, i.e., on the
leading open source Java runtime and support libs.
But they don't. They're going out of their way to make OO.o dependent
on _not just_ the Java language (which is not a problem in itself), but
specifically on their own proprietary Java support software -- and they
duck all questions about that.
> I do believe that Simon Phipps cares about open source
> So Sun apparently feels the need to send out mixed messages
That may be, but, as usual, intentions (not to mention perceived or
professed intentions) and "messages" are the booby prize: One of my old
sayings is that the only things that matter in the long term, in open
source, are code and licensing, licensing and code.
What ultimately matters is not Simon Phipps's style of expresion, but
rather Sun Microsystems's code and licensing, licensing and code. The
style of expression merely gives one hints.
> But I do think that we should have more sympathy and gratitude for
This seems a bit of a non-sequitur to my factual observation, which was
that Phipps had ducked your question. It was:
MP: some people on the OpenOffice.org list are saying that the
GCJ is getting good enough to be a substitute for a JRE, others
are saying that's still far from being the case. What do you
Phipps's "answer" was to first say he "haven't really looked into
[gcj] much", and then blather at great length about "Java platform
compatibility" -- which as you note was coded anti-IBM marketing-speak.
Anyhow, _say what_? You asked about gcj -- impliedly asking whether Sun
was going to ensure that OO.o 2.0 works with it -- and his sole answer
is to say he "hadn't really looked into" the flagship open-source Java
implementation, and then change the subject?
What's up with that? Are we supposed to believe that one of Sun's chief
technology guys has never even given serious _thought_ to the main
open-source stack? If so, what does _that_ say about Sun's intentions
Saying we should be sympathetic and grateful to Sun strikes me as all
very well, but as changing the subject. The question was: Is Sun
intending to ignore the open source community's concerns and make new
OO.o versions (avoidably) dependent on its proprietary software? The
answer seems to be "yes" -- and they're not even forthcoming enough to
come right out and answer the question.
> Remember, too, that we in the open source world need corporate
Not really. Corporate allies come and go. _Code and licensing_ matters.
(Well, patents do, too, but only for 20 years from filing.)
> Microsoft Office delivers a sufficiently high level of
> competition that AbiWord and Koffice alone are insufficient to
> respond to.
You know where the big advance in MS-Word import/export came from?
Not mostly from OO.o, but rather from Caolan McNamara's wvWare.
Anyhow, if you're saying that we should turn a blind eye to the
gratuitous and less than forthright introduction of gratuitous
dependency on proprietary software out of gratitude to Sun's other
contributions, I call that a non sequitur.
> So while some healthy skepticism of Sun is warranted, IMHO, the
> burden of proof shifts to Sun critics to come up with a functional
> replacement for Java and for OOo [...]
IMVAO, this is changing the subject and trying to distract attention
from the question at hand.
I have an idea: Why don't you pose my questions to Sun directly, and
see if _you_ can get a straight answer? I'll bet you $10 that you'll
get nothing of the sort.
> > The unstated (and perhaps a bit too subtle) subtext of that
> > write-up is that, yes, one can write Java that runs entirely on
> > open-source Java tools -- but the problem is that almost all
> > extant Java codebases are written by people who don't give a damn
> > about free software,
> Is that hyperbole?
Not one bit.
Don't take my word for it; look (for example) at the Java entries in my
list of all known Linux MUAs -- as I suggested. Notice the overwhelming
presence, indicated in the details of system requirements for each
program, of dependencies on proprietary Sun libraries. You will find
that that tendency is nearly universal, among Java codebases.
> What about Mono? It is my understanding that Mono is Miguel de
> Icaza's answer to .Net, and that .Net was Microsoft's answer to
> Java. In other words, as Mono gets better, will Sun have to end of
> life Java?
I may be missing something, but I think you just took a left turn,
there. Mono is not a Java implementation at all. It's some sort of
bloatware C# / gtk+ / gnomelibs / glade developer thing. You seem to be
attempting to start some entirely different discussion. Developments or
lack of same in Mono would have no effect on OO.o whatsoever.
But I you eventually reach:
> This is another reason that I am not worried about Java. Under the
> competitive pressures of Mono and other truly open solutions, Sun will
> eventually have to let Java go. Am I way off base there?
I would have no idea. And I try not to make any plans based on
estimates of far-off possible future changes to proprietary codebases
that may or may not still exist at that time. The history of software
teaches us, among other things, the lesson that proprietary codebases,
like software companies, are subject to coming and going without notice.
Which is part of the reason why many of us got tired of that game and
went open source, instead.
> Again, is this something which will eventually be commoditized by
> the open source community (Mono?) or will Sun be forced by the
> rising tide to let Java go?
Not a clue. I'm more comfortable talking about what's here and now --
such as dependency on proprietary software or not. But, since you
asked: All of those MUA codebases would have to be fundamentally
rewritten from scratch to run on the Mono runtime or anything like it,
so at that point they might as well not exist.
> > When Phipps says "the Java community is one of the most free
> > software communities that has ever existed", he is of course
> > brazenly lying,
> ooh, harsh!
I'm entirely serious, and willing to put my money where my mouth is.
As I said on LinuxToday:
[Phipps] includes howlers like "the Java community is one of the most
free software communities that has ever existed". (I'll pay Phipps $10
for each publicly archived Java project he lists that's fully functional
using only open-source JVM and library software, if he'll pay me $1 for
every one I cite that isn't.)
Please feel welcome to tell him that I'll be glad to deliver on that
wager, any time. Meanwhile, again, don't take my word for it: Examine
a dozen or so randomy selected Java codebases, and notice their
dependencies. You'll see.
> > for reasons amply demonstrated by my MUAs
> > example. But I feel a little sheepish pointing that out, because
> > it was probably there as flamebait.
> also harsh!
It's probable flamebait because the claim is blatantly ridiculous to
anyone who's actually _looked_ at typical Java projects' system
requirements. Again, you can check.
> I understand that Simon is a corporate spokesperson, but I have
> trouble believing that he was either lying or flamebaiting.
Check for yourself. You'll have to conclude that he was either
dissembling or is unforgiveably ignorant of his own firm's technology.
> After the interview, Simon told me that he is discouraged by what he
> perceives as religious type discussions, and he thinks that heated
> discussions are not helpful for the OOo community.
That's all very well, but it's a distraction from, and evasion of, the
actual matters under discussion: Is Sun going to release OO.o 2.0 in a
form that is only semifunctional without Sun's proprietary Java support
software, yes or no? And why on earth has he "not really looked into"
the flagship open-source Java implementation? And why is he blandly
talking about Java-based hsqldb's selection as the OO.o Base engine
without saying even _word one_ concerning the loud and ongoing attempts
to convince it to adapt sqlite, instead?
And I'm sorry to hear about their business problems, but what really
matters in the long term is code and licensing, licensing and code.
> If I get it correctly, Harry's point is that to avoid conflict or
> because of intellectually laziness or intellectual weakness, we
> have a cultural tendency to accept spin as an unavoidable evil.
Not really. I've read that essay a while back (also published in an
ironically-padded paperback volume by my alma mater, and the gist of it
concerns the existence of an entire category of non-knowledge that we
accept as a routine part of our lives, that he dubs "bullshit", that is
accepted as inherently not fair game to testing against reality, but
rather as transforming the basis of discussion from fact into tall tales
-- claims as performance art. Deliberate humbuggery. Consciously empty
talk. Verbal fakery that is nonetheless accepted as OK even though it's
accepted to be probably fake.
Frankfurt points out that bullshit can serve a useful social role, i.e.,
bullshit answers to otherwise incendiary personal questions are an
accepted way to deflect them without seeming rude.
In other circumstances, it is just a social convention to agree to
regard the truth as not mattering, sometimes solely for the speaker's
benefit, sometimes not.
> Harry apparently ends his book by saying that "sincerity is
> bullshit." One Amazon reviewer of the book says that Harry "points
> to one source of bullshit's unprecedented expansion in recent
> years, the post modern skepticism of objective truth in favor of
> sincereity, or as he defines it, staying true to subjective
Well, actually, his closing paragraph is sort of a rhetorical flourish.
But it is preposterous to imagine that we ourselves are determinate,
and hence susceptible both to correct and to incorrect descriptions,
while supposing that the ascription of determinacy to anything else has
been exposed as a mistake. As conscious beings, we exist only in
response to other things, and we cannot know ourselves at all without
knowing them. Moreover, there is nothing in theory, and certainly
nothing in experience, to support the extraordinary judgment that it is
the truth about himself that is the easiest for a person to know. Facts
about ourselves are not peculiarly solid and resistant to skeptical
dissolution. Our natures are, indeed, elusively insubstantialnotoriously
less stable and less inherent than the natures of other things. And
insofar as this is the case, sincerity itself is bullshit.
As I was suggesting earlier, Princeton University Press's promoting this
tiny 8000-word essay as a "book" is... well... bullshit. ;->
> Simon's statement is more complex, because there is a type of
> suicidal isolationism present in the FSF's insistence on using only
> Ututo-e as a GNU/Linux distro.
Amazingly enough, FSF isn't the topic of discussion, any more than
Phipps's (or Sun's) intentions are. Even less, really.
> I am willing to go to great lengths
> to use OOo, rather than Microsoft Office or WordPerfect Suite as my
> office suite, but I am not willing to rely on only AbiWord and
> Gnumeric for my office suite productivity needs.
And when/if Sun's proprietary Java suddenly vanishes from availability
and ceases to be maintained, what then? Back to OO.o 1.3.x with the
rest of us? ;->
> I think that it would probably NOT be correct to say that OOo is
> more free than AbiWord....
Software that's dependent on proprietary software _is_ less free than
software that isn't. If you think otherwise, then you're playing games
> As a side point, I am always a bit bugged by seeing people lug Apple
> products around to open source events, and yet where is the
> criticism of Apple's lockdown?
I'm not very keen on the Shiny Happy Evil Empire<tm>, either, as it
turns out. ;->
More information about the conspire