Musing out loud... [Why not Smalltalk?]

[Eric Clayberg]

"Piercarlo Grandi" <pg_nh at sabi.Clara.co.UK> wrote in message
news:yf3hezkihkz.fsf at sabi.ClaraNET.co.UK...
>
> clayberg> Then explain why Sun bought Anamorphic and their HotSpot
> clayberg> technology for *several* tens of millions of dollars. If Sun
> clayberg> already had all this in house, they had no reason to buy
> clayberg> HotSpot.
>
> This argument is based on the premise that the relevant people at Sun
> knew what they were doing. Well, lots of companies buy other companies
> for huge prices without any good reason to do so.

On what basis do you insinuate that they did not have a good reason - a
*very* good reason, in fact - to buy Anamorphic and their HotSpot
technology? My team at ParcPlace-Digitalk (the former developers of
VisualWorks Smalltalk) reviewed HotSpot extensively before Sun bought it. We
would have bought it in a heartbeat had we been able to meet Sun's offer.

> clayberg> Were you around when the Anamorphic team was shopping HotSpot
> clayberg> around to the highest bidder? Did you see their technology in
> clayberg> action?  I think you are severely underestimating the
> clayberg> significance of what the Anamorphic/HotSpot team
> clayberg> developed. Apparently Sun did not...
>
> This is just handwaving, or worse: it sounds like insinuating that I
> have in some way passed judgement on Anamorphic's stuff.

That appears to be exactly what you are doing...or do you routinely
"handwave" about things you know nothing about? Go back and read the
original comment that started this thread:

"Javaites probably know that the Self VM project was where Sun's 'hotspot'
technology was developed."

and Eric L.'s and my replies about Anamorphic. Sun's revisionist history
conveniently leaves out the contribution of the Anamorphic team and the fact
that Sun *bought* them and *bought* HotSpot. The Anamorphic team, BTW, went
on to become the core of Sun's HotSpot team.

> Well, I don't think that I have underestimated anything about
> Anamorphic; I haven't (IIRC) emitted any opinion as to whether their
> stuff worked or not or was significant or not.

It worked amazingly well for Smalltalk and was a very significant
achievement at the time. The techniques pioneered by the Self team
represented only part of what HotSpot did. The Anamorphic team not only
pulled together the earlier Self work but quite a few other ideas that, up
until that point, had only been discussed as possibilities in various papers
(from OOPSLA, ACM, etc.). No one prior to Anamorphic had ever attempted to
pull all of those ideas together.

> I have just stated the easily checked _fact_ that the Self research
> group did themeselves the "pull them all together and prove that they
> worked in concert" act themselves for their adaptive compilation
> technology, and it was quite a good act, so whatever the significance of
> Anamorphic's stuff is, it's not that it was a singular achievement of
> Anamorphic's, contrarily to what is stated here:

"Contrarily to what is stated here"?!? No one is disputing that the work
done by the Self team was quite significant. What I don't understand is why
you seem to want to minimize the work of the Anamorphic team whose work you
state you know very little about.

>   clayberg> [ ... ] Several of the ideas for HotSpot originated with Self
(or
>   clayberg> were well known in the literature), but it took the
>   clayberg> Anamorphic team to pull them all together and prove that
>   clayberg> they worked in concert. [ ... ]

Which was *exactly* what Anamorphic did in fact do. Sun bought Anamorphic
and HotSpot for precisely the reason that it went well beyond the work that
was available to them from any other source. Why do you find that so hard to
believe?

> I know very little about what Anamorphic developed

For knowing very little about it, you seem to have a lot to say about it.

> so I am prepared to
> believe that they did eventually achieve results similar to or perhaps
> even better than those of the Self team, and that they were highly
> significant; but not uniquely significant, as the Self team themselves
> used those ideas and managed to "pull them all together and prove that
> they worked in concert", indeed quite easily, glaringly, publically so.

The work done by the Self team represented a *subset* of the work done by
Anamorphic. There was a lot more to HotSpot than just the ideas poineered by
the Self team. Sun wouldn't have shelled out several tens of millions of
dollars for it otherwise. If all Sun needed was the Self work, they already
had that. Doesn't common sense dictate that they thought they were buying
something quite significant and crucial to what they were trying to achieve
with Java?

> Also, in the case of the Self team I don't have to resort to faith; the
> demonstration that their stuff works, and works well, "all together" and
> "in concert" is well known, published, and can even be downloaded.

Which in no way disminishes the significance of Anamorphics HotSpot work or
its signicance to Sun's Java efforts (both technical *and* marketing).

> perhaps I should have mentioned this before: I have met Mario
>     Wolczko long ago (but haven't seen him for a few years) and quite I
>     liked him and his work

Wow. That certainly lends credibility to your arguments. Should all of us
list off the various experts and industry luminaries that we have met at one
time or another. Hey, I shook Bill Gates hand once. Does that count?