[Date Prev][Date Next][Thread Prev][Thread Next][Author Index][Date Index][Thread Index]

Re: the yearly flurry of email messages...

To contribute to the flurry, a few things that I disremember if they made it out.
A fix to a bug in green :  the bug is an = instead of an == in a for loop
the complete smalltalk image which I don't remember if we had released before.
these are currently in my home (rocket) website http://halfwaytoanywhere.com
at http://halfwaytoanywhere.com/xu
in particular   http://halfwaytoanywhere.com/xu/xanadugold/src/image
has the image stuff for parcplace smalltalk.  

IIRC the patched version runs correctly on 64 bit linux but not Mac, 64 bit int I think.  I don't remember if the ent files are compatible, I may have changed the disk allocation bitmap incompatibly to get past some size limitations.  It hadn't been changed since 1988, and wasn't ready for current sized disks, though I forget where it broke.

Feel free to contact me if you have any problems accessing all this,  bandwidth may be limited, right now I host this at home, if there is more demand, I'll put it somewhere better.

On Feb 10, 2011, at 11:06 AM, John Dougan wrote:

More flurrying...

I've taken the Xanadu email data that is/was at Sunless-Sea and mbox-ified it so it should easily slide into most email programs:

You missed the Abora project, which has used ST in some of it's Xanadu-related efforts:

BTW, whatever docs you manage to get, make sure the curse is removed first. One wonders if a careless use of black magic is the real problem with getting Xanadu working. Anyone know a competent witch who could clean up the bad energies? ;-)

Put me down as another person interested in getting Gold running, then seeing where it can go from there. I would also dearly love to see the internals documentation.

  -- John

On Thu, Feb 10, 2011 at 07:43, John Ohno <john.ohno@xxxxxxxxx> wrote:

Udanax green (xu88) still works out of the box for me. Gold / xu92
(according to the udanax website last I checked) is missing the code
that compiles the pseudo-smalltalk language 'X' to c++; if the 'X'
language documentation is floating around then reimplementing a
compiler should be trivial.

A few months ago I made a tarball of all the current official and
semi-official xanadu websites I'm aware of (udanax.com, xanadu.com,
xanadu australia, hyperland, sunless-sea, and the trans-literary
project). If any documentation is floating around on the web (rather
than being locked up in an XOC member's desk drawer) it's probably in
there. If you like, I can dig it up for you.

On Thu, Feb 10, 2011 at 3:54 AM, Craig Latta <craig@xxxxxxxxxx> wrote:
> Hi Hans-Martin!
>     I'm delighted to contribute to the flurry. :)
>     I have all the extant bits and pieces from back in the day (and
> even the SPARCstation hardware it ran on)... but... yes, the ParcPlace
> stuff is the sticking point. Udanax Gold runs on ParcPlace Objectworks
> Smalltalk circa 1988, and the nominal latest-and-greatest version I got
> from Roger Gregory crashes the Objectworks virtual machine.
>     I need the virtual machine sources so I can debug. I had a
> conversation going about it with James Robertson and Georg Heeg several
> years ago, but it fizzled out. Perhaps you could revive that part of
> things. :)  Of course, this would be with the current appropriate people
> (if any :). James has since left what-used-to-be-ParcPlace for greener
> (golder?) pastures.
> -C
> p.s.
>     Some background for those who are not familiar, or forgot, or have
> blotted it out... ;)  I first heard about Ted and the Xanadu team when I
> was working at ParcPlace in 1992, when Ted came to Stanford to give a
> talk. I read Gary Wolf's infamous Wired article on the project in
> 1995[1]. In 1999, when I was at Interval Research, I met Ted, Roger,
> Mark, and Andrew at the O'Reilly Open Source Conference in Monterey,
> where they announced Udanax Green and Gold.
>     I contacted Roger after that to see about getting Gold running
> again. I came over to his place and we spelunked his garage for old disk
> drives, CPUs, memory, etc. I got as far as watching Objectworks crash on
> a SPARCstation 10 on my dining room table, then joined in the grand
> tradition of Waiting For The Next Victim. :)  Curse you, day jobs!
>     I live in Amsterdam now. The hardware is in a storage space in
> California, so not terribly convenient, but the bits are all right here.
> If Roger (or whomever) doesn't mind I'm happy to share them. Hopefully
> we don't really need the hardware anymore, because there are emulators
> which run a hundred times as fast and don't complicate the situation?
> (Ha ha ha :)
>     I've been a Smalltalk hacker since 1991 at UC Berkeley, first with
> Objectworks 2.5, then with the first few releases of VisualWorks while
> at ParcPlace, then to Squeak in 1996 and thereafter. I've served on the
> Squeak board since... oh, god. ;)  I'm preparing a release for the 20th
> of next month, a Squeak distribution called Spoon[2,3]. It features a
> minimal object memory and virtual machine (the better to learn and
> understand the system), and a live module system (the better to
> collaborate and distribute changes). The module system works with other
> Smalltalks, too (in particular, with VisualWorks). Yes, I know, severe
> error 33 risk[4], but I'd suggest that error 34 (being so afraid of
> error 33 that you don't do anything) is even worse. :)
>     Oh, and I'm available for work. :)
> [1] http://www.wired.com/wired/archive/3.06/xanadu.html
> [2] http://netjam.org/spoon (main site)
> [3] http://thiscontext.wordpress.com (blog)
> [4] http://en.wikipedia.org/wiki/Error_33
> --
> Craig Latta
> www.netjam.org/resume
> + 31  06 2757 7177
> +  1 415  287 3547

John Ohno

John Dougan