Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The End of Oz

lareviewofbooks.org

Post reply on HN
Print view View original

11–14 of 14 posts

Re: The End of Oz

#11

Post by mncharity »

kwhitefoot wrote: post #8
mncharity wrote: post #2

For anyone else who wondered, Mozart/Oz 2 has seen few commits this last year. And no new releases. It's still missing constraints and distributed objects. There was some work[1] porting 1.3 to 64-bit a year ago, but it's not quickly clear to me that it was ever merged. And I still don't know of a page to point people at, which conveys the wonderfulness that was Oz. Call it an extensible language, distributed, merged…

Is there a document anywhere that gives a short description of the language and a reason for looking into it? The official site is most unwelcoming to a casual reader.

> Is there a document anywhere that gives a short description of the language and a reason for looking into it?

Not that I know of. And there now seems little reason for looking into it, except when doing language research.

EDIT: The SICP-like textbook is apparently online. It's not short, but FWIW: "Concepts, Techniques, and Models of Computer Programming" http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.102...

> The official site is most unwelcoming to a casual reader.

Mozart/Oz documentation... sigh. Think: "bad old days"; "just buy the tome textbook"; that textbook being online (discouraging other docs), and then not (so then there's nothing); academic project; majority of users were students taking mandatory CS courses; in European academia; research papers and talk slides as documentation; and now zombie/death. Years ago I ended up generating my own "virtual manual"[1], but the links have rotted since.

[1] http://www.vendian.org/oz/wiki/index.cgi?VirtualManual But the links are rotted. And it's no longer a wiki, just a snapshot of that page.

Re: The End of Oz

#12

Post by mncharity »

nextos wrote: post #5
pubby wrote: post #3

Earlier quoted context omitted.

Ha, I don't think you realized what the article was about.

I came to see the comments thinking about Mozart/Oz, without noticing the domain. Still, it's very sad that Mozart/Oz gets no attention. IMHO, the associated book CTM is on par with SICP. I wish someone ported all their ideas to a Lisp, if that made development more lively.

> I came to see the comments thinking about Mozart/Oz, without noticing the domain.

And I.

> no attention [...] CTM is on par with SICP.

It looks like CTM is available online at the moment: "Concepts, Techniques, and Models of Computer Programming" http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.102...

I don't recall what year the SICP pdf became free for download from MIT Press. But while CTM drafts were available during writing, they vanished upon publication. Which given the poor state of other Mozart/Oz documentation, was an adoption barrier.

Scheme also seemed to have a broader base of support, with ties into the broader lisp and programming communities. M/Oz seemed more a European academic instructional-and-a-bit-of-research VM/language. A very different cultural context. And broader societal open-source culture and infrastructure was less developed back then, so there was less to fall back on.

I'm reminded of Poplog, a polylingual system of Common Lisp, Prolog, ML, and a C-like language, out of Sussex. But the culture from which the authors came, and the economic choices they made, took something with seemingly great potential, and reduced it to near-zero impact. EDIT: Looks like it's again ported to 64-bit as of this summer, FWIW: http://www.cs.bham.ac.uk/research/projects/poplog/freepoplog...

Also perhaps, while Scheme was a lisp, Oz was weird in its own unique way, so it hit "eww, parentheses"-like resistance without sanctuary. Which I fuzzily recall motivating a second language for the Mozart VM, dividing attention.

> ported all their ideas to a Lisp

Some ideas from the Oz language might be ported perhaps. But the underlying Mozart VM, which provided the unification, constraints, and distributed programming... hmm. Are any of the lisp implementations of prolog for instance, at all performance competitive?

Re: The End of Oz

#13

Post by nextos »

mncharity wrote: post #12
nextos wrote: post #5

Earlier quoted context omitted.

I came to see the comments thinking about Mozart/Oz, without noticing the domain. Still, it's very sad that Mozart/Oz gets no attention. IMHO, the associated book CTM is on par with SICP. I wish someone ported all their ideas to a Lisp, if that made development more lively.

> I came to see the comments thinking about Mozart/Oz, without noticing the domain. And I. > no attention [...] CTM is on par with SICP. It looks like CTM is available online at the moment: "Concepts, Techniques, and Models of Computer Programming" http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.102... I don't recall what year the SICP pdf became free for download from MIT Press. But while CTM drafts were av…

I think CTM drafts were always available online, even after publication.

Re: The End of Oz

#14

Post by mncharity »

nextos wrote: post #13
mncharity wrote: post #12

Earlier quoted context omitted.

> I came to see the comments thinking about Mozart/Oz, without noticing the domain. And I. > no attention [...] CTM is on par with SICP. It looks like CTM is available online at the moment: "Concepts, Techniques, and Models of Computer Programming" http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.102... I don't recall what year the SICP pdf became free for download from MIT Press. But while CTM drafts were av…

I think CTM drafts were always available online, even after publication.

At least circa 2004, while working on improving Mozart/Oz online documentation, I did not know of any such. The drafts previously available, had all been taken down. Shrug.
Post reply on HN