Qi is now an open source project on http://code.google.com/p/qilang/
for people wanting to contribute to Qi's development. After some
long discussions we've sorted out the implications
of the GPL license that has been the main arguing point for Qi since
its inception in 2005. (see
http://groups.google.co.uk/group/comp.lang.lisp/browse_frm/thread/d2c922ed5826fac4/fd3c4569c9eaaa8e?lnk=gst&q=Qi&rnum=1&hl=en#fd3c4569c9eaaa8e
for the original thread) and the relevant web page for viewing Qi and
the GPL is http://www.lambdassociates.org/licence.htm.
all the best
Mark (back from India)
On Apr 3, 12:29 am, "Mark Tarver" <··········@ukonline.co.uk> wrote:
> Qi is now an open source project onhttp://code.google.com/p/qilang/
> for people wanting to contribute to Qi's development. After some
> long discussions we've sorted out the implications
> of the GPL license that has been the main arguing point for Qi since
> its inception in 2005. (seehttp://groups.google.co.uk/group/comp.lang.lisp/browse_frm/thread/d2c...
> for the original thread) and the relevant web page for viewing Qi and
> the GPL ishttp://www.lambdassociates.org/licence.htm.
Neato.
The last sentence of the home page says:
"The type discipline of Qi is based on sequent notation which is
a much more powerful and flexible tool for defining types than that
used in ML or Haskell. To find out why click here."
Except "click here" is not a link. I clicked on the most interesting
part and nothing happened!
Marshall
> On Apr 3, 12:29 am, "Mark Tarver" <··········@ukonline.co.uk> wrote:
>
> > Qi is now an open source project onhttp://code.google.com/p/qilang/
> > for people wanting to contribute to Qi's development. After some
> > long discussions we've sorted out the implications
> > of the GPL license that has been the main arguing point for Qi since
> > its inception in 2005. (seehttp://groups.google.co.uk/group/comp.lang.lisp/browse_frm/thread/d2c...
> > for the original thread) and the relevant web page for viewing Qi and
> > the GPL ishttp://www.lambdassociates.org/licence.htm.
On 3 Apr, 16:45, "Marshall" <···············@gmail.com> wrote:
>
> Neato.
>
> The last sentence of the home page says:
>
> "The type discipline of Qi is based on sequent notation which is
> a much more powerful and flexible tool for defining types than that
> used in ML or Haskell. To find out why click here."
>
> Except "click here" is not a link. I clicked on the most interesting
> part and nothing happened!
>
> Marshall
I think you are looking at a different page - not the home page. The
sentence you quote is on http://www.lambdassociates.org/aboutqi.htm.
But the link does work - it goes to
http://www.lambdassociates.org/advtypes.htm
That was the first study I wrote to illustrate the power of Qi.
But a better study might be
http://www.lambdassociates.org/studies/study02.htm
called the 10 cent Lisp type checker. There are a fair number of
studies
in http://www.lambdassociates.org/studies.htm which are resistant to
ML and Haskell - http://www.lambdassociates.org/studies/study07.htm
(which is about computer algebra) is another worth looking at.
Mark
Aha. Right - ambiguity. I thought Marshall was referring to my web
site in which the link does work. He was in fact referring to the
open source Google page.
Actually I'm not in charge of that! Entropyfails runs that and he's
probably pasted the text from my site and forgot to carry over the
link. I'll tell him and he'll fix it.
Mark
On 3 Apr, 19:55, "Matthias Benkard" <··········@gmail.com> wrote:
> Hi,
>
> > I think you are looking at a different page - not the home page.
>
> http://code.google.com/p/qilang/
>
> Matthias
I would be very interested in Qi.
Unfortunately looks like it runs only on CLISP.
Any plans to make it run on SBCL on lispworks ?
Currently i'm using these 2 lisps.
On 3 Apr, 23:26, "Vagif Verdi" <···········@gmail.com> wrote:
> I would be very interested in Qi.
> Unfortunately looks like it runs only on CLISP.
> Any plans to make it run on SBCL on lispworks ?
> Currently i'm using these 2 lisps.
It does run under CMUCL as well. Yes, eventually it will
run on SBCL and whatever version of CL we can find - ditto
for the Mac OS X which somebody has asked for.
Porting to another Lisp is not a priority for me right now.
I can't keep up with all the combinations of Lisp and OS out there! :)
For this reason principally Qi was made open source. So I think this
is a good request for the people running the open source to port Qi to
SBCL. I'll put in a request.
Mark
> Porting to another Lisp is not a priority for me right now.
> I can't keep up with all the combinations of Lisp and OS out there! :)
install.txt tweak for Mac OS/X CMUCL PPC
http://code.google.com/p/qilang/issues/detail?id=1&can=2&q=
also other CLs might not auto :USE '("COMMON-LISP")
in MAKE-PACKAGE "qi"
> For this reason principally Qi was made open source. So I think this
> is a good request for the people running the open source to port Qi to
> SBCL.
There are only 4 cond reader blocks,
the 1st one mentioned here, looks like the hard one? -
http://programmingkungfuqi.blogspot.com/2007/03/tour-of-qi-72-and-62-diff.html
-
I was using it, at the moment, to explore Prolog...
When I cut the Qi Prolog from between the quotes in
the defprolog of Einstein.qi and fed it to SWI-Prolog,
it reported meeding only half as many assertions.
Is this a difference between a AUM and WAM machines,
tuning on SWI's part (memoize?), or is Qi counting twice ?
On 4 Apr, 22:54, ····@yahoo.com wrote:
> > Porting to another Lisp is not a priority for me right now.
> > I can't keep up with all the combinations of Lisp and OS out there! :)
>
> install.txt tweak for Mac OS/X CMUCL PPC
>
> http://code.google.com/p/qilang/issues/detail?id=1&can=2&q=
>
> also other CLs might not auto :USE '("COMMON-LISP")
> in MAKE-PACKAGE "qi"
>
> > For this reason principally Qi was made open source. So I think this
> > is a good request for the people running the open source to port Qi to
> > SBCL.
>
> There are only 4 cond reader blocks,
> the 1st one mentioned here, looks like the hard one? -http://programmingkungfuqi.blogspot.com/2007/03/tour-of-qi-72-and-62-...
>
> -
>
> I was using it, at the moment, to explore Prolog...
> When I cut the Qi Prolog from between the quotes in
> the defprolog of Einstein.qi and fed it to SWI-Prolog,
> it reported meeding only half as many assertions.
>
> Is this a difference between a AUM and WAM machines,
> tuning on SWI's part (memoize?), or is Qi counting twice ?
Assertions - do you mean inferences?
Depends I think on how you count logical inferences. Qi counts
every attempted inference while some implementations count
only successful inferences. I recall from memory that Christopher
Hogger discussed this ambiguity somewhere. A Quintus benchmark
says "The measure of logical inferences per second (Lips) used
here is taken to be procedure calls per second"
You can read a Qi logical inference as a procedure (i.e. function)
call. So 600
KLIPS means in Qi 600,000 procedure (function) calls per second which
is about the maximum you get if you run it under CMUCL on a bog
standard
desktop PC.
Mark