On Sunday, 15 November 2015 at 04:19:21 UTC, Lionello Lunesu wrote:
> On 09/11/15 04:38, Richard Davies wrote:
>
> Yeah, I got curious too. I spend some time on it yesterday and had a stab at writing it in D.

Hi. I send a email to John Gustafson yesterday, re this thread.

He replied as follows:

"There are efforts underway worldwide to support unums in C, Python, Julia, Java, and now D. And the book has only been out for nine months!"

Nick

On 17/11/15 07:52, Nick_B wrote:
> On Sunday, 15 November 2015 at 04:19:21 UTC, Lionello Lunesu wrote:
>> On 09/11/15 04:38, Richard Davies wrote:
>>
>> Yeah, I got curious too. I spend some time on it yesterday and had a
>> stab at writing it in D.
>
> Hi. I send a email to John Gustafson yesterday, re this thread.
>
> He replied as follows:
>
> "There are efforts underway worldwide to support unums in C, Python,
> Julia, Java, and now D. And the book has only been out for nine months!"
>
> Nick

Ha, and I think we can do a better job than the others! Surely in terms of performance!

Funny, I was indeed looking at the Julia and Python (and John's Matematica code) to learn how some of the operations on Unums work.

Still a lot to do though. I've basically only declared the type, still needs all the calculations.

L.

Hi

John Gustafson was in town (Wellington, NZ) for the Multicore World Conference 2016 ( http://www.multicoreworld.com/) conference. I caught up with him, tonight, and spoke to him for about two hours. Here is a quick summary of what we discussed.  John has just redesigned Unums, to address the design issues in version 1.0.  He presented his Powerpoint presentation to the conference, with the details of Unums 2.0 (this is a tentative name at present).  Its a improved design, but I will only brief detail it:  "It  will have  more dynamic range with 16-bit values than IEEE half-precision, but only by a small amount. Still remarkable to be uniformly better in dynamic range and precision, with support for inexact values and perfect reciprocation. If a language supports just one unum data type, John believes it should be the 16-bit one".  John has agreed to provide a link to the Powerpoint presentation, in a couple of weeks, and then later, a link to his new published paper on the subject, when it is ready.  There will likely be a new book, building on version 1.0, and, again, tentatively titled 'Unums 2.0'. I also discussed with him, about integrating it with D. At the present, there is a 'C' codebase under construction, but this could be rewritten in D in the future.  D may require some language changes, and a new phobos library, to support this advanced functionality. Of course Walter will have decide if he wants this advanced numbering system as a part of D.

As an aside, John mentioned that Rex Computing (http://www.rexcomputing.com/) is using Unums with the Julia language, for their new hyper-efficient processor architecture. It will be interesting to see what these whiz kids deliver in time.

cheers
Nick

On Wednesday, 17 February 2016 at 08:11:21 UTC, Nick B wrote:
> Hi
>
> John Gustafson was in town (Wellington, NZ) for the Multicore World Conference 2016 ( http://www.multicoreworld.com/) conference. I caught up with him, tonight, and spoke to him for about two hours. Here is a quick summary of what we discussed.  John has just redesigned Unums, to address the design issues in version 1.0.  He presented his Powerpoint presentation to the conference, with the details of Unums 2.0 (this is a tentative name at present).  Its a improved design, but I will only brief detail it:  "It  will have  more dynamic range with 16-bit values than IEEE half-precision, but only by a small amount. Still remarkable to be uniformly better in dynamic range and precision, with support for inexact values and perfect reciprocation. If a language supports just one unum data type, John believes it should be the 16-bit one".  John has agreed to provide a link to the Powerpoint presentation, in a couple of weeks, and then later, a link to his new published paper on the subject, when it is ready.  There will likely be a new book, building on version 1.0, and, again, tentatively titled 'Unums 2.0'. I also discussed with him, about integrating it with D. At the present, there is a 'C' codebase under construction, but this could be rewritten in D in the future.  D may require some language changes, and a new phobos library, to support this advanced functionality. Of course Walter will have decide if he wants this advanced numbering system as a part of D.
>

I would be interested in the Powerpoint when it becomes available.

As for getting it to work in D, I'm not sure how much language changes would be necessary. If they can get it to work in C, then surely it would work in D. After all, the book has an implementation in python (albeit this is not the latest and great version apparently).

Wrt phobos, I would just recommend that whatever unum library gets eventually written has a companion with the equivalent of the functions from std.math.

On Wednesday, 17 February 2016 at 16:35:41 UTC, jmh530 wrote:
> On Wednesday, 17 February 2016 at 08:11:21 UTC, Nick B wrote:

>
> Wrt phobos, I would just recommend that whatever unum library gets eventually written has a companion with the equivalent of the functions from std.math.

Having just looked at the slides again, I believe this will break compatibility with std.math, (for example it throws out NaN), just as D has broken full compatibility with all of C++.

I hope to have a link to the revised presentation within 7 days.

Can anyone tell me who are the maths experts, and hard science users, around here ?


Nick

On Saturday, 20 February 2016 at 23:25:40 UTC, Nick B wrote:
> On Wednesday, 17 February 2016 at 16:35:41 UTC, jmh530 wrote:
>> On Wednesday, 17 February 2016 at 08:11:21 UTC, Nick B wrote:

>
> Having just looked at the slides again, I believe this will break compatibility with std.math, (for example it throws out NaN), just as D has broken full compatibility with all of C++.

UNUM II  is also proposing to break completely from IEEE 754 floats and gain Computation with mathematical rigor .......
>
> Can anyone tell me who are the maths experts, and hard science users, around here ?
>
>
> Nick

If anyone is still interested in this concept whatsoever, we are holding a mini-workshop on the current developments of Unums at the University of California Santa Cruz on Oct 24th.  We'd love to have some participation from interested parties, including presentations on any attempts to implement (in D?) etc.  Please see https://systems.soe.ucsc.edu/2016-symposium or contact me via here.  Nic.


On Saturday, 20 February 2016 at 23:38:52 UTC, Nick B wrote:
> On Saturday, 20 February 2016 at 23:25:40 UTC, Nick B wrote:
>> On Wednesday, 17 February 2016 at 16:35:41 UTC, jmh530 wrote:
>>> On Wednesday, 17 February 2016 at 08:11:21 UTC, Nick B wrote:
>
>>
>> Having just looked at the slides again, I believe this will break compatibility with std.math, (for example it throws out NaN), just as D has broken full compatibility with all of C++.
>
> UNUM II  is also proposing to break completely from IEEE 754 floats and gain Computation with mathematical rigor .......
>>
>> Can anyone tell me who are the maths experts, and hard science users, around here ?
>>
>>
>> Nick

On Tuesday, 20 September 2016 at 22:52:57 UTC, Nic Brummell wrote:
> If anyone is still interested in this concept whatsoever, we are holding a mini-workshop on the current developments of Unums at the University of California Santa Cruz on Oct 24th.  We'd love to have some participation from interested parties, including presentations on any attempts to implement (in D?) etc.  Please see https://systems.soe.ucsc.edu/2016-symposium or contact me via here.  Nic.
>

Nic

Thanks for the heads up.  John Gustafson will have the best understanding as to the progress to implement this in "C" I believe.  Perhaps you could post back an update after the conference ?

Nick

On Tuesday, 20 September 2016 at 22:52:57 UTC, Nic Brummell wrote:
> If anyone is still interested in this concept whatsoever

Now that we've announced we're doing multiplayer games. One of the guys was saying we'd need a fixed-point library. My response was "Why not unums?" Thus, in the very near future I'll be looking at evaluating the available C++ and D libraries for unums to see how suitable they will be for deterministic multiplayer gaming.

Is there some central repository with links to the active projects? I'll try and wrap my head fully around the math before we get to that point though.

On Wednesday, 21 September 2016 at 07:52:18 UTC, Ethan Watson wrote:
> On Tuesday, 20 September 2016 at 22:52:57 UTC, Nic Brummell
>
> Is there some central repository with links to the active projects? I'll try and wrap my head fully around the math before we get to that point though.

Ethan

There is the other thread on this subject. Its called:

» General » Unum II announcement

there is a lot a technical discussion on Unums at:

https://groups.google.com/forum/#!forum/unum-computing

Note that a Skip Cave is even claiming to have UNums running in the J programming language.

You may also want to read John G post at the bottom of this thread Called"The Great Debate: John Gustafson and William Kahan (Video?)"

cheers N.