[Ironpython-users] numpy implementation [was Re: Community Meeting]

[Slide]

Moving this discussion on numpy into its own thread so searching for it
will be easier for folks in the future.

On Wed, Jul 13, 2016 at 8:59 AM Uncle Meat <3000.sjg at gmail.com> wrote:

> OK, I was able to track down some source code for the NumpyDotNet project.
> It looks like some people did a lot of work on this:
> https://github.com/numpy/numpy-refactor
> I'm not sure what state it's in or even if it builds. I'm not sure who did
> the .NET portion (I think it was Enthought) but it looks like the
> "refactor" branch includes/inherits the numpy license.
>
> Steve
>
>
> On Wed, Jul 13, 2016 at 8:36 AM, Uncle Meat <3000.sjg at gmail.com> wrote:
>
>> You can still get the install binaries and Python code here:
>> https://store.enthought.com/repo/.iron/ but you have to create a login
>> at the Enthought site. Looks like the key to this is the NumpyDotNet dlls
>> that I cant locate. Maybe we can ask Enthought if they'd like to open
>> source it (or maybe this is just a dead end)?
>>
>> Steve
>>
>> On Wed, Jul 13, 2016 at 1:59 AM, Markus Schaber <m.schaber at codesys.com>
>> wrote:
>>
>>> Hi,
>>>
>>>
>>>
>>> At a first glance, IronLab looks nice.
>>>
>>>
>>>
>>> The License is LGPL, so I think it cannot be part of the main IronPython
>>> distribution.
>>>
>>>
>>>
>>> OTOH, they’re using “NumPy and SciPY for IronPython / .NET” which seems
>>> abandoned, but used to be maintained by the PTVS people:
>>>
>>>
>>> http://pytools.codeplex.com/wikipage?title=NumPy%20and%20SciPy%20for%20.Net
>>>
>>> However, I could not locate the source repository of those libraries…
>>>
>>>
>>>
>>>
>>>
>>> Btw, I found a serious drawback on my previous suggestion to use
>>> C++/CLI: It’s hardly platform independent that way, as there’s no C++/CLI
>>> Compiler (yet) which works on non-windows platforms.
>>>
>>>
>>>
>>>
>>>
>>> Best regards
>>>
>>> Markus Schaber
>>>
>>> *CODESYS®* a trademark of 3S-Smart Software Solutions GmbH
>>>
>>> *Inspiring Automation Solutions *
>>> ------------------------------
>>>
>>> 3S-Smart Software Solutions GmbH
>>> Dipl.-Inf. Markus Schaber | Product Development Core Technology
>>> Memminger Str. 151 | 87439 Kempten | Germany
>>> Tel. +49-831-54031-979 | Fax +49-831-54031-50
>>>
>>> E-Mail: m.schaber at codesys.com | Web: codesys.com
>>> <http://www.codesys.com> | CODESYS store: store.codesys.com
>>> CODESYS forum: forum.codesys.com
>>>
>>> *Managing Directors: Dipl.Inf. Dieter Hess, Dipl.Inf. Manfred Werner* | *Trade
>>> register: Kempten HRB 6186* | *Tax ID No.: DE 167014915*
>>> * ------------------------------ *
>>>
>>>
>>>
>>> *This e-mail may contain confidential and/or privileged information. If
>>> you are not the intended recipient (or have received this e-mail in error)
>>> please notify the sender immediately and destroy this e-mail. Any
>>> unauthorised copying, disclosure or distribution of the material in this
>>> e-mail is strictly forbidden.*
>>>
>>> *From:* Uncle Meat [mailto:3000.sjg at gmail.com]
>>> *Sent:* Tuesday, July 12, 2016 6:32 PM
>>> *To:* Markus Schaber
>>> *Cc:* Haymo Kutschbach; Tim Orling; ironpython-users at python.org
>>>
>>> *Subject:* Re: [Ironpython-users] Community Meeting
>>>
>>>
>>>
>>> A project exists called IronLab (https://github.com/rwg0/ironlab) that
>>> works with both numpy and scipy. Maybe we can leverage what was done in
>>> IronLab to support numpy? I think it uses the code developed by Enthought,
>>> not sure if it's OSS.
>>>
>>>
>>>
>>> Thanks,
>>>
>>> Steve
>>>
>>>
>>>
>>> On Tue, Jul 12, 2016 at 11:56 AM, Markus Schaber <m.schaber at codesys.com>
>>> wrote:
>>>
>>> Hi,
>>>
>>>
>>>
>>> Another way might be to use C++/CLI and the NumPy sources to directly
>>> write a .NET wrapper for NumPy, with the API optimized for the IronPython
>>> use case.
>>>
>>>
>>>
>>> Might be more work, but AFAICS, should give the best results
>>> (performance-wise).
>>>
>>>
>>>
>>> Best regards
>>>
>>> Markus Schaber
>>>
>>> *CODESYS®* a trademark of 3S-Smart Software Solutions GmbH
>>>
>>> *Inspiring Automation Solutions *
>>> ------------------------------
>>>
>>> 3S-Smart Software Solutions GmbH
>>> Dipl.-Inf. Markus Schaber | Product Development Core Technology
>>> Memminger Str. 151 | 87439 Kempten | Germany
>>> Tel. +49-831-54031-979 | Fax +49-831-54031-50
>>>
>>> E-Mail: m.schaber at codesys.com | Web: codesys.com
>>> <http://www.codesys.com> | CODESYS store: store.codesys.com
>>> CODESYS forum: forum.codesys.com
>>>
>>> *Managing Directors: Dipl.Inf. Dieter Hess, Dipl.Inf. Manfred Werner* | *Trade
>>> register: Kempten HRB 6186* | *Tax ID No.: DE 167014915*
>>> * ------------------------------ *
>>>
>>>
>>>
>>> *This e-mail may contain confidential and/or privileged information. If
>>> you are not the intended recipient (or have received this e-mail in error)
>>> please notify the sender immediately and destroy this e-mail. Any
>>> unauthorised copying, disclosure or distribution of the material in this
>>> e-mail is strictly forbidden.*
>>>
>>> *From:* Ironpython-users [mailto:ironpython-users-bounces+m.schaber=
>>> codesys.com at python.org] *On Behalf Of *Haymo Kutschbach
>>> *Sent:* Tuesday, July 12, 2016 5:39 PM
>>> *To:* Tim Orling; Uncle Meat
>>> *Cc:* ironpython-users at python.org
>>> *Subject:* Re: [Ironpython-users] Community Meeting
>>>
>>>
>>>
>>> Thanks Tim,
>>>
>>> Ironclad looks interesting. Will investigate it. Thanks for the
>>> reference!
>>>
>>> The promising thing regarding efforts: most, if not all of it has been
>>> done already in our Computing Engine. In fact people maintain wrappers
>>> around the (Matlab'isch) ILNumerics which mimic the numpy API and tend to
>>> be rather thin. The effort therefore would establish in getting rid of such
>>> wrappers and to handle the internal storage differences directly in
>>> ILNumerics. But of course, this needs to be done also...
>>>
>>> Best
>>> Haymo
>>>
>>> On July 11, 2016 10:23:28 PM Tim Orling <ticotimo at gmail.com> wrote:
>>>
>>>
>>>
>>>
>>>
>>> On Mon, Jul 11, 2016 at 6:56 AM, Uncle Meat <3000.sjg at gmail.com> wrote:
>>>
>>> Hi Haymo,
>>>
>>>
>>>
>>> I am very interested in having IPy support Numpy. I would be willing to
>>> support the effort if you need some help.
>>>
>>>
>>>
>>>
>>>
>>> Numpy is a problem because of the amount of C extensions. The best hope,
>>> IMHO, is ironclad which marshalls objects around. Pawel and I spent a bit
>>> of time getting a clean build of ironclad, but whether it was actually
>>> functional or not remains to be seen. We were going for clean builds as a
>>> first step. Actually getting Numpy to be .NET native is going to be a LOT
>>> of work.
>>>
>>> https://github.com/IronLanguages/ironclad
>>>
>>> Unfortunately, I have changed employers and am not that focused on .NET
>>> any more. My main interest these days would be to see IronPython3 moving
>>> forward (out of fondness for the platform).
>>>
>>> Cheers,
>>>
>>> --Tim (Intel Open Source Technology Center)
>>>
>>>
>>>
>>> Thanks,
>>>
>>> Steve
>>>
>>>
>>>
>>> On Mon, Jul 11, 2016 at 7:10 AM, Haymo Kutschbach <
>>> h.kutschbach at ilnumerics.net> wrote:
>>>
>>> Hi Alex and all!
>>>
>>>
>>>
>>> Thanks for the interesting news and for all the work done so far! There
>>> are considerations at ILNumerics to add numpy support to IronPython. So I
>>> would like to closely follow plans and progress of the main project. If
>>> there is a chance to attend the upcoming meeting I would like to
>>> participate as well.
>>>
>>>
>>>
>>> My timezone is CEST / Berlin.
>>>
>>>
>>>
>>> Best regards
>>> Haymo
>>>
>>>
>>>
>>> --
>>>
>>> -------------------------
>>>
>>> Haymo Kutschbach
>>>
>>> ILNumerics GmbH
>>>
>>>
>>>
>>> h.kutschbach at ilnumerics.net
>>>
>>>
>>>
>>> ILNumerics GmbH
>>>
>>> Danziger Str. 219
>>>
>>> 10407 Berlin
>>>
>>>
>>>
>>> Tel: +49 30 4208 7799
>>>
>>> Fax: +49 30 4208 7775
>>>
>>>
>>>
>>> http://www.ilnumerics.net
>>>
>>> http://twitter.com/ilnumerics
>>>
>>>
>>>
>>> Registergericht: Amtsgericht Potsdam
>>>
>>> Registernummer: HRB 26715 P
>>>
>>> Geschäftsführung: Haymo Kutschbach
>>>
>>>
>>>
>>> *From:* Ironpython-users [mailto:ironpython-users-bounces+h.kutschbach=
>>> ilnumerics.net at python.org] *On Behalf Of *Slide
>>> *Sent:* Montag, 11. Juli 2016 07:02
>>> *To:* ironpython-users at python.org
>>> *Subject:* [Ironpython-users] Community Meeting
>>>
>>>
>>>
>>> Hello IronPythonistas!
>>>
>>>
>>>
>>> Jeff Hardy has handed over the mangement of IronPython to myself and
>>> Benedikt Eggers. We are hoping to recruit some new people to contribute to
>>> IronPython and get some releases out. We would like to have a community
>>> meeting to discuss future direction for IronPython. We're planning on
>>> holding the meeting on Gitter, you can find the room here:
>>> https://gitter.im/IronLanguages/main.
>>>
>>>
>>>
>>> I would like to cover the following agenda topics, but would definitely
>>> like to hear from community members what they would like to discuss as
>>> well.
>>>
>>>
>>>
>>> - IronRuby - What do we want to do here?
>>>
>>> - Issue disposition - there are currently about 961 issues, with most
>>> having been ported from CodePlex.
>>>
>>> - Release schedule - I would like to do more rolling releases, getting
>>> features and fixes out faster, but still maintaining stability
>>>
>>> - Continuous Integration - I was, for some time a contributor on the
>>> Jenkins project, I am approaching CloudBees to see what might be available
>>> for open source projects
>>>
>>> - IronPython 3 - Roadmap to IronPython 3, which would be Python 3
>>> compatible.
>>>
>>>
>>>
>>> Please reply back if you are interested in attending the meeting and
>>> specify your timezone. I'd like to make the meeting as easy to attend as
>>> possible for the most amount of people. We won't be able to get everyone,
>>> but I would like to get as many as possible. I would like to hold the
>>> meeting this Thursday.
>>>
>>>
>>>
>>> Thanks!
>>>
>>>
>>>
>>> Alex Earl
>>>
>>> (slide)
>>>
>>>
>>>
>>> _______________________________________________
>>> Ironpython-users mailing list
>>> Ironpython-users at python.org
>>> https://mail.python.org/mailman/listinfo/ironpython-users
>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> Ironpython-users mailing list
>>> Ironpython-users at python.org
>>> https://mail.python.org/mailman/listinfo/ironpython-users
>>>
>>>
>>>
>>>
>>>
>>
>>
> _______________________________________________
> Ironpython-users mailing list
> Ironpython-users at python.org
> https://mail.python.org/mailman/listinfo/ironpython-users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/ironpython-users/attachments/20160713/d689071d/attachment.html>