I think your idea is an excellent one. I originally migrated J to the Android platform as a side project. It took a little bit of source code tweaking to get everything compiled and linked and I did have to make some tradeoffs in the implementation of some foreigns. I was fortunate in that the J code base is already well structured for cross platform development and I benefitted from previous work for the arm. Once the preponderance of unit tests succeeded, there were some minor adjustments in a few of the libraries and the startup code but I got an interactive console which could render pdfs from 'plot' and display them and a similar capability for bitmapped 'view'.
Chris Burke and Bill Lam were extremely supportive, giving me repo access and providing arm-linux builds of the libraries for an environment that only supported zips. I gave Eric an early demonstration in a downtown pub in Toronto using a laptop emulator and ultimately jsoftware adopted the project. It was not long before Bill had fixed a dozen subtle mostly floating point bugs but the configurations seemed to require a more advanced setup than my own. I do a fair bit of J-ing on my phone and I can't tell you how delighted I am that I don't have to support it. Thanks guys! On Sun, Nov 29, 2020, 14:47 emacstheviking <[email protected]> wrote: > Greg, > the 'vision' would be to be able to not only create J code that runs in the > browser, but one that has full DOM, Canvas, WebGL capability etc. For the > web assembly version I don't think that the current set of "!" features > need all be present especially if the JS interface was fluid enough and > powerful enough. > > I am def. going to start looking into this as a retirement project. I > should be so lucky... > > > > On Sun, 29 Nov 2020 at 16:48, greg heil <[email protected]> wrote: > > > >>no, favoring emacs over vi is the most insane thought > > > > >Now, now no editor wars on Jprograming. Personally i prefer QED ... but > i > > have not seen an implementation since QUED. Maybe Joey can update us... > in > > JCHAT! > > > > >As to a web assembly version... a great idea! Especially if one can > bring > > the full capabilities of the native Javascript through. It should be > > maintained by > > jsoftware.com alongside of their releases for the likes of Android, Pi > etc > > > > >Joe Bogner has a great sifting to Javascript with Emscriptem and said he > > was interested in a WA version. Hopefully he can chime in with updates. > > > > ~greg > > https://picsrp.github.io > > > > -- > > > > from: Hauke Rehr <[email protected]> > > to: [email protected] > > date: Nov 29, 2020, 5:16 AM > > subject: Re: [Jprogramming] J and Web Assembly > > > > >no, favoring emacs over vi is the most insane thought > > > > -- > > > > from: emacstheviking <[email protected]> > > to: [email protected] > > date: Nov 29, 2020, 3:56 AM > > subject: [Jprogramming] J and Web Assembly > > mailing list: Programming forum > > > > > > >Anybody even tried this? > > > > >I have spent an hour reading the web assembly docs and I don't see why > it > > couldn't be done apart from the !: commands which could be problematic. > > > > >I'm seeing a 'core' of the J language that provides basic language > > capability and then uses the WA interface to providea JS way to call > JInit > > and JDo. > > > > >Mad idea? > > > > >I am def. going to investigate more over the coming weeks. I have a > > custom build of J on my iMac but I think it used gcc rather than clang > > which is what's required by the LLVM translation process. > > > > >This could be my most insane thought this weekend.... > > ---------------------------------------------------------------------- > > For information about J forums see http://www.jsoftware.com/forums.htm > > > ---------------------------------------------------------------------- > For information about J forums see http://www.jsoftware.com/forums.htm > ---------------------------------------------------------------------- For information about J forums see http://www.jsoftware.com/forums.htm
