On Jan 22, 2008 5:41 AM, mhampton <[EMAIL PROTECTED]> wrote:
>
> I agree - an optional package makes more sense for the moment. The
> spkg is less than a mb, fortunately, so adding it to the standard
> packages eventually wouldn't inflate the total size that much.
I propose the following:
(1) Json support is made an optional package
(2) Once there are some actual interesting uses of it, then we
seriously consider making it a standard package. (This could
be a week from now, etc.)
Regarding 2, I never remember being convinced to put anything
into Sage unless there was a really cool application that _used_
it. E.g., I didn't want to put libgd in sage and wouldn't until Tom
Boothby showed off a bunch of very high resolution fractal images
he could render very quickly using libgd, which matplotlib sucked at.
That was pretty convincing.
> There are several optional packages that I use a lot and I hope to
> eventually have in sage as standard (biopython, polymake, phcpack) but
> I am waiting to make my final case (for different reasons for each
> package, but that's another thread I guess).
>
> I think this could be an exciting way to get all the java applet
> makers out there interested in sage, although I don't completely
> understand the architecture of what this is supposed to do.
>
> Cheers,
> Marshall Hampton
>
> On Jan 22, 6:23 am, "David Joyner" <[EMAIL PROTECTED]> wrote:
>
> > On Jan 22, 2008 5:05 AM, Ted Kosan <[EMAIL PROTECTED]> wrote:
> >
> >
> >
> >
> >
> > > William wrote:
> >
> > > > > If further testing is successful, I would like to have simpleJSON
> > > > > included in SAGE. What procedure do I need to follow in order to make
> > > > > an official software addition request?
> >
> > > > (1) Convince us it's a good idea. You basically just did that.
> >
> > > > (2) Create a trac ticket and then to attach some code that
> > > > we can try it out.
> >
> > > A simpleJSON spkg has now been created and more information about it
> > > can be found in its trac ticket:
> >
> > >http://www.sagetrac.org/sage_trac/ticket/1510
> >
> > > William has added the following comment to this ticket which requests
> > > that the spkg be discussed on sage-devel before deciding on whether or
> > > not to include it in Sage:
> >
> > > >I think some sort of general voting and discussion should occur
> > > before including
> > > >any new packages standard in Sage, especially ones that don't cover
> > > some very clear
> > > >mathematical area that is completely unconvered by Sage (e.g., R and
> > > >PolyBori?
> > > >did address a clear mathematical area). In particular, it is
> > > _critical_ that there be
> > > >more than one person who really wants the package to go in before we
> > > even consider
> > > >putting it in. I suggest that:
> >
> > > > 1. simplejson be made an optional package,
> > > > 2. there be a post to sage-devel to start some discussion about
> > > > whether this actually
> > > >belongs in Sage. That it is is easy to put in
> > > >Sage (it's pure python) is a plus, but is
> > > >definitely not enough of an argument (to put it mildly).
> >
> > > >Remember, every package that goes into Sage will cause Michael
> > > Abshoff, and me,
> > > >and others headaches at some point, and will add
> > > >to the horrendous problem we
> > > >already have with packages getting out of date with upstream.
> >
> > > >Also, perhaps there should be somebody -- probably Ted in this case
> > > -- who very clear
> > > >volunteers to keep the package up to date for the next year, and find
> > > somebody to take
> > > >over if they can't continue.
> >
> > > >The above was quick brainstorming. It is not meant to be some well
> > > thought out procedure,
> > > >which is something I don't think we have yet. -- William
> >
> > > The purpose for adding simpleJSON to Sage is to give clients an
> > > object-based, language-neutral method for communicating with the Sage
> > > server. This package is not about adding mathematical capabilities to
> > > Sage, but rather, it is for allowing clients to more easily access the
> > > mathematical functionality it already has. I am specifically using
> > > simpleJSON to allow applets, which are loaded from a separate server
> > > into a worksheet, to pull data out of Sage. Allowing applets to be
> > > loaded from a separate server removes the need for them to be included
> > > with Sage itself which I think is a good route to follow.
> >
> > > I have estimated that there are probably hundreds of math, science,
> > > and engineering-oriented Java applets in existence that are capable of
> > > adding value to Sage. My goal with these applets is to not only make
> > > them available in the notebook, but to also attract a significant
> > > number of their developers to the Sage project so that they can add
> > > even more value to it.
> >
> > > If there is no easy way for these applets to communicate with the Sage
> > > server, however, there is not much point in embedding them in a
> > > worksheet.
> >
> > > As for the idea of making simpleJSON an optional package, my thought
> > > is that there is little point in this. If a newbie Windows user wants
> > > to use a calculator applet in a worksheet because this is the level of
> > > technology they are comfortable with, they are probably not going to
> > > be very successful installing an optional package.
> >
> > I can see you don't like the idea of making JSON merely an optional
> > package but one possibly way to hopefully help your proposal would be to
> > first make an optional package and then see how that goes. I think for
> > William and Michael, the effort required to change an optional package
> > to a standard package isn't great. On the other hand, it does make it
> > easier for people to test (and even for you to create bug fixes) for
> > optional packages. You could also add some details about it to the
> > wiki athttp://wiki.sagemath.org/optional_packages_available_for_SAGE
>
> > So, I agree with William.
> >
> >
> >
> > > Anyway, I researched a number of alternatives before choosing a
> > > JSON-based solution to this client communications problem. It is
> > > relatively clean and it seems to work fairly well. If people have
> > > alternative ways to solve this problem, however, I would like to hear
> > > them :-)
> >
> > > Ted
> >
>
--
William Stein
Associate Professor of Mathematics
University of Washington
http://wstein.org
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to sage-devel@googlegroups.com
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---
