On Thu, May 29, 2014 at 7:54 AM, kcrisman <kcris...@gmail.com> wrote: > That said, I think you are referring to software contributions themselves, > and here we come upon something William already stumbled on when he started > psage (though I haven't checked in on that project in a while). A maturing > project necessarily builds different self-governance structures (see Weber, > "The Success of Open Source" and his analysis of several famous Linux > episodes, for example) than when it was early in development, and at this > time there is definitely a procedural barrier that tries to touch on some of > the issues of quality that rjf and others occasionally bring up, but which > makes it harder to get contributions in. (I'll be commenting longer on this > in another thread.)
Update: My conclusion from psage is that it's very important to have easy ways to share in-development code. Using trac (e.g., git branches) is also fine for that. Psage is pointless (or more like the famous combinat queue back then). I didn't stumble upon anything with psage. Unlike some ecosystems (e.g., with Node, general random python packages, etc.), it's much, much better for us to just have one very large repository that gets tested together. All good code should get included into standard Sage. This is because Sage is a mathematics package, and mathematics is incredibly rigid/rigorous/inter-related/etc., much, much more so than some other areas of software development. > As some examples, I would point out: > > * The San Diego State tutorials - nicely done, now open for more > development, but never either fully in Sage nor really fully taken advantage > of. > > * The "Use Sage" idea from Springer, which didn't go anywhere (by consensus > of developers) but also didn't really (as far as I know, I could be > mistaken) pan out in the samizdat Sage world as planned. I very concretely decided not to pursue a "Use Sage" series with Springer due to lack of time (or rather, deciding my time is better spent on SageMathCloud right now). Springer would definitely 100% been ready to run with it otherwise. > [...] but rather from a lack of resources to manage or solicit contributions > from a wide scope of areas, some of which are more peripheral to the software > proper. Having some dedicated editor(s) to ensure patches get properly reviewed in a timely basis, like journals have, would indeed go a long way. And money could buy that. > So the issue is finding ways to promote some of the newer developments in > the ecosystem without coercing, finding random revenue, or the like. SMC is > a notable exception here, possible because William has some dedicated time > set aside for this. Several categories I wish I had time to shepherd are: I have the next year full time set aside for SMC. SMC has the potential to generate *non*-random revenue... > * Curated set of examples; interact.sagemath.org was a great start in this > direction, but sort of petered out. Maybe the interface was just a little too much of a pain. I think wiki.sagemath.org/interact was doing well and not petering out, and was objectively like interact.sagemath.org, but worse. The nbviewer project that IPython has is a successful collection of examples of IPython, with much positive momentum -- we could learn from (and maybe even use) that somehow. > * Having set of people who could go around country (world?) doing low-cost > tutorials for research or teaching. Is this sort of motivated by what Wolfram, inc maybe does? What's the model here? > * Making a high-quality set of screencasts for Sage use. I can't even > count how many people have told me they wanted to do this, but I don't think > there ended up being very many, and they certainly aren't organized in a > nice central way. Compare this to Geogebratube. Can you also not count how many people have told you that they _want_ to watch such screencasts? I ask this, having made nearly 100 screencasts involving Sage this year alone... https://www.youtube.com/user/wstein389/videos I guess they get about 75-125 views on average. I have a group of CS undergrads working with me on SMC now, and they argued that from their point of view screencasts about Sage are a waste of time. I argued for screencasts, but I have to admit that they made my question my assumptions about their value. We agreed that they are useful in some contexts -- e.g., I think this one was useful http://youtu.be/Xj5-WKeecns in that it totally answered somebody's question. However, more technical/programming ones, which would naturally come up with Sage, and more pain to watch than just reading a list of steps. > * Localization. This is harder because of the commands being in English, > but there has been a lot of trouble even getting translations in because not > enough people feel qualified at both the Sage and the language to review > them properly. See e.g. [1]. This is also one of those canonical areas where money can solve problems. For example, if I had funding, I can instantly think of three people who would all be all over doing Russian translations. And of course there is strong French skills in our community. Watching Google analytics, the most useful second language would probably be Spanish, then maybe Chinese. > I'm sure you can think of others. > > > Unfortunately, after seven years or so of working with Sage, I'm not sure I > have any good answer for how to make that happen. Geogebra has a good > model, I think, and to some extent WeBWorK as well, but Sage is primarily > aimed at a different place. Suggestions would be very welcome indeed. Geogebra / WebWork does indeed have a much different audience than Sage... > > > - kcrisman > > > [1] > http://trac.sagemath.org/query?status=needs_info&status=needs_review&status=needs_work&status=new&status=positive_review&component=documentation&summary=~translation&col=id&col=summary&col=component&col=status&col=type&col=priority&col=milestone&order=priority > > -- > You received this message because you are subscribed to the Google Groups > "sage-devel" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to sage-devel+unsubscr...@googlegroups.com. > To post to this group, send email to sage-devel@googlegroups.com. > Visit this group at http://groups.google.com/group/sage-devel. > > For more options, visit https://groups.google.com/d/optout. -- William Stein Professor of Mathematics University of Washington http://wstein.org -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at http://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
