Le dimanche 30 juin 2024 à 16:45:41 UTC+2, Dima Pasechnik a écrit :
On Sun, Jun 30, 2024 at 4:14 AM Matthias Koeppe <matthia...@gmail.com> wrote: > > Another migration route that we could consider: > Importing the messages from AskBot into Google Groups (merging with the sage-support group). Google Groups' interface is pretty bad, one cannot do maths, or properly format things. This is important for our subjects. : - For code, one can make do with the "Markdown Here" plugin, but - I am not aware of any convenient way to use LaTeX. The only workaround I found was to use emacs' sage-shell's inline typesetting , export the result to PNG and paste that *as a picture* in Google groups' Web interface. Choreish enough to be dissuasive, and sometimes ugly as hell... BTW : if anybody is aware of a way to get typeset snippets in the damn Google Groups interface, I'd be delighted too hear frim him/her... (and who knows for how much longer they keep it free, or support it at all...) I'd rather seriously consider moving to Discourse (which does have an importer from askbot) and is yet another nosy cookie monster... I have to use it for mc-stan (a probabilistic (Bayesian) modeling tool) Q&A, and do not relish it... My two (€)cents... > > I don't visit ask.sagemath.org much, so others who are regulars there may correct me, but it seems that the whole downvoting/upvoting / karma / accepting answers mechanism of this forum may have little value, as most questions have 0 or 1 answers. So if that's lost in the migration, perhaps not much is lost. > > Importing stuff via mailbox files to Google Groups is explained for example in https://geek.co.il/2018/05/16/migrating-google-groups-archives-between-accounts > > On Saturday, June 22, 2024 at 10:18:45 AM UTC-7 Matthias Koeppe wrote: >> >> Reviving this old thread from 2021, as we are currently taking stock of the project infrastructure (discussion in https://groups.google.com/g/sagemath-admins, updating the wiki page https://github.com/sagemath/sage/wiki/Infrastructure). >> >> - server uses Askbot 0.7.59 (a 2020 hotfix of a 2015 version); current: 0.12.2 (2023) >> - https://ask.sagemath.org/questions/ looks at least 10 years out of date re web design >> - the "0 years ago" bug still shows >> >> >> On Wednesday, December 15, 2021 at 9:00:33 AM UTC-8 William Stein wrote: >>> >>> Good idea -- I'll do that. >>> >>> On Wed, Dec 15, 2021 at 8:24 AM Dima Pasechnik <dim...@gmail.com> wrote: >>> > >>> > >>> > >>> > On Wed, 15 Dec 2021, 15:59 William Stein, <wst...@gmail.com> wrote: >>> >> >>> >> On Wed, Dec 15, 2021 at 6:49 AM Dima Pasechnik <dim...@gmail.com> wrote: >>> >> > >>> >> > On Wed, Dec 15, 2021 at 12:44 AM William Stein <wst...@gmail.com> wrote: >>> >> > > >>> >> > > Here's some very quick thoughts, some on topic and some off: >>> >> > > >>> >> > > - The original author of the software that ask.sagemath.org runs did initially do the configuration at some point long ago. He was starting a business to support instances in exchange for money. I wonder if that is still around? >>> >> > >>> >> > contact details, etc are here: https://askbot.com/ >>> >> > (their own instance is running version 0.10.3) >>> >> >>> >> >>> >> Thanks. It looks like the obvious thing we would have to buy would be >>> >> his hosting service, which is $45/month and provides this (see below). >>> >> It seems like he doesn't offer any sort of "manage your existing site" >>> >> service anymore. An obvious question is whether or not he would also >>> >> migrate our existing site over to his hosted service. Also, we would >>> >> be switching from "trust Thierry for backups, long term existence, >>> >> etc." to "trust random guy in Chile who has nothing to do with Sage >>> >> and also trust that William can ensure the $45/month gets paid >>> >> somehow." So it seems he doesn't really provide the service >>> >> (namely -- help us upgrade to the latest version) that he used to >>> >> provide, and I can see why (it is probably a pain in the butt to do). >>> > >>> > >>> > they offer consulting. It won't harm to ask whether they are willing to do an one-off job of upgrading. >>> > >>> > >>> >> >>> >> >>> >> >>> >> 30 day free trial >>> >> All standard features >>> >> SSL Support >>> >> Single Sign On >>> >> Up to 750 users >>> >> Daily backups >>> >> >>> >> >>> >> > >>> >> > >>> >> > >>> >> > > If so, maybe we could pay him to update our ask.sagemath.org instance? In any case, many thanks to Thierry for maintaining ask.sagemath.org all these years! >>> >> > > >>> >> > > - The ziglang community has an interesting approach to the community discussion problem, which is just to welcome and try *everything* that somebody will set up, and see what sticks. They have a list of all the resources that various community members are running here, which is useful for ideas, in case there's interest in expanding what's available for Sage: https://github.com/ziglang/zig/wiki/Community >>> >> > > My impression with zig is that the two main ways they communicate are (1) github issues, and (2) their Discord chat server, which is *very* active. Note that Discord has absolutely nothing to do with Discourse, and solves a completely different problem. We also use Discord for CoCalc, and it has a manageable steady stream of activity (in fact, some discussion there is about Sage). One aspect of Discord is that answers don't seem to get indexed by Google; it's very different than ask.sagemath.org or Discourse. I think the Sage community currently has no *active* general realtime community chat, though we used to all use IRC for Sage very heavily a long time ago, and I think Gitter was also popular at one point. There's the Zulip for Sage here https://www.sagemath.org/help-zulip.html but it's not that active. A point in favor of Discord is that it is what all the kids are currently using to discuss video games, hence they are likely to already be using it, and get notifications, etc; it's also very functional for free (unlike Slack). >>> >> > > >>> >> > > - Github has a free thing that is supposed to compete with Discourse, I think, called "Github Discussions". I tried encouraging people to use Github Discussions for CoCalc (see https://github.com/sagemathinc/cocalc/discussions), e.g., by linking to it in a few places, but it is totally dead. Nobody uses it. I'm not sure why. >>> >> > > >>> >> > > -- William >>> >> > > >>> >> > > On Tue, Dec 14, 2021 at 3:43 PM Thierry < sage-goo...@lma.metelu.net> wrote: >>> >> > >> >>> >> > >> On Sat, Dec 11, 2021 at 07:00:54AM -0800, kcrisman wrote: >>> >> > >> > Some hopefully-useful comments: >>> >> > >> > >>> >> > >> > Overall, I am not sure why the project is running this server. I think we >>> >> > >> > > are spreading ourselves pretty thin here. There is also: >>> >> > >> > > - https://groups.google.com/g/sage-support >>> >> > >> > > - https://stackoverflow.com/questions/tagged/sage >>> >> > >> > > >>> >> > >> > 50 questions in last 6 months - thankfully, not as many about the Sage ERP >>> >> > >> > software as in the past - so nowhere near the volume of ask.sagemath? But >>> >> > >> > see below. >>> >> > >> > >>> >> > >> > > - https://stackexchange.com/search?q=sage >>> >> > >> > > - https://mathoverflow.net/search?q=sage >>> >> > >> > > >>> >> > >> > negligible >>> >> > >> > >>> >> > >> > Also https://math.stackexchange.com/questions/tagged/sagemath - 50 >>> >> > >> > questions in the past 18 months. >>> >> > >> > >>> >> > >> > Though I haven't had time to contribute to any of these regularly since >>> >> > >> > somewhat before the pandemic, my observations are twofold on this: >>> >> > >> > 1) The same people were helping on both the SE sites and ask.sagemath >>> >> > >> > 2) The response time and quality (and number of answers) were all better on >>> >> > >> > ask.sagemath >>> >> > >> > It's possible this has changed recently, but more likely it is just that >>> >> > >> > some of the networking issues have made it less reliable, but haven't >>> >> > >> > improved service on the other sites. Contributors to answers have liked >>> >> > >> > ask.sagemath. >>> >> > >> >>> >> > >> An additional benefit from ask.sagemath is that is it well referenced on >>> >> > >> search engines, so that answers posted there can be used as a knowledge >>> >> > >> database. >>> >> > >> >>> >> > >> > I do notice that ask.sagemath has for some time only reported "0 years >>> >> > >> > ago" for recent questions - I don't think the askbot developer >>> >> > >> > (https://askbot.com) has been super productive lately and of course we >>> >> > >> > don't have the ability to support that software as well as Sage >>> >> > >> > itself. That seems to be more of a sticking point, though I think >>> >> > >> > Thierry and Samuel will have some relevant comments on that. >>> >> > >> >>> >> > >> This "0 years ago" comes from a change in some javascript where, there >>> >> > >> is a jump from hours to years, without knowing about days and months. >>> >> > >> >>> >> > >> Maybe it is fixed on another upstream branch. Note that we are still on >>> >> > >> the 0.7.x branch since migrating requires some work. Indeed, when we >>> >> > >> upgrade, we need to run some migration scripts that update the database >>> >> > >> structure. Unfortunately, our database scheme is not exactly the same as >>> >> > >> the official one since in the early years, ask.sagemath was installed by >>> >> > >> askbot developper himself, who adapted the db to some needs (perhaps >>> >> > >> William can say more about that). We already fixed a lots of such >>> >> > >> things, but when i moved ask.sagemath in emergency last year from Ohio >>> >> > >> State University to Paris North University, i updated the versions until >>> >> > >> i fall on some database migration issue and did not go further. This >>> >> > >> might not be that hard to fix, but it is pretty boring (one have to >>> >> > >> understand where is the bug, do some sql to adapt our tweaked database >>> >> > >> to what it is supposed to be, and try again to see if the migration >>> >> > >> script pass). >>> >> > >> >>> >> > >> If someone from the community is willing to give a try, i could set up a >>> >> > >> separated container, that is a duplicate of the current production >>> >> > >> server and provide an SSH access. >>> >> > >> >>> >> > >> > As for sage-support, it's always surprising to me how many "customers" >>> >> > >> > simply disappear as soon as you mention asking a question there - >>> >> > >> > somehow they either want a Q&A site, or to ask on Facebook, or >>> >> > >> > whatever. I don't really know why this happens, but clearly the Q&A >>> >> > >> > sites (in toto) are necessary. >>> >> > >> >>> >> > >> It is a pity that sage-support is not well referenced by search engines >>> >> > >> since there are some pretty good answers there. If i understand >>> >> > >> correctly, on discourse it is possible to ask and answer questions by >>> >> > >> email, so that people that prefer the sage-support way could contribute >>> >> > >> as well. However, i think that merging ask.sagemath and sage-support >>> >> > >> into some discourse instance and keep all links well redirected will >>> >> > >> require quite some work. >>> >> > >> >>> >> > >> Ciao, >>> >> > >> Thierry >>> >> > >> >>> >> > >> >>> >> > >> > Finally, to anticipate a question that has come up in the past on this >>> >> > >> > issue, I had previously inquired with a SE employee about transferring >>> >> > >> > data, and it seems it's not really a viable option. (As a very tangential >>> >> > >> > note, one could make a similar observation about Trac versus GH/GL - could >>> >> > >> > one take all those very useful discussions and import them?) >>> >> > >> > >>> >> > >> > -- >>> >> > >> > 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+...@googlegroups.com. >>> >> > >> > To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/3ab714cf-5987-4a1d-9f99-471e616e83a8n%40googlegroups.com. >>> >> > >> >>> >> > >> -- >>> >> > >> 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+...@googlegroups.com. >>> >> > >> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/20211214234339.GC5262%40metelu.net. >>> >> > > >>> >> > > >>> >> > > >>> >> > > -- >>> >> > > William (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+...@googlegroups.com. >>> >> > > To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/CACLE5GBNueo5hK04X0s4jcuykF9G3P%2BjG_WdL9Ko0epostaFjg%40mail.gmail.com. >>> >> > >>> >> > -- >>> >> > 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+...@googlegroups.com. >>> >> > To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/CAAWYfq3qUz-eAF1Yg%3DN9NaGsD3VFj4sb6sVp_-PYaCThdtf2Sw%40mail.gmail.com. >>> >> >>> >> >>> >> >>> >> -- >>> >> William (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+...@googlegroups.com. >>> >> To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/CACLE5GAFXyEzWEHWLVEoiMoF-6Hw7unwb0Zd%2Bsi6_DvreCa7WA%40mail.gmail.com. >>> > >>> > -- >>> > 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+...@googlegroups.com. >>> > To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/CAAWYfq3UdJgiSacn-nRjnW_QqPMYKvHOTEu-509gqnobKGwkgw%40mail.gmail.com. >>> >>> >>> >>> -- >>> William (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+...@googlegroups.com. > To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/156bc1f2-d4c5-4b9f-8ce7-cbac89f2d07bn%40googlegroups.com. -- 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 view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/e44884d8-bc2a-4185-9dd1-07fdf0665215n%40googlegroups.com.
