[sage-devel] sage and jupyterhub
My question is possibly stupid, but I could not find a satisfying answer by googling... Can sage work with *jupyterhub*? Or will it be possible? I am planing to build a jupyterhub server for the French mathematician community, and, of course, it would be cool to run sage in it. Yours, t;d. -- 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. attachment: tdumont.vcf
Re: [sage-devel] sage and jupyterhub for classroom teaching
On Wed, Apr 1, 2015 at 9:54 PM, William Stein wst...@gmail.com wrote: If somebody from JH decides to explain how every point above is slightly wrong, that would be nice, since I can't imagine that they aren't wrong! Actually, I don't see any major errors in there :) Just to add some detail, the basic design of JH is: out of the box, assume: 1. local Unix auth (PAM) 2. for each authenticated user, you start their notebook as a local subprocess. That makes it very easy to run it in your typical 'university research group' environment, where there's some NFS server under a desk a beefy shared unix box where the group members all SSH in. But both #1 and #2 above are separately customizable. What Jess did was to replace #1 with Github auth, and #2 with a fairly sophisticated Docker setup (she ended up contributing fixes to Docker Swarm just as it was released). With said customizations, the same basic architecture should be usable in some other scenarios, just as Jess has demonstrated. JH doesn't have out of the box any of the many sophisticated features that SMC has, nor is most of that stuff on our roadmap. Lots of what William has done there is needed to provide a robust, 24/7 service with (hopefully :) paying customers, and we're not doing any of that. I do hope that SMC will be very successful, and we do talk to William every time we have a chance to share ideas, brainstorm, etc. Cheers f -- Fernando Perez (@fperez_org; http://fperez.org) fperez.net-at-gmail: mailing lists only (I ignore this when swamped!) fernando.perez-at-berkeley: contact me here for any direct mail -- 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.
Re: [sage-devel] sage and jupyterhub for classroom teaching
On Thu, Apr 2, 2015 at 3:00 PM, Fernando Perez fperez@gmail.com wrote: On Wed, Apr 1, 2015 at 9:54 PM, William Stein wst...@gmail.com wrote: If somebody from JH decides to explain how every point above is slightly wrong, that would be nice, since I can't imagine that they aren't wrong! Actually, I don't see any major errors in there :) Thanks for adding additional clarifications! Just to add some detail, the basic design of JH is: out of the box, assume: 1. local Unix auth (PAM) 2. for each authenticated user, you start their notebook as a local subprocess. That makes it very easy to run it in your typical 'university research group' environment, where there's some NFS server under a desk a beefy shared unix box where the group members all SSH in. But both #1 and #2 above are separately customizable. What Jess did was to replace #1 with Github auth, and #2 with a fairly sophisticated Docker setup (she ended up contributing fixes to Docker Swarm just as it was released). With said customizations, the same basic architecture should be usable in some other scenarios, just as Jess has demonstrated. JH doesn't have out of the box any of the many sophisticated features that SMC has, nor is most of that stuff on our roadmap. Lots of what William has done there is needed to provide a robust, 24/7 service with (hopefully :) paying customers, and we're not doing any of that. I do hope that SMC will be very successful, and we do talk to William every time we have a chance to share ideas, brainstorm, etc. The feeling is very mutual. My dream would be that SMI = (SageMath, Inc.) is very successful and can be one of JH's industry partners, and that SMI can financially (and otherwise) help to support JH development. William P.S. speaking of github auth, I happened to add support for Github auth (and Google, Facebook) to SMC a few days ago, in case that's interesting to anybody, look at https://github.com/sagemathinc/smc-public/blob/835966ac41f75ae26247aa12ff49c482f0aa2180/hub.coffee#L347 There have now been 403 links between an auth provider and SMC since I made this live about 3 days ago, with 31 facebook, 50 github, and 322 google, which is perhaps interesting. Cheers f -- Fernando Perez (@fperez_org; http://fperez.org) fperez.net-at-gmail: mailing lists only (I ignore this when swamped!) fernando.perez-at-berkeley: contact me here for any direct mail -- 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 (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.
Re: [sage-devel] sage and jupyterhub for classroom teaching
On Thu, Apr 2, 2015 at 3:11 PM, William Stein wst...@gmail.com wrote: The feeling is very mutual. My dream would be that SMI = (SageMath, Inc.) is very successful and can be one of JH's industry partners, and that SMI can financially (and otherwise) help to support JH development. That would be awesome, and in the meantime, we do hope that everything we're doing on JH can be of use to SMC. Cheers f -- Fernando Perez (@fperez_org; http://fperez.org) fperez.net-at-gmail: mailing lists only (I ignore this when swamped!) fernando.perez-at-berkeley: contact me here for any direct mail -- 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.
[sage-devel] sage and jupyterhub for classroom teaching
It appears that sagenb is in maintenance-only, and that for graphical interface, the IPython notebook is the way forward. The IPython notebook looks wonderful and will probably be a very able replacement for single-user scenarios, but it lacks the multiuser capability that sagenb provides. I noticed that IPython notebook is now Jupyter, and that there *IS* a multiuser offshoot for that now: Jupyterhub. Does anyone here have experience deploying Jupyterhub and/or using it to interface with sage? It looks like a very attractive option for cases where SageMathCloud isn't appropriate. The blog post here: https://developer.rackspace.com/blog/deploying-jupyterhub-for-education/ looks promising as far as how mature Jupyterhub is, but I have no idea how it would work with sage. Comments and insights welcome! Nils -- 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.
Re: [sage-devel] sage and jupyterhub for classroom teaching
On Wed, Apr 1, 2015 at 9:21 PM, Nils Bruin nbr...@sfu.ca wrote: It appears that sagenb is in maintenance-only, and that for graphical interface, the IPython notebook is the way forward. The IPython notebook looks wonderful and will probably be a very able replacement for single-user scenarios, but it lacks the multiuser capability that sagenb provides. I noticed that IPython notebook is now Jupyter, and that there *IS* a multiuser offshoot for that now: Jupyterhub. Does anyone here have experience deploying Jupyterhub and/or using it to interface with sage? It looks like a very attractive option for cases where SageMathCloud isn't appropriate. The blog post here: https://developer.rackspace.com/blog/deploying-jupyterhub-for-education/ looks promising as far as how mature Jupyterhub is, but I have no idea how it would work with sage. Comments and insights welcome! Jupyterhub (JH) and SageMathCloud (SMC) are basically approaching the same problem but from somewhat different directions. The following comparisons are difficult to write and probably wrong; they might be hard to write for the JH people, since they probably barely understand SMC, and for me since I surely barely understand JH. So take all this with a huge grain of salt, as it is probably more a reflection of my own ignorance than anything else. That said, I had some long conversations with Fernando Perez and some other JH devs in Berkeley a few months ago, and he helped me get a sense of what's going on. - JH has a bit more funding and resources right now than SMC -- they have several fulltime people, significant funds from Moore/Sloane, etc. SMC is 99.9% me in my spare time... - JH is (probably) all BSD licensed, where SMC is GPLv3. (SMC is GPLv3 because of the data management plans in several of my NSF grants, which were written more around Sage, but still apply.) - SMC is -- more or less -- only designed to run multi-data center on a large number of computers, and there's no automatic process to just get a copy of SMC up and running. Also, SMC requires constant human monitoring and maintenance. - JH is written in Python as much as is reasonably possible, whereas SMC is mostly written in CoffeeScript. - Both SMC and JH have target audiences that include large undergraduate courses with students using Python and other languages. - JH's money model is a Moore/Sloane grant, along with industry partnerships; this combination is the main plan for longterm funding, and it will probably work at a small scale at least, since Fernando et al. are very effective at grantsmanship. The JH devs explicitly decided not to go the route of starting a for-profit company.In contrast, SMC's money model is a for-profit company (SageMath, Inc.), with investment and a (yet to be revealed) business model, etc.; there's no evidence that I'm any good at this approach to funding what we're doing, but this approach definitely has the potential to scale up substantially in size. - Realtime collaboration is central in the design of everything in SMC, and of a lot of things I haven't implemented yet, but have laid the foundations for. JH has very limited collaborative capabilities (e.g., the coLaboratory stuff from Google), but of course much is planned. I hope that they are an order of magnitude better at implementing that stuff than me, since it is frickin' hard. - IPython is a heavily used and central part of SMC. Sage worksheets in SMC are much different than IPython notebooks (which are very similar to sagenb notebooks), and I view the two as complementary. If somebody from JH decides to explain how every point above is slightly wrong, that would be nice, since I can't imagine that they aren't wrong! -- William Nils -- 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 (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.
