Perl 6 - the future is here, just unevenly distributed

IRC log for #webwork, 2012-03-26

| Channels | #webwork index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
13:22 JohnTravis joined #webwork
16:16 JohnTravis joined #webwork
18:08 ionparticle hi mgage
18:19 mgage hi -- back for a little bit between classes
18:20 mgage ionparticle: hi
18:21 ionparticle just answering your hi from yesterday
18:22 mgage thanks for submitting the patch for openwebwork -- it's useful to practice how this collaboration will work
18:24 ionparticle I'm working on adding create course and add/drop users for webservices
18:24 ionparticle so more will be coming?
18:24 ionparticle lol
18:24 mgage I have some changes to the tools in webwork2/clients which now register with the xmlrpc structure for the repo at openwebwork.  Basically in order to get the xmlrpc webservice to respond you need to specify a course, an instructor login name and a password  -- and the webservice checks that the login name is an instructor in the given course before responding.  It's how the library browser 3   javaScript is calling back into the server to get prob
18:24 mgage rendered.
18:26 mgage creating courses and add/drop users will be good additions -- I'll plan to let you do that.  You should be able to use a back end that is similar to the one that the webworkSOAP service uses to perform those functions.
18:26 ionparticle didn't know soap already have it
18:27 ionparticle so already wrote it and it's in my personal fork :/
18:27 mgage eventually I'd like to fuse those two services so they are easier to maintain. (have both SOAP and xmlrpc (and JSON for that matter) transport but with as much reused code as possible).
18:28 mgage take a look at webworkSOAP.pm -- it's pretty easy to read and to see what functionality has been implemented.
18:28 ionparticle ok, taking a look at it
18:29 ionparticle would be nice to have the same backend, yeah
18:29 mgage Does your personal fork have the library browser 3 code in it? have you already been supplying credentials to the xmlrpc to get registered?
18:29 mgage right
18:29 ionparticle I forked from openwebwork
18:29 ionparticle so it's whatever's in there
18:30 mgage I can push the client version that I have up to my personal fork (mgage) tonight.  It still needs a little work before I push it to openwebwork but I don't mind sharing it for now.
18:30 ionparticle it's fine, I do supply login/password/course
18:30 mgage If you are synced to what is currently on openwebwork then you have needed to supply userID, password and courseName in order to connect to xmlrpc.
18:30 ionparticle figured that's what it's supposed to do from the code
18:31 ionparticle only concern is that it's all transmitted in plaintext
18:31 mgage ok -- in that case you are up-to-date.  previously (before December say) there was no checking credentials
18:31 ionparticle so will have to figure out how to ssl it
18:32 mgage yeah -- that would be useful.  I do have it set up so that for the library browser -- as for everything else -- the password is only used once and then you use a session key for the subsequent calls.
18:32 mgage if you are working with https server is that sufficient ssl protection? or do you need (and can you get?) more
18:32 ionparticle https is sufficient
18:33 mgage ok -- most of our sites now run under https .
18:34 mgage Not sure I've done that with the xmlrpc calls or not however -- I don't think it would be too hard to have them respond only to https requests.
18:34 ionparticle I'm not sure if xmlrpc gets protected too
18:34 ionparticle it's something I was planning on checking out
18:35 mgage I'm not either, but I think changes to webwork.apache2-config would be enough to insure that -- probably don't have to change any of the other code.
18:35 ionparticle nice
18:35 ionparticle oh speaking of conf files
18:35 mgage yes
18:36 ionparticle would it be possible not to have /opt/webwork hardcoded in global.conf
18:37 mgage yes, if you look at the latest version global.conf has been split into prelocal.conf (looked at first) then global.conf.dist (which will seldom need local reconfiguration) and then postlocal.conf -- where you can override configuratons
18:38 ionparticle yeah, I'm overriding in postlocal.conf
18:38 mgage I'm not sure I have everything adjusted perfectly yet among those three files -- but the principal is that you can set them up and unless you are doing something really unusual you can update code and not make any changes.
18:39 ionparticle still, would be nice to get something like the $webwork_root variable for /opt/webwork
18:39 mgage OK.  that sounds about right -- I want to have some default assignment in global.conf.dist for those who haven't figured out what they want to override yet.
18:40 ionparticle $webwork_dir, I mean
18:40 mgage hmmm.  so that is available in global.conf.dist but not in postlocal.conf?
18:41 mgage that could be -- I read through that code when making the changes but I didn't write that originally so there might be some tweaks that would make this division more useful
18:42 mgage I'll try to take a look this week and see if there is something that would make this more useful -- or if you want to email me a suggestion of what you would like I'll look at that also
18:42 ionparticle ah, it's a minor inconvenience, no worries
18:43 mgage need to run -- next class --- but this would be a good time to tweak these configuration files so that they are maximally convenient for everyone.  I've made a substantial change so now is the right time to add additional tweaks.   ttyl
18:44 ionparticle thanks for your time!
18:44 mgage np

| Channels | #webwork index | Today | | Search | Google Search | Plain-Text | summary