12:37:09 #startmeeting 12:37:09 Meeting started Tue Jun 4 12:37:09 2013 UTC. The chair is rebelsky. Information about MeetBot at http://wiki.debian.org/MeetBot. 12:37:09 Useful Commands: #action #agreed #halp #info #idea #link #topic. 12:37:20 Good morning those of you who are here. 12:38:09 Good morning 12:38:36 Some changes to the schedule this morning. 12:38:44 Had planned two 45 minute sessions on tools. 12:38:55 Now - One 45 minute session on tools, one 45 minute session on git. 12:39:22 And maybe we'll even get some food. 12:39:32 #topic TOols 12:39:35 #topic Tools 12:39:52 Here we're not really talking about how to use the tools, but much more about how the tools fit in the culture. 12:40:09 Slide: FOSS Communication Needs 12:40:24 Talk to students about this in terms of building and guarding your professional image. 12:40:30 And how open source helps with all of that. 12:40:40 Interesting characteristics 12:40:47 Most are distributed. Many are globally distributed. 12:41:06 Spread out in many ways - not just by location, but also by times (eg., irregular schedules) 12:41:20 And you need to deal with turnover and inconsistency of activity. 12:41:26 There are people who have bursts of activity. 12:41:29 And you want to have people come in. 12:41:37 Amazingly, all of these are also characteristic of business. 12:41:46 Certainly, turnover is much higher than it used to be. 12:41:54 And businesses are much more distributed than they used to be. 12:42:04 Slide: FOSS Team Communication 12:42:17 Ask students: Given all these challenges, how do they get the work done? 12:42:35 What are the processes and culture? (Get the students to think about things other than the technology.) 12:43:16 Communication that is ... online, decentralized, transparent, both transient and archival, etc. (Archival - Meetboot etc.) 12:43:24 rebelsky: off topic : http://people.biology.ufl.edu/ulan/ 12:43:29 Slide: FOSS Communication Tools 12:43:34 Slide; IRC 12:43:44 Interesting. It looks like clunky chat to students. 12:43:52 A place to explain the process and culture and how it fits. 12:44:26 Meetbot adds some interest. 12:44:54 Peter runs IRC channel for his department. Helps support students who go through the curriculum in different ways. 12:45:00 Holds virtual office hours on the chatbot. 12:45:19 Now there's a community. Students who aren't in his classes still participate in the IRC. 12:45:36 (Whoops ... Holds office hours on the IRC.) 12:46:06 Now writes his own bots for the channel, and students also write with the bots. 12:46:37 #idea (from Peter): Department IRC that gives students a place to hang and to support each other. 12:46:57 #idea (from Peter): Having students write useful meetbots 12:47:10 Silly one IRC confessions 12:47:16 And then run analytics on the data 12:47:40 #link Peter's code is on bitbucket 12:48:13 #idea (from Joanie) Written communication breaks down barriers. If you have text, you don't have to worry about accents and more. 12:48:33 And if you don't speak English, you can use Google translate to get some sense as to what's happening. 12:48:48 Low bandwidth is also important. You can do it any where. 12:49:01 sbenthall: Thanks 12:49:22 And you can just drop in (and lurk). 12:49:28 #info https://bitbucket.org/pbui/bobbit 12:49:45 #link https://bitbucket.org/pbui/bobbit 12:50:01 Slide: IRC Social and Process Impact 12:50:21 Has many of the effects of the nearby water cooler at work. You can hear interesting things by just hanging around. 12:50:33 And then you can contribute. 12:50:41 And you can tap Heidi on the shoulder. 12:51:14 The power really comes from the ability to check in with "real people" in the project. 12:51:36 Remember that some use of IRC is just tradition. "That's the way open source has been done in the past." 12:51:49 Slide; Wikis 12:52:01 Think about the cultural role. 12:52:16 Students have used Wikipedia for years, but haven't thought deeply about them. 12:52:25 E.g., How are they maintained? How do you get info in? Etc. 12:52:43 Cliff K has editing a Wiki page as a first step in getting started with open source. 12:53:02 In case you hadn't noticed, Etherpad has a slider on it. 12:53:34 Slide; Wikis and FOSS 12:53:42 Meets important criteria: Quick publication, Decentralized control, etc. 12:53:50 Slide: Wiki Examples 12:54:35 Note that Media Wiki's Wiki has lots of interesting "how to material" = How to comment well, etc. 12:54:37 #link Why poeple contirbute to wikis https://www.youtube.com/watch?v=WghdsOz9KwA 12:54:45 And the Fedora Wiki gives you lots of examples of an active, technical Wiki 12:54:48 Slide: Blogs 12:55:01 #link Fedora project wiki fedoraproject.org/wiki 12:55:24 In open source projects, 'blogs are a way of how you establish how you are 12:55:39 And the 'blogs of the people on the team help establish the "brand" of the project. 12:55:49 Slide: Blogosphere and Beyond 12:56:06 Need to get students to think beyond the blog as itself. 12:56:22 For example, blog-to-blog links. 12:56:41 In his experience, students (like faculty) don't know that you can subscribe to blogs. 12:56:52 And planets make things even more powerful in terms of flows of information. 12:57:01 Slide: Blog Value for FOSS 12:57:09 Students need to learn that "In an online world, you are what you write." 12:57:33 Some employers even demand access to Facebook page. 12:57:44 But that's becoming illegal (e.g., in Illinois and CA) 12:58:22 #link http://planet.fedoraproject.org/ 12:58:45 Look, lots of people blogging about stuff. 12:59:28 #link http://en.wikipedia.org/wiki/Hackergotchi 12:59:35 And floating heads. Hackergotchi (or something) 13:00:08 Slide; Summary 13:00:34 FOSS communities are large, and they are not magic, but they are often far ahead of industry in how they manage distributed teams. 13:00:45 And keep the framework of "Your professional image matters" 13:01:23 (Sam's note to self: Where does "building your professional image" fit in our curricula? Maybe in the new "Learning from Alumni" class.) 13:01:27 * ruehle what is the name again? 13:01:29 #topic Breakfast 13:01:44 patti: Hackergotchi 13:04:11 rebelsky: things you can make zodbot do https://fedoraproject.org/wiki/User:Rsuehle 13:04:39 #link actually useful zodbot commands http://fedoraproject.org/wiki/Zodbot 13:06:05 #chair rsuehle 13:06:05 Current chairs: rebelsky rsuehle 13:06:21 Aww ... Zodbot doesn't give any snarky messages when someone becomes chair. 13:06:46 d#chair pnutzh4x0r 13:06:51 Whoops. 13:07:06 #chair pnutzh4x0r 13:07:06 Current chairs: pnutzh4x0r rebelsky rsuehle 13:09:43 .chair rebelsky 13:09:43 rebelsky is seated in a chair with a nice view of a placid lake, unsuspecting that another chair is about to be slammed into them. 13:09:50 It does if you use a . instead of a # :) 13:10:29 .moarbacon 13:10:29 Really rsuehle? I'm not giving up my bacon! 13:10:47 Haha, zodbot! I already HAVE bacon! 13:11:36 rsuehle: But do you have the white bread that absorbs all the fat? 13:11:55 That seems like a personal question! 13:12:32 (Just observing the platter.) 13:12:45 .moar whitebreadthatabsorbsallthefat rsuehle 13:12:45 here rsuehle, have some more whitebreadthatabsorbsallthefat 13:13:00 rsuehle: A more serious question. When there are multiple simultaneous conversations going on, how do people keep track of which comment goes with which conversation? 13:13:12 Practice :D 13:13:34 Usually you can kinda figure out who's talking to whom and work it out, and sometimes you have to say, "Wait, what?" 13:14:00 In some meetings, there's protocol to keep it form turning into chaos. Basically when you want to speak, you say "!" and get called on. 13:14:41 I was thinking it might be useful to tag conversations (e.g., with *irc use*) 13:15:03 My bigger problem is how to keep track of where something is happening--I have 37 channels open at any given time. (I just counted.) 13:15:26 What do you mean by tag them? 13:15:27 So, do people do analyses of conversation threads in IRCs? (Maybe that's a questions for sbenthall) 13:16:08 *irc use* For example, I might indicate that I'm in a conversation with this. 13:16:22 *breakfast* Or talk about the wonder of the paper cup scoops like this. 13:16:42 *irc use* Although stars are probably the wrong symbol. 13:17:23 But doing something like that would let me type "grep '\*irc use\*' sillychatlog" to see a conversation. 13:17:42 *irc use* But of course, that prior line shows the difficulty of remembering to type the prefix. 13:19:14 *irc use* Comment from Peter: You can create a new channel by just joining it 13:19:21 *irc use* And once you've created a channel, you probably want to take it over (as moderator) 13:20:19 *irc use* A recent grad who is in graduate school participates in his chat and can give advice on grad school. 13:21:54 *irc use* Peter recommends weechat as a terminal client 13:22:03 *irc use* Or iirsi 13:22:54 Theoretically that's what #topic does 13:22:58 *irc use* Joanie notes that iirsi lets her connect from wherever she is and see a history of what's been happening. 13:23:12 Although it's at the top of a conversation instead of for concurrent threads 13:23:26 rsuehle: But #topic assumes that there's only one conversation going on at a time. 13:23:30 right 13:23:36 irssi + {screen,tmux} + ssh connection 13:23:39 .laugh about simultaneous comments. 13:23:58 #topic More Tools and Stuff 13:24:02 Slide: Release Cycle 13:24:24 Think about it in terms of papers. When is the paper due, the review, the notifications, etc. It happens every year for SIGCSE. 13:24:36 (Or for every other conference) 13:24:50 Software projects are also on release cycles. 13:25:01 And this matters in terms of course planning. 13:25:24 When you get involved in a project, you should check the calendar first. The week of a release is not a week that people will be happy to hear from you. 13:25:28 (Or maybe even notice you.) 13:25:38 .moar bandwidth Drexel 13:25:38 here Drexel, have some more bandwidth 13:26:14 #link example release schedule methodology https://fedoraproject.org/wiki/Fedora_Release_Life_Cycle?rd=LifeCycle 13:26:30 #link https://live.gnome.org/Schedule 13:26:40 Look for freezes. 13:27:01 #link https://live.gnome.org/ThreePointNine 13:27:29 Also sometimes things like this happen (note how many times the dates changed) https://fedoraproject.org/wiki/Releases/18/Schedule?rd=Releases/18#Key_Milestones 13:27:45 Remember that "No more changes can be made" includes "and you've gone through all of the feedback iterations" 13:28:24 There is such a thing as "freeze break" to get something in after, but you'd better have a really, really good reason, and since studnets aren't going to be working on show stoppers, that's largely irrelevant. 13:28:40 (But I also wouldn't want some student to see that somebody got a freeze break and assume they could easily too.) 13:28:50 Note that there are even things like "String Freezes" so that translators have time to change all the strings. 13:30:05 Everyone has freezes 13:30:09 Slide: Roadmaps 13:30:25 Usually on a per-project basis. Listing what they want done before the next release. Sometimes who is responsible. 13:30:36 Sometimes they'll tell you which release that they are targeting. 13:31:00 #slide Upstreaming 13:31:11 Make sure to watch the video that's linked on the slide! 13:31:22 #info Make sure to watch the video that's linked on the Upstreaming slide. 13:31:38 Slide; Development Process 13:31:47 You code until you get stuck. What happens then? 13:31:56 Don't wait a few days. 13:32:00 So track your work and explain what you're doing. 13:32:09 'Blogs are a good way to keep track of what you're doing. 13:32:18 And then when you ask for help, you can give links to what you've been doing. 13:32:43 And you may be done when you report it in. 13:32:51 Since someone may pick it up. 13:34:22 Sam needs to remember that in third generation VCs, it's not "don't check in nonworking code" but rather "don't upstream nonworking code" 13:34:34 Slide: Summary 13:36:19 Side note: Don't forget about pastebins - a place to put short snippets of code. 13:36:52 #topic PasteBins 13:37:05 Typical use: You're in an IRC and want people to see some lines of code. 13:37:14 Copy lines (e.g., errors or whatever) 13:37:17 Paste into pastebin 13:37:19 Put link into IRC 13:37:33 But it doesn't have to be code. It can be any big chunk of text. 13:37:56 DON'T PASTE TEN LINES OF CODE! PLEASE! 13:38:02 hey sam, check my code http://pastebin.com/u2GLsfiw 13:38:34 It can do syntax highlighting and such. 13:38:38 And you can set expiration dates. 13:38:59 camm: I refuse to help you until you meet the GNU conventions for C code. 13:39:36 #topic git 13:39:48 This is about using git to interact with open source communities. 13:39:59 git is unwieldy. Get a book. 13:40:04 git-scm.com has a nice free book. 13:40:10 And other resources. 13:40:44 #info: GitHub has something similar to Pastebin called Gist. E.g. https://gist.github.com/kwurst/5705980 13:41:05 #link http://git-scm.com/ 13:41:25 Okay, we're going to work in the terminal. 13:43:06 Looking at OpenMRS you can see that the source code is at git 13:44:25 Don't do the command yourself (at least not here at Drexel, with the overloaded guest network.) 13:44:43 #link Gnome runs their own git server https://git.gnome.org 13:45:24 Many companies use git, even proprietary software companies. But their repositories are private. 13:45:37 Branches and tags are part of the version control vernacular. 13:46:10 git came out of Linux project so it's very popular. 13:46:39 git tries to do "just the basics" 13:47:13 People who want a bit more perspective on version control can read all sorts of things. 13:47:19 #link http://www.ericsink.com/vcbe/ 13:49:06 We're playing with https://git.gnome.org/browse/accerciser/ 13:49:20 http://www.catb.org/esr/writings/version-control/version-control.html 13:49:21 #link http://www.catb.org/esr/writings/version-control/version-control.html (ESR on Version Control) 13:59:48 (SamR was offline due to the wonders of freechat.freenode.net) 13:59:57 #topic Creating Your Own Project 14:00:04 git init is where you start 14:00:07 Add your files 14:00:13 #link git cheat sheet https://help.github.com/articles/git-cheatsheet 14:00:27 (Note that you only do git init once in a while.) 14:00:34 git add filename 14:01:51 Then you need to commit it in. 14:02:01 git commit -m "Message" (the short version) 14:02:11 (Okay, I'm not going to type all of the git instructions here. Sorry.) 14:02:48 #action Cam should create a page on foss2serve or wherever summarizing his lecture. 14:03:16 #action Joanie and RSuehle should add their deep wisdom to that page once it's created. 14:03:40 * joanie thinks Cam is doing just fine. :) 14:04:26 agreed! 14:04:33 joanie: I also think Cam is doing just fine. But I've observed that you often have a few subtle additional things to add. 14:05:48 rebelsky: indeed. But he's doing exactly what I would do in the order I would do it and saying exactly what I would saying. Modulo the VI thang. ;) 14:05:59 !emacs 14:06:00 pnutzh4x0r: A previous girlfriend of mine switched to emacs. Needless to say, the relationship went nowhere. 14:06:05 s/would saying/would be saying/ 14:06:09 joanie: Modal editors rock! 14:06:35 emacs would not have had that problem. Just sayin'. 14:06:42 And for those of us old enough to remember having to mount a second disk in order to access emacs, knowing vi meant that we could use ex. 14:08:33 Remember that git diff helps you find the latest change. 14:09:12 Danger! git checkout FILENAME will grab the latest one from the repository, overwriting any changes you've made. 14:09:33 Note that all that we've been doing right now has been on the local machine. 14:09:55 Which means that when Cam's daughter throws your laptop in the lake or river, your work is gone. 14:10:03 Needs to go to a server. 14:10:36 WIth your local work, it's often advantageous to do lots of commits. 14:11:07 Try to keep commits small. It makes it easier to keep track of what's been done. 14:11:39 Drexel says we shouldn't have a coffee break. 14:13:00 #topic On to Github 14:13:10 Create a new repository. Make sure to include a README. 14:15:08 Cam's is cmadconnell https://github.com/cmacdonell 14:16:27 Note that it's good to start by making the repository on github, rather than on your local machine. 14:16:58 Note: Write git commits in present tense. 14:17:11 Cam gets irritated if you write in the past tense. 14:17:22 Cam gets even more irritated if you write "This commit does ..." 14:17:38 Remember that git commit is local. How do you share it with the world? 14:17:51 The repo you clone from is called the "origin". 14:19:10 You need to sync your local version with origin. 14:20:16 So you push it upstream to the origin 14:20:22 Command: git push origin 14:22:03 how does git push know where to push to? how does it link to github? 14:22:19 #link Eric Sink on the airplane example http://www.ericsink.com/vcbe/html/dvcs_disconnected_operation.html 14:22:41 lorip: When you cloned the repo it created an invisible file that says where the master is. 14:22:51 You can also name other repos and push to them. 14:24:07 lorip: You'll often find that when you push it, you'll get prompted for an account and password. (Well, you can configure that away.) 14:24:27 ok - I'm missing a piece of how this works - if I clone on my local machine - it knows the original source, but I don't want to push back there, I want it to my github account - how does this work? 14:25:18 Yay! github does changesets rather than the wonder of "move is remove/add" from 2nd gen version control systems. 14:26:07 lorip: u can specify the remote target inplace of origin 14:26:22 I know someone will correct me, but here's my model. Suppose that I'm using Ushahidi. I make a clone of Ushahidi in my github account. I then clone my clone onto my local machine. I make the changes on my local machine. I push to my github copy of Ushahidi. 14:26:35 Finally, if what I've done is useful, I push it back to upstream. 14:27:10 lorip: Did one of our answers help? 14:27:39 thanks - I was missing the piece of cloning to my github account and then cloning again to my local machine - that makes sense! 14:35:29 (Whoops ... SamR got distracted making organizations and didn't take notes.) 14:49:00 Feels like the attack of the clones in here (ba dum dum) 14:50:43 IMportant command: git format patch master 14:51:19 #topic OpenSource.com 14:51:32 Started out staying away from code. 14:51:39 Opensource.com is a great way to publish your thoughts on FOSS 14:51:45 Look ... there's even a channel for education. 14:51:52 And look! We can contribute to OpenSource.com 14:51:56 And also a great source of material to be used in class. 14:52:17 There used to be a massive timeline. Ruth will send it to you. 14:52:47 Open source is a culture ... not just a way to develop code. 14:53:30 There's a "Submit an Article" link. It may sound a bit intimidating. But don't worry. It's mostly to give them a way to filter "the weird crap". 14:53:47 As long as we can write complete sentences (and interesting stuff) she'll pretty happpy. 14:54:04 And you can email her: ruth@redhat.com No need to figure out whether u comes before e. 14:55:58 They even post a summer reading list. 14:56:20 lorip: Here's a cute little intro to different VCS activities: http://www.ericsink.com/vcbe/html/basics.html 14:56:34 (I like it more in the PDF, but, this is ok.) 14:56:39 Question: Who reads this? 14:57:00 It's an evolution of RedHat magazine. 0.5+ M users. 15:00:19 s/users/email subscribers -- not the same :) 15:00:41 I guess what I didn't quite finish saying was that it started as a largely technical audience but has evolved and grown to open-other-things people 15:05:09 Sharing work is hard. 15:05:16 Often doesn't get credited by the institution. 15:05:22 So you have to think about it in terms of service to your community. 15:05:32 And not just "I'm not sure that my assignment is good enough." 15:05:37 We need to be more open. 15:05:44 And we also live in a world of peer-reviewed publications where we want everythi ng to be finished and done. 15:05:49 This is a tough nut to crack. 15:05:59 There have been a variety of projects to get faculty to share their stuff. 15:06:10 And they haven't been very successful. 15:06:16 But it's useful. 15:06:21 And it fits within the FOSS community. 15:07:07 Swenet or whatever has assignments with some comments (using word comments) that explain to students what goes where. 15:07:14 And Greg knows that it gets used. 15:07:40 If you're lucky, it becomes a community. You get comments on what has and hasn't worked with an assignment. 15:08:06 (Sam is sick of getting kicked off of webchat.freenode.net) 15:08:22 Slide: Defining Learning Activities 15:08:33 Putting it out there helps. 15:08:36 But it also helps to have additional stuff there for those who might adapt/adopt. 15:08:58 So they have a template. It's a checklist. 15:09:12 Like many checklists, it helps the author think about this. 15:09:30 But the checklist is also potential metadata, so people can search on it. 15:14:35 We're doing some brainstorming. 15:14:50 We're going to use Darci's CS2 as an example. 15:15:25 http://xcitegroup.org/softhum/doku.php?id=f:50ways 15:15:41 Open source is a way to help address "Why are we learning stacks and queues and ...." 15:15:48 She'll be working with the OpenMRS projects. 15:15:59 Her hope is to find existing code that exemplifies the topics they teach in CS2. 15:16:18 Can she then ask students to evaluate those sections of code. 15:16:39 Maybe even an opportunity for them to document. 15:17:24 What are things we can commit back to the community as a class? 15:17:33 (Comfort of acting as a class, rather than as individuals.) 15:19:11 Sam's suggestion ... playing with a data feed. 15:19:23 They may not be able to contribute back too much, but perhaps feature requests and bug notes. 15:19:35 Cam: You can't expect much more contribution than that from CS2 students. 15:20:01 But manipulating the data that comes from the data feed provides ways to explore some of the key ideas of the course. 15:20:41 In terms of pulling out pieces (back to Darci's idea): You want to pull out the right amount. You don't want them to feel overwhelmed. (Or at least not too overwhelmed.) 15:21:10 Does making them too overwhelmed make it more like a weed out course? 15:21:27 At some point students need to understand about large systems. It helps motivate revision control systems and such. 15:21:49 Scaffolding is important. 15:22:43 How do I know what's the right piece? Greg's hope is that the groups we have here will be a good source. 15:23:18 Question (to Darci): Would you keep using the same code each year, or would you look for new ones? 15:23:35 Answer (I think): Continue to use the same code ... it's going to take awhile to learn. 15:24:28 Patti, Stoney, and Peter were talking about CS1/CS0. 15:24:42 At that level, it's hard to get them to contribute. But you can talk about methodologies and tools. 15:24:57 Gives opportunities for other things, such as peer review. 15:25:23 Or an assignment in which there are multiple iterations. You have to commit something each week, but the final product is not due until week three. 15:25:32 And there's feedback along the way. 15:26:33 Issue of motivation. (E.g., why do you need a portfolio? when you graduate, what do you want your portfolio to be?) 15:26:49 Is it reasonable to get them thinking about that, and how do you get there? 15:27:11 Question for students: How far does your coursework get you? And what does that mean about what you have to do beyond coursework. 15:27:30 Motivation of assignments. Open eyes to a whole new world of things. 15:27:57 Learning objectives are motivational, rather than technical. Help them understand their field. 15:28:53 Where are we going to find the examples for Darci and company? 15:28:56 That's a great thing for us to share. 15:29:53 Greg: What's the difference between requirement spec and design spec 15:30:14 You have to be clear on the difference. 15:31:32 Different courses? 15:31:58 sbenthall is teaching a course to master's students in the iSchool (Berkeley) about open source communities. 15:32:04 Fulfuills "management requirement" 15:32:41 #link The Architecture of Open Source Applications http://www.aosabook.org/en/index.html 15:32:58 Would that help with any of Darci's stuff. (It's not one project, but many.) 15:33:09 (And I don't know how deep the analysis is.) 15:33:32 Back to sbenthall's stuff: The students will participate in communities and report back on the community. 15:33:42 He's happy to share the final papers (projects?) with us. 15:34:53 Note: FOSS is also now a research opportunity for people (questions of how projects are developed, etc.) 15:35:16 E.g., mining software repositories. 15:35:37 sbenthall: hmmm what if we made a badge system for foss 15:35:46 so like klout but for open source contributions 15:36:00 you could data mine mailing lists, irc logs, planets, etc. 15:36:09 bug trackers 15:36:11 #topic DataKind 15:36:16 pnutzh4x0r: Like Mozilla's badge program? 15:36:30 right 15:36:48 #link http://datakind.org/ 15:36:53 "Using data in the service of humanity." 15:37:07 Provides an opportunity for certain kinds of hackathons. 15:37:15 People tend to navigate to those with similar skillsets. 15:37:20 Programming teams of 3-5. 15:37:59 heidie: mozilla badge is more about learning or measuring skills 15:38:22 im thinking more like klout: http://klout.com/home 15:38:31 which measures your influence 15:38:38 in this case it would measure your contribution 15:38:43 They tend to be in big cities. 15:39:09 (just to be clear, I'm staying on the DataKind thread, rather than the badge system) 15:39:41 Random hacks of kindness, driven by data rather than by software. 15:39:58 pnutzh4x0r: Doesn't GitHub and/or Ohloh do something like this already? 15:40:06 Greg likes to keep throwing things out to see if we're interested in things. 15:40:12 #link http://spaceappschallenge.org/ 15:40:17 some ideas for a hackathon ^ 15:40:41 heidie: but this would integrate all forms of contributions, or that would be the idea anyway 15:40:58 pnutzh4x0r: great idea. I think that's what Ohloh tries to be. I wonder how open their contributions are? 15:41:38 #topic Misc. 15:42:04 Colorado has a degree in "ICT in Development" (e.g., in developing nations). 15:42:05 pnutzh4x0r: one problem is heterogeneity between projects. scraping the software repositories is one thing, but a lot of participation is often on external infrastructure (mailing lists, plants, blogging) 15:42:21 Greg has started to talk to them about the use of open source software (which sounds like it wasn't on their radar) 15:42:34 sbenthall: that's the challenge :] 15:42:35 pnutzh4x0r: I hear that Mailman 3.0 is under development I've been thinking of asking them about how to make it easier for external instrumentation 15:43:35 Today's guest: Tim S. He's a librarian. He pays attention to the various stuff that's happening in the iSchool. 15:43:47 pnutzh4x0r: there's also a metrics-wg@theopensourceway.org list where they talk about this. There's a project that gets into this already for scraping from e.g github 15:43:56 What do you do when you have a community that needs something, but no obvious open-source solution. 15:44:16 #link http://metricsgrimoire.github.io/ 15:44:20 And a different model: MouseTrap. 15:45:00 Sam wishes that IRC was better threaded (either by design or by common usage) 15:45:10 pnutzh4x0r: see also openstack's new dashboard 15:45:11 #topic Our Lunch Speaker 15:45:13 #link http://activity.openstack.org/dash/browser/ 15:45:31 Before being here, he was a corporate librarian. "Special Libraries Association" 15:46:05 This group really knows their internal customers well. 15:46:26 pnutzh4x0r: a lot of the metrics work has been driven by the hackers, I believe. I hadn't heard of much academic work on the subject; partly because I think it's hard to really formulate a problem statement 15:46:31 (Like all librarians, they know about moving data around.) 15:46:43 He was a member of the Emergency Preparedness and Recovery Council. 15:47:09 sbenthall: i guess the questions would revolve around the social network 15:47:31 "We are not an open source community by any long shot." 15:47:48 But they want their services to be available where they are useful 15:47:52 sbenthall: you could also ask questions like how active are the core developers on the bug tracker or mailing list? 15:48:09 So how do you make the membership list available. What kinds of requests come up in times of emergencies? 15:48:42 "What if all 10K of us could sign up to be accessible for emergencies." 15:49:00 Some of the requests are obvious: I need this document. 15:49:25 Some of it relies on particular expertise: How do I get 100 tractors (or whatever). It's likely that the librarian at certain companies would know this. 15:49:52 Idea; Make the member database available as a gated community. (Available to selected folks, and maybe to research community.) 15:49:55