Difference between revisions of "OE IRC 29June06"
|Line 525:||Line 525:|
midori will send transcripts as usual ...
midori will send transcripts as usual ...
Latest revision as of 03:20, 13 July 2006
midori: hi guys
jrichter: Hi folks.
jrichter: Does anyone know if Jen is coming today?
midori: She's taken the day off, so nope.
midori: (i.e. no, she's not coming)
midori: Amelia's also on vacation; Jane should be along soon ...
jrichter: Sourceforge (aka "the permanently fucked system you can never use") won't let me view her merge bug she posted yesterday. Does anyone know what the problem was?
Alex_MGI: It was a repeat of a bug she noted at an earlier time, but I can't remember the details.
midori: Not firsthand, but SourceForge has graciously allowed me to read the entry:
midori: merge "refused to work" until she moved the subsumed term to be a sibling of the term to do the subsuming; then it created two copies of the merged term, one is_a and one part_of the parent.
jrichter: I just logged out and back into sourceforge, and now I can finally see the bug.
midori: You can log in now?
jrichter: Finally. This is the merging with children conundrum again.
midori: Login was still broken a couple hours ago.
MelissaH: Hi folks
jrichter: I think I'm going to email the main GO list and suggest that we install bugzilla on the stanford servers. I'm really tired of sourceforge's tomfoolery.
midori: Hi Melissa!
midori: Hi Jane - good timing! Is it the advocacy group (or who) that's looking into alternatives to SourceForge?
j-lo: have we started yet?
midori: not really ...
Alex_MGI: This may be our quorum for today.
j-lo: it's not advocacy
j-lo: Chris was looking into developing something
midori: Shall we start? Here are the topic suggestions I emailed:
j-lo: but nothing's happening at the mo
midori: 1. Paper
midori: 2. 1.1 release plans
midori: 3. User testing
midori: anyone have any others?
jrichter: One quick thing:
jrichter: I still haven't gotten detailed feature request lists from Mel or Harold. Time is almost up!
MelissaH: I have been trying, but sourceforge has been down and I need to make a few more requests...
Alex_MGI: Update on graphical browsing?
Alex_MGI: or editing?
Harold: I put my request in for the graph plugin ages ago
midori: John - did you get Karen C's picks? I know of two or three that she'd vote for ...
jrichter: Yeah, you sent a short list of 3 things to the list. I incorporated all those requests, Harold, but I thought you might have others.
MelissaH: I am at home, will email them tomorrow even if no sourceforge.
jrichter: I didn't get Karen's list.
midori: SourceForge has just come back (within the past 2 hr).
midori: John - OK, I'll email you.
MelissaH: lets hope it stays up!
jrichter: Graphical browsing/editing update: I'm not doing any more development until I get approval from the WG. I've done all the proof-of-concept work. At this point, I'd actually be modifying OBO-Edit if I went any further.
jrichter: Let's get into the paper stuff, then...
midori: Graphical browsing fits into the 1.1 topic (#2), so we can come back to it.
jrichter: I liked Jane's suggestion that we submit a brief application notes paper to Bioinformatics.
MelissaH: I like that idea too.
Alex_MGI: Sounds good to me, except the author list might exceed the length of the main text.
jrichter: I think that we should have at most 3 or 4 named authors, but we credit the OBO-Edit Working Group as an author, and we include a footnote (or whatever) saying who that is.
midori: We've managed that with the GO Consortium once or twice.
j-lo: yeah - me too
j-lo: why don't we just have it as John + OE WG
jrichter: I need to check with Suzi on whether she would like to be credited. At one point she did, but then she suggested just crediting the WG in bulk.
MelissaH: If so, i vote for john +wg, keep it simple.
jrichter: I'd like to credit one or two other named authors, in the hopes one or two folks will put in a significant editing effort.
MelissaH: I guess you can always decide when its finished.
midori: John, do you plan to write the text and just have us comment/edit? Or do you want some of us to contribute to the writing?
jrichter: I think I'd like to do the first pass, and then send it to the WG for comments.
Alex_MGI: John, I think you should pick the named coauthors.
jrichter: Here's the general outline I was thinking of:
midori: ok - less work for me
Alex_MGI: Not necessary now.
jrichter: File Formats
jrichter: But Visualization would come before Searching/Filtering.
jrichter: That's it.
MelissaH: sounds good, it mirrors the most important points of the help guide.
jrichter: I'm redundant like that.
MelissaH: the intro will be really important, as it needs to draw in folks from many disciplines.
jrichter: Also, I'm redundant like that.
midori: Recycling is good.
midori: So ... John, do you need anything more from us during this chat?
jrichter: As far as named co-authors, I think I'd like Melissa to be one of the main editors (if she'd like), because she's always so thorough about pointing out when I've become incomprehensible.
MelissaH: I am laughing, ......
jrichter: And then whichever other person is very excited about this.
MelissaH: be happy to make sense of your recycling can.
midori: I'm willing (but don't want to step on toes, if anyone else really really wants it).
jrichter: Is everyone happy with John, Mel, Midori & the WG as the author list?
MelissaH: we don't have a lot of folks here today, maybe you should email everyone.
midori: good idea
jrichter: Good idea. I'll send out an email. I'll also have some kind of draft out to the group by next Thursday.
rama: Karen is out at a meeting. I agree we should wait.
jrichter: If anyone has any helpful suggestions about wording in the intro, please let me know.
jrichter: I'm ready to move on. Progress!
midori: John can start on his draft in the meantime, though (don't want to lose momentum).
j-lo: info for authors for bioinformatics are here btw: http://www.oxfordjournals.org/bioinformatics/for_authors/general.html
midori: Agenda item 2: OBO-Edit 1.1.
jrichter: Oh crap, wait.
jrichter: Let's back up for one sec-
jrichter: Someone suggested that we notify Bioinformatics before we finish the paper.
jrichter: Is that a good idea, how is that done, and who should do it?
j-lo: oh yes - want me to do that?
j-lo: i review papers for them...
MelissaH: or maybe suzi should, at least talk to her about it.
j-lo: as does midori
MelissaH: maybe both of you !
jrichter: Suzi knows what we're up to, and she's quite busy. I think she'd be happy if Jane spoke to Bioinformatics for us.
jrichter: (I give Suzi a report on what we're doing each Monday)
MelissaH: ok, good.
MelissaH: good, got to keep those PIs in the loop.
jrichter: Jane, can you give them a heads-up?
jrichter: Okay, back to OBO-Edit 1.1:
jrichter: I had two questions about this...
jrichter: First: Do I need WG approval for architectural improvements to OBO-Edit? For example, I want to upgrade the drag-n-drop system, and make some major revisions to the command-line processing system. But these are changes no one is going to vote for, because they're no fun at all. Can I just go ahead, or do I have to make the case to do these things?
MelissaH: what exactly do you mean by architectural? behind the scenes?
Alex_MGI: I think you should go ahead, because you determine the software architecture. If they break things, we'll let you know.
midori: I think you can go ahead and do things that will make OBO-Edit do what it does now, only better (e.g. more efficiently).
j-lo: yeah - we trust you
midori: We should have a say about anything that would change how we interact with O-E.
jrichter: Yeah. When both these things are done, OBO-Edit will seem to be exactly the same (except that you'll be able to drag and drop between windows in Mac OS). Both these things will make interesting changes possible down the road though.
MelissaH: I agree, as long as you don't change any features.
jrichter: Cool. I'll go ahead, then!
midori: Sounds good.
Alex_MGI: The point is, now we have an official version, so you can play with things in new betas.
jrichter: I will consult the working group if I think an architectural change will take a very long time.
jrichter: Item 1.5: I'm planning on doing a CVS branch for OBO 1.1. This way I can bug fix in OBO-Edit 1.0 while I develop OBO-Edit 1.1
midori: Yes, the one thing we might want to comment on is how to divvy up precious developer time.
jrichter: Anyone who works with the CVS sources should know that, because you're going to have to say which branch you want.
midori: I'm all for "OBO-Edit 1.0 while [you] develop OBO-Edit 1.1"; not sure I much care how you implement!
jrichter: This only matters if you're compiling the CVS sources, so I thought I'd mention it in case anyone is doing that.
MelissaH: I think we are all used to browsing cvs by now, right?
jrichter: It gets really weird once a branch has happened, though. It's like that movie where Ashton Kutcher goes back in time. But, you know, not as boring.
Alex_MGI: Obviously, you won't be adding new features to 1.00x, so it makes sense.
MelissaH: Perhaps when this happens you can email out a big warning not to get stuck in time.
jrichter: Item 2: How do we decide what makes it into OBO-Edit 1.1. Do we just have a straight "X votes" or you don't get in, or do we add as many things as we can until 3 months are up, or what?
j-lo: prioritise things with lots of votes, and then keep adding until the 3 months are up
Alex_MGI: I think you should do all the easy things you can in half the time and spend the other half on big bits.
midori: We have to take into consideration how long the additions would take.
midori: Related to Alex's comment, I guess ...
MelissaH: Lets collect the votes, and then you can assign a time estimate to the top votes and see if they fit into the 3 mo window.
Alex_MGI: I agree, and I think John has already addressed that. I think he should make a proposal based on our votes of what he thinks he can finish in 3 months.
jrichter: But we want to weight priorities towards faster items, right?
Alex_MGI: I think quick user friendliness things are important.
MelissaH: I think alex is right, it should be a balance, like an investment...
Alex_MGI: Also graphical editing can wait to 1.2 if necessary.
Alex_MGI: (but I still want it)
midori: Well, there are two variables: how much a feature is wanted (how many votes) and how much time/effort it takes. I think a small easy-to-do feature would have a lower "barrier to entry" than a toughie.
midori: (Note: I am not volunteering to try to quantify!)
Alex_MGI: That's why I propose that John do a proposal himself, based on our existing input.
MelissaH: yes, there might also be things he is more excited about, which never hurts.
jrichter: On Sunday or Monday (when all the votes should be in)., I'll compile the list of feature requests and come up with a proposal that I'll email to the list. Then we can let the sniping begin!
Alex_MGI: With a fixed date in mind.
jrichter: Let's pick that date now.
Alex_MGI: Sepember 1 ?
jrichter: (Realizing that I'll be having a baby around August 15th, and I'll probably not be much use for a couple weeks in there)
Alex_MGI: Oh, forgot.
midori: Sorry, is this the date for the feature list, first beta, or release target?
jrichter: Release target.
Alex_MGI: Release date then October 1.
jrichter: I'll be releasing constant betas as a I add features and fix bugs.
jrichter: Oct 1 sounds good to me.
MelissaH: I think the release date should be chosen one john's proposal has been agreed on.
j-lo: one thing:
midori: We also have to allow enough time for testing ... Sept. will be full of meetings (but that doesn't mean we can't still make Oct. 1).
Alex_MGI: Isn't his proposal based on the amount of time he needs to do the development.
jrichter: Yeah, it's easier for me to come up with a proposal if I have a deadline.
MelissaH: I see.
j-lo: if we want to switch to OBO 1.2 format for GO pretty soon, we need to get the spec up-to-date
MelissaH: I think oct1st is overly optimistic with the meetings and the baby.
Alex_MGI: It depends on how much we're adding.
jrichter: Does getting the spec up to date include those obsolescence changes?
j-lo: i think so...midori?
midori: Melissa - John can scale the list down to accommodate meetings and mini-John.
MelissaH: gotcha, I understand the idea now.
midori: Sorry, was gonna ask anyway ... do you mean the 'use_term' business?
midori: Then yes indeed.
jrichter: Okay, here's what I think we need to do: I'll talk to Chris, and then fix the spec immediately. I think we should get the obsolescence changes into OBO-Edit 1.002 (or whatever).
jrichter: This breaks our "no major revisions" rule, but this is a very special case.
j-lo: it is
midori: which reminds me ... has anyone tested that 1.002 beta?
jrichter: BUT: I'm not changing the tag names.
midori: ok then.
jrichter: At the last IRC meeting, we decided that we would assign little 3 person task forces to test a specific beta release.
jrichter: Let's figure out who that is now...
midori: I've done a tiny bit today ... the newline bug does seem to be fixed, and it adds terms normally.
midori: The two or three searches I did were fine too.
jrichter: Volunteers? Realizing that the changes to OBO format will be in this release...
jrichter stares with disturbing intensity at Midori and Jane
j-lo: okay - I volunteer!
midori can't type anymore ...
jrichter: Okay, I'm going to have to pick at random for the other two...
midori might as well continue testing 1.002 anyway since she started
jrichter: How's about... Rama or Harold? Can you do it? (If you volunteer now, you're off the hook the next few times)
midori: Is there anyone who can test that command line stuff?
jrichter: Mike Cherry's on it.
jrichter: Although Jane will love love love it once she tries it.
rama: I cannot test during the next two weeks. I am swamped with Annotation camp stuff (eurie and karen are away). After July 15th.
jrichter: Harold who is called Hurdl? How about you?
midori: I'm going to annotation camp too, and then will be on vacation. But I'll fit in testing around that somewhere ...
jrichter: How about Petra or Alex, then?
jrichter: (While we wait... Jane: Have you looked at that command-line stuff yet? Go-slims from the command-line!)
j-lo: not yet...
j-lo: sounds good though!
Alex_MGI: I'm busy tomorrow and out most or all of next week.
MelissaH: I can test.
MelissaH: I have a bunch of editing to do anyway.
jrichter: Okay, Melissa, Midori & Jane are the testers for this round. This means that they have to sign off before we get an official release.
jrichter: Anyone else can test and submit bug reports too! Those three just have signoff authority.
jrichter: Brief reminder to the beta-testers: Don't commit to CVS with a beta version.
jrichter: Especially this one, since we'll be majorly f-ing with the obo data adapter.
Harold: Lately, I'm also swamped
jrichter: Any other issues before we close this chat?
midori: For testers generally: can we have a checklist of what to test in each beta?
Alex_MGI: A list of changes.
j-lo: i want to know about the command line stuff - are there any notes John?
jrichter: You guys know that I always submit a detailed list of changes with every release, right?
midori: I guess 5 minutes isn't really enough time to do justice to usability testing ... but we shouldn't forget all about it.
jrichter: They're in the README file, and if you click the changelog button on the sourceforge download page.
jrichter: I also updated the user's guide with complete documentation for the obo2obo utility.
jrichter: (or you can type "obo2obo -?" at the command line)
jrichter: (The change list is at the very bottom of the readme file, if anyone's looking for it)
MelissaH: user testing.... are you all going to the GO meeting in sept? can this be where we coordinate some official testing?
midori: John - yes, and I always look at it.
jrichter looks fondly at Midori
jrichter: So I guess the beta testers should test the new stuff, and try to do your daily editing on the beta version. (Just use the official release to save your changes)
midori: I've just been wondering whether there are ever any specific aspects of changes that we should pay special attention to ...
Harold: yes that would be very helpful to know
midori: ... or anything left over from previous versions...
jrichter: The stuff so far is pretty trivial. The real tricky stuff will regard the parent/child merge and the obsolescence changes. So look for those in 1.002-beta2, coming soon.
jrichter: Everyone agrees that parent/child merge is a good idea?
jrichter: If everyone wants it, I'll make the changes, but parent/arbitrary descendant merge has got to stay illegal. You're begging for cycles otherwise.
midori: yup ... I suspect it's the most common merge situation these days.
jrichter: You're always merging a parent with a direct child, right?
j-lo: mostly, yes
midori: Yup, and restricting it is fine.
jrichter: Okay. I'll see how that comes out!
jrichter: Is that all, or shall we go have breakfast/lunch/dinner/morphine?
midori is ready to go home ...
jrichter: "or" was the wrong conjunction there too. Haven't had my morphine yet!
jrichter: Okay, bye folks.
midori: bye, and thanks!
midori will send transcripts as usual ...