OE IRC 8Feb07

From GO Wiki
Revision as of 12:14, 4 August 2008 by Midori (talk | contribs) (New page: <pre> 5:25 PM jclark has joined the channel hjd_mgi: Avast maties! jrichter: Hello, folks. jclark: Hi Alex_MGI: Been ages, it seems! jclark: true tberardi has joined the channel ...)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search
The printable version is no longer supported and may have rendering errors. Please update your browser bookmarks and please use the default browser print function instead.


5:25 PM

jclark has joined the channel

hjd_mgi: Avast maties!

jrichter: Hello, folks.

jclark: Hi

Alex_MGI: Been ages, it seems!

jclark: true

tberardi has joined the channel

5:30 PM

midori: OK, it's 5:30. Let the fun begin!

jrichter: Hooray!

midori: Do you have any topics?

jrichter: I'd like to say a quick thing about what I've been doing, and what I will be doing this month. Otherwise, I've got nothing beyond your agenda topics.

midori: OK, go ahead.

Anon205 has joined the channel

jrichter: Because of some funding weirdness in our group, this month I am funded by NCBO, not GO.

jrichter: This has to do with the grant renewal and the way LBL releases funds.

midori 's head starts to spin ...

jrichter: Anyway, because of this, I've been working with Phenote this month, and I will be the entire rest of the month.

jrichter: The good news is that Phenote is built on top of the OBO-Edit sources, so I'm still really working on OBO-Edit 2.0

kchris has joined the channel

jclark has quit the server saying: Ping timeout

Anon205 has quit the server saying: Quit: Anon205

midori: Ah ... that's highly relevant to one of the topics I thought up, namely the 1.1release prognosis.

jclark has joined the channel

5:35 PM

jrichter: I may have some time to do OBO-Edit 1.1 development too. I just can't focus on it 100%.

jrichter: What I've been doing for Phenote will be really useful for us, though. Mark needs the reasoner to have a much smaller memory footprint, and I've gotten some good initial results when I started putting the reasoner data in a SQL database.

jrichter: Anyway, that's all I wanted to say on that topic.

MelissaH has joined the channel

midori: Stuff about the reasoner sounds good.

MelissaH: Hi there.

j-lo has joined the channel

MelissaH: sorry I am late.

midori: Given that you can't work as much on 1.1, can you guesstimate how long before we might be able to release it? (Not trying to crack the whip here; just looking to stay informed.)

tberardi: What

tberardi: (sorry)

Alex_MGI: I'm getting concerned that the 1.1 release is taking a long time, but I guess that can't be helped too much this month.

jrichter: It's hard to say. I can probably clean up all the (real) outstanding tracker bugs in about 3 days, but there's also documentation and turnaround time...

jrichter: ...for testing.

jrichter: Hmmm....

jrichter: How about this?

5:40 PM

jrichter: What if I work on OBO-Edit bug fixes Monday - Wednesday of next week. There will be a new beta release on Thursday, even if I haven't gotten through all the bug reports. That way we can at least make some headway.

midori: OK by me ...

Alex_MGI: Good.

jrichter: Anything else on that topic, then?

jclark: has there been much targeted testing? I've been out of the loop for a while

MelissaH: yeah me too!

midori: very good question!

jclark: also I tried b18 today and it has new features that don't show in my usual layout. I was wondering if you could modify my layout so I can test those. It's the don't search obsoletes box and whotnot.

jrichter: The OBO-Edit 1.1 development I've done since the consortium meeting has been tricky, focused bug fixes. I'm still trying to fix an incredibly painful link filter interface bug, and I just tracked down an insidious reasoner problem. So there hasn't been much room for targeted testing.

jclark: they look cool.

5:45 PM

jclark: sorry I didn't mean you testing I meant us

jrichter: Jen - try picking the "Use default layout" option in the Configuration Plugin. The default layout is a lot like yours, and it has the new buttons.

jclark: ah!

jclark: thanks.

jrichter: I don't know what the ole' users have been up to.

jclark: Does someone in the group know?

jclark: I've been travelling a lot and got out of date

jclark: who has been using the betas?

j-lo: i've been using b18 quite a lot

tberardi: I haven't tested since b17.

MelissaH: I've been using b18 a lot the last week

hjd_mgi: I've been using it for browsing during annotation, and the occasional term. I'll be using it today to fix some defs

kchris: I've used b18 a little, I think I did my last set of edits with it.

jrichter: Just to make it clear, beta18 has a horrible reasoner bug. Don't trust the beta18 reasoner.

midori: I've done some editing in beta18 ... term adds, moves, etc. And I did a very quick check on the cross-product tab.

MelissaH: that would explain things!!

hjd_mgi: I'm still not happen with performance during a save, even with the editor checks turned off

MelissaH: yeah, performance has been terrible for us too/

jclark: so should we wait for that fix and then do some concerted testing?

jrichter: Is there anyone who *doesn't* have performance problems while saving?

jrichter: (I'm trying to figure out why I can't reproduce this on any of my machines)

midori: For me saving is back up to essentially the same speed as 16 and previous (but 17 was unusable).

jclark: I haven't tried I don't think.

midori: ... both with checks on and off -- b18 "pauses" to do the checks, but for the same amount of time as b16

j-lo: my save's fine

kchris: I don't think I noticed anything other than the fact that the checker-thing (sorry spacing on what it's really called) is very slow, especially with so many minor errors

MelissaH: my performance isn't so much save as it is opening many paths.

midori: (Yes, I actually timed it. I'm a geek.)

5:50 PM

hjd_mgi: Is there any correlate with mac tiger platform?  I don't remember 16 having any lags

kchris: I did notice that when I did my last edits, which involved adding a lot of paths, that it became VERY VERY SLOW to open paths, but I opened some of the same paths in the official version and it was slow too

MelissaH: it keeps crashing the cell ontology.

MelissaH: because the CL has soooo many paths.

midori: That reminds me ... I've noticed the DAG viewer not updating promptly when I commit a change. (I haven't put in a bug report because I thought John knew about the problem from the Hinxton hands-on session.)

jrichter: So the problem here is the DAG Viewer, yes?

midori: Harold, are you using tiger? Jane and I are.

MelissaH: I think so.

hjd_mgi: I haven't had issues iwth CL; maybe I have an old version (Oct 10 2006)?

hjd_mgi: yes, I'm using tiger

jrichter: And you're actually using the DAG Viewer as you edit, right?

midori: hmm, argues against a big-cat correlation, then ...

midori: John - yes.

MelissaH: We did get a newer version from Minna, not sure if its up on sourceforge yet.

MelissaH: But there aren't that many changes.

jrichter: Here's what I'd like to see...

jrichter: Melissa: Can you send me a copy of the cell ontology that you're using, along with a list of things that I should do to bring about some of the really bad performance you're describing?

MelissaH: happy to.

midori: I just tried a quick test now: I added a term, and when I hit 'create new child', <new term> appeared in the DAG viewer. But when I gave it a name and committed (I've got autocommit off), the DAG viewer continued to show '<new term>'.

kchris: I saw that too, I had to click on some other term to get it to update

5:55 PM

midori: yep

jrichter: The DAG Viewer problem needs to be fixed in OBO-Edit 1.1, it seems. If it's taking so long that it's providing misleading information, we need to do something about it.

midori: Fixing it would be  a Good Thing (tm), I agree.

kchris: I wouldn't say it's taking so long that it's providing misleading info, it just doesn't show anything till it loads, but I agree that fixing this would be great

jrichter: The DAG Viewer tries to calculate the paths in the background, and I'll bet that this is some kind of thread synchronization thing.

midori: That's what you said in hinxton too.

jrichter: Back to Harold's problem...

jrichter: Does anyone else have the problem where there's a long pause before saving, even with all the verification checks disabled?

midori: I don't.

MelissaH: nope

midori: (I also no longer have the beta17 problem, which was that save progress was agonizingly slow.)

hjd_mgi: I will actually time it today when I use it to make some def changes; I'll post it to the group

jrichter: Harold: once the save gets going, the speed is fine, yes? It just takes a long time to start?

jrichter: Also for Harold: You're saving to a local disk right? Not an NFS partition or some kind of virtual network disk?

hjd_mgi: Once the screen actually comes up with where and how to save it, it's fine; and yes, it's a local disk, not virtual

midori: I just tested, and got no detectable pause (ten 10 sec to save).

hjd_mgi: ok, let me try now

6:00 PM

jrichter: Okay. Harold, can you (once again) zip up your config directory and send it to me. Also, please open the verification plugin configuration screen and send me a screen shot.

tberardi has to go to her 10 am meeting, will catch up via transcript

jrichter: Bye Tanya.

midori: bye Tanya

midori: My test did remind me of one other thing:

Alex_MGI: I have only a 10-sec pause myself.

hjd_mgi: ok, 10-12 seconds from uinsg file, save as (all verification off)

midori: beta18 keeps resetting to "write current id rules", even with save profiles that have been set to "don't write id rules".

jrichter: Midori - That's just an interface thing?

midori: Jane has been contemplating what to put in a bug report about the problem (under what circumstances it resets, etc.)

midori: John - not a-tall sure I understand the question.

jrichter: It saves correctly if you remember to set it back to "don't write id rules"?

j-lo: oh yeah - it doesn't reset every time, just sometimes

midori: Oh, yes -- it does respect what the pulldown says. It's just that what the pulldown is set to changes without human intervention.

j-lo: I think it's when you save with a filter, and then then when you turn the filter off, that menu re-sets

j-lo: perhaps

6:05 PM

midori: That can't be it -- I'm having the same problem, and I haven't saved with a filter in ages!

j-lo: not that then

jrichter: This deserves a bug report, in any case. If you can learn more about how to make it happen, we can update the bug report later. I'll see what I can do to track it down on Monday.

midori: I also think I'm not having the problem all the time ... I wonder if it's how I quit OBO-EDit (I often use command-Q, once I've got all the settings the way I want 'em).

hjd_mgi: I had a similar problem a while back with my dbxrefs, and noted that it happend when I would switch back and forth between 1.002 and 1.1betas.'

hjd_mgi: Just timed with verification on: 2.5 minutes + for the warning found screen to appear'

midori: oh wow -- mine's more like 20-30 seconds.

jrichter: Electromagentic signals move a wee bit slower up in Maine.

hjd_mgi: it's the cold

jrichter: Should we move on to Chris's paper responses?

hjd_mgi: yes

midori: OK

6:10 PM

MelissaH: sure

jrichter: I'm not sure what the process is at this point. Would a more experienced publisher talk us through?

MelissaH: What did the editor last tell us, to resubmit with changes, or to submit elsewhere?

MelissaH: Its been so long that it will be a whole new review.

midori: Resubmit with changes within 90 days ... that was on Nov. 14th (who can do the calculation quickly -- are we too late yet?)

kchris: about 3 months, we probably have a few days

Alex_MGI: Just about

jrichter: We have < 1 week, I think.

MelissaH: Then ASAP, we send back with Chris suggested changes and comments. I can help with this.

6:15 PM

Alex_MGI: If you need to add another page, Bioinformatics may allow this but assess a page charge.

MelissaH: His comments were very helpfuland couldn't have come from any of us I don't think.    

midori: I just added up ... if I got it right, we have until Monday.

jrichter: !

jrichter: Great.

midori: from the email: "The system will automatically remove the revision option if a revised paper has not been submitted within 90 days of the original decision. If this happens, a new submission will need to be made."

jrichter: Okay, I've got the latest drafts on my machine, so I'll make the actual changes to the paper and send it around to the working group for comments tonight.

Alex_MGI: A new submission might generate nicer reviews.

Alex_MGI: Or not.

jrichter: Can someone forward Chris's email to me? I don't seem to have it in my inbox.

jrichter: ?

hjd_mgi: If the issues are addressed to the satisfaction of the editor and reviewers, then it would be a definite accept. Otherwise, we start from scratch

midori: on the way ...

MelissaH: I think (hope) the editor will take into account Chris' comments about that (ugly)reviewer.

jrichter: Everybody MUST look at the draft I send around tonight, because we're out of time. I'm resubmitting tomorrow morning no matter what.

jclark: cool app corroborates midori's counting: http://www.jeffpalm.com/dayssince/

jrichter: Midori: Can you be available on Skype tomorrow until around 6pm (your time)? In case I have any questions about the submission process?

6:20 PM

midori: I think so ... I'll keep you informed if I have to go away from skype between, say 3pm and 6pm.

jrichter: You know, now that I look over Chris's email more carefully, I think it needs to be edited before we can submit his comments to the journal (to take out some of his snarkier insider comments, say).

MelissaH: definitely.

jrichter: Would anyone like to sign up for the task of drafting the edited response?

MelissaH: I can help prepare the reply to the editor.

Alex_MGI: But in general, his comments are quite good.

MelissaH: Why don't I take a first stab, and then Midori can edit?

midori: ok

jrichter: Sounds good. Would you be able to send out your draft by tonight, Melissa?

midori: That would let me deal with it tomorrow morning, UK time.

MelissaH: (it just took 65 sec to load 638 paths to 'gamma-delta intraepithelial T cell'

MelissaH: yes

jrichter: A quick style question - do we present our response to the editors as "an edited version of Chris Mungall

jrichter: 's responses" or...

jrichter: do we just present them as "this is what we, the authors, have to say in response"?

MelissaH: You can sign it yourself, but say the authors respond....

kchris: I think it should be from we the authors

hjd_mgi: It should apply to all of  the lised authors

hjd_mgi: llisted

hjd_mgi: listed   #!!#%%

jrichter: Consensus! Sweet.

6:25 PM

MelissaH: You are the corresponding author, but all authors are responsible for all aspects of the paper.

midori: I agree ... the editors will care about (1) what it says and (2) the fact that all authors are behind it.

MelissaH: (held accountable)

Alex_MGI: (Melissa, I just loaded that term and it was almost instantaneous, with the reasoner on)

MelissaH: on b18?

MelissaH: this is happening on two of our windows machines here.

jrichter: Any other issues before we go? I now feel the crushing weight of a billion urgent tasks.

Alex_MGI: on b18, latest publically available Cell Ontology

midori: In that case, our work is done.

MelissaH: hmmm, my reasoner is off, anything else?

MelissaH: one thing is that I have a global renderer on with a bunch of terms marked as a cell slim category.

Alex_MGI: It was the same with the reasoner off.

jrichter: Melissa - It's probably the fact that it's Windows. I think Windows may be having a problem with the way the DAG Viewer does its threaded screen updates.

Alex_MGI: No global render here.

kchris: I had some similar problems with very slow loading of paths, on Mac OS X, and my term only had 214 paths

jrichter: I'll see what I can do about this on Monday. I'm sure it's the threading, and I already have mapped out some strategies to fix it.

midori: I've also had paths load slowly (tho I haven't had anythingake quite as long as that cell term).

jrichter: Now I must go. I've got a ton of stuff to do.

Alex_MGI: Bye.

jrichter: Goodbye, gentle working group. See you in two weeks.

midori: OK, see you all in two weeks!

jclark: good luck

kchris: bye!

jrichter has quit the server saying: Client exited

MelissaH: ok, I'll send along the revised reply to Midor,John and Chris ASAP.

jclark: bye

midori: I'll send the transcript as usual.

kchris has quit the server saying: Quit: I'll be back!

6:30 PM

Alex_MGI has quit the server saying: Quit: Quitting!

MelissaH: btw, sorry all that I have been away in CARO land for the last few months....shirking my oboedit duties.

jclark: bye

jclark has quit the server saying: Quit: jclark

MelissaH: see ya.

hjd_mgi:  - happy

hjd_mgi: good bye all

hjd_mgi has quit the server saying: Quit: Chatzilla 0.9.77 [Firefox 2.0.0.1/2006120417]

MelissaH has quit the server saying: Quit: MelissaH