[00:06] wgrant: beuno told me earlier to, so why not? [00:08] wgrant, what's the bug? beuno is currently working on a fix for the questions page and a translation page. See https://dev.launchpad.net/LaunchpadTestPlan/2.2.6 [00:10] matsubara: Bug #392333 and bug #392335 [00:10] Launchpad bug 392333 in malone "List of related questions broken by sprite changes" [Undecided,New] https://launchpad.net/bugs/392333 [00:10] Launchpad bug 392335 in malone "Subscriber list on bug action page has broken icon" [Undecided,New] https://launchpad.net/bugs/392335 [00:12] thanks wgrant === mthaddon is now known as afk === afk is now known as mthaddon === mthaddon is now known as afk [00:23] wgrant, what are bug action pages? [00:26] matsubara: Something that is going to send email. [00:26] Let me find an example.. [00:27] +addbranch is one. [00:27] wgrant, thanks [00:28] Hi guys, I keep having recurring problems with LP with a "Sorry, there was a problem connecting to the Launchpad server." message keep coming up, it's not really a *problem* as such, more of an annoyance, just making you aware in case there is a developing server problem :) [00:30] xteejx: launchpad.net, bazaar.launchpad.net, or edge.launchpad.net? [00:30] wgrant, standard lp.net [00:31] xteejx, do you have the URL for the page? [00:31] sure [00:31] https://launchpad.net/ubuntu/+bugs?assignee_option=none&field.assignee=&field.component=1&field.component=2&field.component-empty-marker=1&field.has_no_package.used=&field.has_patch.used=&field.importance%3Alist=Undecided&field.omit_dupes=on&field.omit_dupes.used=&field.owner=&field.searchtext=&field.status%3Alist=New&orderby=-datecreated&search=Search&start=13200 [00:33] xteejx, wfm. do you get those errors always on that page? [00:35] no just randomly, i am triaging old bugs with assigned packages, and working in reverse order to get rid of the old ones, so maybe its because i keep refreshingn after doing 3 of them. Its not really a problem I'll just try later, just thought I should let you guys know in case there's a developing problem thats all :) [00:41] matsubara: I can reproduce, but only around ~1/20 failed for me. [00:42] (not that same URL, of course) [00:42] It takes a very long time, then dies with the usual 502. [00:43] wgrant, I suppose you don't get an OOPS for that, do you? [00:44] or is it that mis-aligned page? [00:44] matsubara: No, it's Squid or Apache, not the app itself. [00:44] It is the misaligned page, yes. [00:44] ah ok. those are trickier to debug [00:45] They are. [00:45] I'll see if the headers give anything useful... [00:47] Hmmm. [00:51] matsubara: The 502 is from Apache, and has no Via header. There's nothing else useful in the headers. [00:57] wgrant, I'm pretty sure we have a bug reported about it but I can't find [00:58] matsubara: About what? It looks like an appserver is dead. [00:58] wgrant, about something going on between apache and the app server and the user getting the "sorry, can't connect" page [00:59] matsubara: Ah. [00:59] I remember adding a dump from live http headers to the report, but I can't find it :-( [01:03] Isn't it more likely that an appserver is just hung? [01:07] wgrant, could be, but if it was the losas would be here already [01:08] matsubara: Ah... === matsubara is now known as matsubara-dinner [01:43] matsubara-dinner: It looks like appserver C is broken. [01:50] hi. I have a question about python-launchpad-lib: Is it possible to search for bugs created after a specific date? e.g. (fakecode): "buglist = launchpad.distributions['ubuntu'].searchTasks(tags="apport-package", bug_creation_date>"05012009") [01:51] Ampelbein: Not directly, but you can order by creation date and stop when you see one early enough. [01:52] wgrant: nice suggestion, haven't thought of that. thanks! === matsubara-dinner is now known as matsubara [02:08] wgrant, how can you tell? [02:09] matsubara: Well, you might notice a few keyboard-mash 404 OOPSes when you look at the reports tomorrow... [02:12] wgrant, what do you mean by keyboard-mash? [02:13] hmm just got one "Please try again" myself [02:13] matsubara: I caused a few dozen 404s with a Launchpad referrer, which gave me a few OOPSes from all appservers except C. [02:17] hey guys with bzr or svn any way to find the number of modified lines [02:17] svn diff | grep "^+" | grep -v "^+++" | wc -l gives you added lines [02:18] svn diff | grep "^-" | grep -v "^---" | wc -l is removed [02:18] but i need modified lines and unmodified lines [02:18] How do you distinguish modified vs. a removal and an addition? [02:19] the --- is removed and +++ is added [02:20] diffstat may be more useful for you [02:20] diffstat? [02:21] Give diffstat a diff, and it will spit back the diff's stats. [02:22] diffstat? [02:22] what is that [02:22] svn diffstat [02:22] ? [02:23] thewrath: svn diff | diffstat [02:24] here is my command right now [02:25] svn diff -r 6:7 mikesats | grep "^-" | grep "^---" | wc -l [02:25] for removed lines [02:25] so i would do svn diff -r 6:7 mikesats | diffstat | grep "^-" | grep "^---" | wc -l ? [02:26] No. Have a look at diffstat's output. [02:26] svn diff -r 6:7 mikesats | diffstat [02:26] ? [02:27] thewrath: Yes. [02:27] wgrant: diffstat doesn't give modified counts [02:27] thumper: Because there's no such thing as a modified line. [02:27] what do you mean [02:27] wgrant: you can do more analysis to determine it [02:27] but most tools don't [02:27] matsubara: So, can anything be done about lpnet3 before London wakes up? [02:28] thumper: Doing such analysis is probably crazy. [02:28] not necessarily [02:28] I've seen it done [02:28] kdiff3 was good at that [02:28] thumper: have any pre defined scripts for that [02:28] thewrath: no [02:28] wat is kdiff3? [02:28] kde based gui diff tool [02:28] still in respository? [02:28] I don't think it made the leap to kde4 though [02:28] perhaps [02:29] so right now i can only find delete and added then [02:29] ? [02:29] there is no way to identify modified lines [02:30] diffs don't work like that. [02:30] ok [02:30] The application would have to guess which were modified and which were replaced. [02:30] And I don't like applications that guess. [02:30] wgrant: added and removed lines are guesses too [02:30] so added and removed are the two i can figure out with --- and +++ [02:31] lifeless: Mm, true. [02:31] thewrath: so as I said before, why are you doing this. Not 'who wants you to do it'. *why* [02:32] lifeless: is for work [02:32] we have several projects that we use svn [02:32] and we want to see productivity [02:32] dont ask me they asked me if it was possible and i said i would ask around and look into it [02:33] that answer you lifeless lol [02:34] its a more useful answer, yes [02:34] lolk [02:34] we use another tool to calculate this all and we have to download each revisiont od o it [02:34] I suggest bzr-stats, or diffstat, which gives different answers [02:34] if you want 'unmodified lines' you'll have to download every revision [02:34] ok [02:35] wgrant, I think we'll have to wait for London to wake up. it's not a major breakage to justify waking people up. [02:35] matsubara: if the appserver is still in the rotation, 1/8 requests will error, no? [02:35] wgrant, I looked at our graphs and you seem to be right. looks like lpnet3 is down [02:36] matsubara: That's what I suspected. It's a bit messier than previous issues of this kind, as the timeline view spews alert()s when a request fails. [02:36] so run bzr diff | diffstats? [02:36] i really need it for svn but we can see lol [02:36] well start with diffstat then [02:37] lifeless, shouldn't apache just take it off of rotation? [02:37] personally though, if you have an existing tool, use it ;) [02:37] matsubara: Not unless it's really dead. [02:37] its a pain bc we have to download every revision [02:37] matsubara: In this case it looks like it's just hng. [02:37] matsubara: pound will, if its not accepting sockets [02:37] lifeless: so what the command is bzr diffstat? [02:37] thewrath: no, 'diffstat' [02:37] so just run diffstat inside my revision? [02:38] no [02:38] you are using svn, so svn diff -r 4:5 | diffstat [02:39] on my stuff i am using bzr [02:39] so i am goign to relate to both [02:39] so bzr diff -r 4:5 |diffstat [02:39] ? [02:39] bzr diff -r 4..5 | diffstat [02:40] i get command not found [02:40] then you haven't installed it [02:40] diffstat? [02:40] i have to install that portion of bzr [02:40] It's not bzr. [02:40] It's diffstat. Entirely independent. [02:40] oh ok [02:40] It works on any diff. [02:40] ah i see [02:41] It doesn't care where it came frm. [02:41] ah i see [02:42] i can run it as a linux emulator command? [02:42] well in a linux emulator [02:43] hello people [02:43] A Linux emulator? [02:43] You mean cygwin? [02:43] yea [02:43] I speak portuguese :s [02:43] thewrath: Cygwin should have an option to install it. [02:43] ig ij [02:43] *oh ok [02:43] what is cygwin ? [02:44] cocacool: A Linux-like environment for Windows. [02:44] ok [02:46] what things does it give me wgrant and lifeless? [02:46] thewrath: What things does what give you? diffstat? [02:47] diffstat [02:47] thewrath: It gives you nice stats on additions/removals in each file in the diff, and in total. [02:47] Without grep and wc -l. [02:48] k [02:50] bzr diff -r 6..7 mikesats | diffstat? [02:50] Something like that. [02:50] Just get a diff into diffstat. [02:50] ok [02:50] u were saying about getting the modified and unmod lines that you would have to download each version? [02:55] wgrant: i get when running bzr diff -r 5..6 mikesats | diffstat [02:55] ipsubnetcheck.pph | 84 ++++++++++++++++.... [02:55] 1 file change, 84 insertations [02:55] that is all ig et [02:56] That's correct. [02:56] says nothing about removed lines [02:56] thewrath: Try diffstat -m [02:57] no such option as -m [02:57] Sigh. [02:57] i have a book at home with options for that [02:57] thewrath: Does that diff actually have any deletions? [02:58] no removals [02:58] Then why exactly would it say anything about removed lines? [02:59] wat -m or -r [02:59] Neither. [02:59] You complained that diffstat didn't tell you anything about removals. [03:00] ok [03:00] That's not surprising, given that you just said that the diff didn't have any removals. [03:01] k [03:01] 6 & 7 [03:01] Pardon? [03:02] hold on [03:02] wats a good free image upload/hoster [03:02] nvm [03:03] wgrant: that is version 6 n 7: http://img32.imageshack.us/img32/894/20090625220152.png [03:03] thewrath: That looks fine to me. [03:03] ok [03:04] wat does the 186 signify [03:04] 186 changes in that file. [03:04] 186 additions and removals? [03:04] wat doe hte + and - means [03:05] It looks like it. [03:05] wat doe hte + and - means [03:05] bc there is not a hundred of each [03:05] well a hundred plus [03:05] Presumably it scales the numbers. [03:06] scales ... explain [03:07] On a small diff here, it gives the right number. On yours, with a fairly big diff on one file, it looks like it scales it down to avoid making that line too long. [03:07] So each + or - will in fact represent multiple. [03:09] k === matsubara is now known as matsubara-afk [03:09] no way to find the number of lines that were modified [03:10] That's what -m attempts to do. [03:10] But your diffstat seems to be braindead. [03:11] shoudl i try it in linux enviroment instead of an emulated [03:11] so what bzr diff -m 6..7 mikesats | diffstat? [03:11] or bzr diff -r 6..7 mikesats | diffstat -m [03:11] If that is what you want to do. [03:11] I've been getting Please try again messages all evening, is there something going on? [03:12] cabrey: Yes, one of the application servers seems to have hung. Just refresh, and things should be OK. [03:12] on launchpad, that is [03:12] wgrant, alright, thanks :) [03:12] wgrant: do you know if this is escalated to is? [03:12] IS* [03:12] just making sure it wasnt my end :O [03:12] wgrant: which one the first or the second one [03:12] and do you think we should? it's a kind of bad time in london [03:12] bc you say diffstat but i had that for diff [03:13] bzr diff -r 6..7 mikesats | diffstat -m [03:13] or [03:13] bzr diff -m 6..7 mikesats | diffstat [03:13] its the second nvm [03:13] okay now found that [03:13] poolie: It was a little while ago deemed insufficiently important to escalate, given that it's 2am for everybody... [03:13] now i need to find lines that have not been messed with [03:14] (I disagree with that assessment) [03:14] mine [03:14] i meant the first one wgrant [03:14] sorry tired [03:14] and doing work stuff at this hour can be deadly lol [03:14] but the nice thing is that i get to go in a little later i think tomorrow [03:14] about an hour later lol [03:15] thewrath: You should use whichever one both works for you and is what you want. [03:15] well bzr diff -m ... gives me errors but the other way doesnt [03:15] bc -m is for diffstat and not diff [03:15] wgrant: i've only seen one failure this morning in light use [03:15] is there any way to find the numnber of unmodified lines [03:15] i guess there's several app servers, like 8? [03:15] poolie: There are 8. [03:15] and the load balancer is meant to cope if they're hung [03:15] lpnet3 is dead. [03:15] :) [03:16] poolie: pound will knock them off the rotation if they are not accepting connections. [03:16] It doesn't care if they time out. [03:16] you certainly know a lot about it :) [03:16] poolie: Oh yes... ran into the same problem with the application that I manage. [03:16] poolie: You won't see many of the timeouts, because you're lp-beta-testers... [03:17] And edge is working fine. [03:17] * poolie reads the logs [03:17] ah of course [03:17] i probably only saw one when i entered an un-edgey url [03:17] is there a way to find the number of unmodified lines [03:18] thewrath: I'm sure you can work that one out. [03:18] you think something with diffstat? [03:18] No. [03:18] any suggestions [03:19] I've filed a bug on them being left in rotation [03:20] i wonder if we could get a non-LO SA to thump them? [03:20] poolie: see my mail to launchpad [03:20] as of 2 minutes ago [03:20] ! for mod lines is something partaiing to diffstat only [03:20] when i do my long grep it does not work [03:21] poolie: they're all in london too, afaik === vorian is now known as stevie [03:21] oh of course [03:21] single point of failure win :) [03:21] what edit request did i have that got forwarded to ubottu? [03:21] I'm not sure the regular sysadmins are in London [03:21] I know the LOSA's are [03:22] they did say "IS sprint" [03:22] it would be nice if https://blueprints.edge.launchpad.net/launchpad-foundations/+spec/system-status-page was implemented [03:22] poolie: yeah, I'm just not sure is all [03:22] oh nvm [03:31] thewrath: anything matching "^!is$" is interpreted as an edit request [03:32] i ran this bc it was similar to everythign else for added and removed [03:32] bzr diff -r 6..7 mikesats | grep "^!" | grep -v "^!!!"| wc -l [03:32] i get base: !": event not found [03:33] use single quotes [03:33] bash interprets ! specially [03:35] ' or ` [03:35] ' [03:35] ` is generally called a "back-tick" [03:36] i get 0 which is not right [03:39] ` is back tick and ' is single quote? [03:39] yep [03:40] bzr diff 0r 6..7 mikesats | grep -v '^!' | wc -l gives 414 which is not right [03:40] bc when i run diffstat it says 94 [03:41] make sure you use egrep to make it interpret regex [03:41] http://img29.imageshack.us/img29/465/20090625224104.png [03:41] i get 0 using egrep [03:42] any final thoughts before i go to bed [03:42] '^!' will match only when '!' is at the start of the line [03:42] right [03:43] that is modified isnt [03:43] ^ is the start of the line, $ is the end [03:43] [^...] is negation [03:43] it should say if a line is modified at the beginner [03:44] it just means "the start of the line" [03:44] *should say if a line is modified at the beginners [03:44] like a line that was added or removed has a + or - at the beginning of a line [03:44] ! should be at hte beginning of hte line if that line was modified [03:44] Error: I am only a bot, please don't think I'm intelligent :) [03:45] an unmodified line will start with a space [03:45] wat about modified [03:45] with a + or a - [03:45] + is added line [03:45] - is removed [03:46] yes, and a space is unmodified [03:46] ! is modified [03:46] Sorry, I don't know anything about is modified [03:46] according how it looks with u run diffstat -m [03:46] that's in diffstat, but not in diff [03:47] diffstat interprets the output from diff [03:47] ok so in diff [03:47] + is added, - is remoeve dns 1 blank space is modified all at the beginning of the line [03:47] ?? [03:48] + is a line added, - is a removed line, and a blank space is unchanged [03:48] that is wat i meant [03:49] (note: there are also +++ and ---) [03:49] whcih means wat [03:49] of the original is "Hello World" then the unmodified line in diff is " Hello World" [03:50] s/of/if/ [03:51] savvas: wat does the +++ or --- mean [03:51] It prefixes the filenames of the original and new version of the file being compared [03:51] Well, for most common forms of diff output. [03:52] thewrath: these are explained at: http://en.wikipedia.org/wiki/Diff#Unified_format [03:52] > the original file is preceded by "---" and the new file is preceded by "+++" [03:52] wats the difference between bzr diff -r 5..6 mikesats | grep "^+" | grep "^+++" | wc -l and bzr diff -r 5..6 mikesats | grep "^+" | wc -l [03:53] the first will only match files changed, the second will match that plus all the added/changed lines [03:53] actually, no [03:54] wait, yes ;) [03:54] i was told the first one was to get new lines [03:54] 23:33 neels thewrath, e.g. << svn diff | grep "^+" | grep -v "^+++" | wc -l >> gives you the number of added lines :) [03:54] the first will match only the "+++ new-file" parts === abentley1 is now known as abentley [03:54] the first 'grep "^+"' is also not needed [03:55] yes it is [03:55] shouldn't + be escaped? [03:55] if grep uses extended regex (not sure) [03:55] tsimpson: its necessary bc between 5 n 6 there were only 84 lines of changes and all additions not 88 [03:55] thewrath: what are you trying to do actually? :) === abentley1 is now known as abentley [03:56] savvas: get the number of added, removed, mod and unmod lines from revision 5 to revision 6 [03:56] between 5 n 6 there was only 84 lines of code changes and all were additions [03:56] so i am just testing that out [03:56] for added lines you might want something more like grep '^+' |grep -v '^+++' [03:56] thewrath: '^+++' matches a line starting with 3 '+'s, '^+' matches a line starting with 1 '+', if it starts with 3 then it also starts with 1 [03:57] svn diff | grep "^+" | grep -v "^+++" | wc -l gives me new lines (added lines) [03:57] yeah, the + needs escaping [03:57] svn diff | grep "^-" | grep -v "^---" | wc -l >> [03:58] that line shows removed lines [03:58] what do you mean especaping tsimpson [03:58] ah, -v [03:58] it needs '^\+' and '^\+\+\+' [03:58] because '+' is a regex special character [03:58] BTW... how is this grep tutorial about launchpad exactly? === richard is now known as Guest39203 [03:59] sorry [03:59] it means repeated at least once [03:59] it was carried into grep [03:59] tsimpson: nope that now only brings up 2 which is wrong [03:59] 84 is right [03:59] it does NOT need to be escaped [03:59] tsimpson: Please test your suggestions before statiung them as fact. [04:00] jmarsden: I did [04:00] $ echo '---'|egrep '^+++' [04:00] --- [04:00] echo '---'|egrep '^\+\+\+' [04:00] egrep != grep [04:00] (nothing) [04:00] The command used in the examples is grep. [04:00] Bug 392371 [04:00] Launchpad bug 392371 in launchpad "appservers are not removed from rotation properly" [Undecided,New] https://launchpad.net/bugs/392371 [04:00] bzr diff -r 5..6 mikesats | grep "^\+.... does not wokr [04:01] no escaping is necssary [04:01] here is an example: http://pastebin.ca/1475012 [04:01] ok [04:01] thewrath: You are corect, tsimpson is mistaken and confused. [04:01] diffstat and bzr with grep show the same thing [04:01] right [04:01] ty [04:01] then 1 blank space before is an unmodified line? [04:02] before is at he beginning [04:02] yes [04:02] savvas: wat ist eh === for [04:02] number of files? [04:02] === is the header for bzr [04:03] === modified file 'surl/surl.py' [04:03] --- surl/surl.py 2009-04-27 12:10:16 +0000 [04:03] +++ surl/surl.py 2009-06-11 23:38:41 +0000 [04:03] bzr diff -r 4..7 | grep '^ ' | grep -v '^ ' | wc -l is for unmodified lines? [04:03] No, you'll only see lines that were used to supply context for a change that way. [04:04] how do i find unmodified or modified lines then [04:04] or cant i [04:04] no I think it's: bzr diff -r 4..7 | grep '^ ' | wc -l [04:04] savvas: If I edit one line in a 100 line file, bzr diff will not output all 99 unmodified lines... [04:04] It only shows a few around each change. [04:05]