Evergreen ILS Website

IRC log for #evergreen, 2016-02-19

| Channels | #evergreen index | Today | | Search | Google Search | Plain-Text | summary | Join Webchat

All times shown according to the server's local time.

Time Nick Message
01:09 artunit_away joined #evergreen
06:27 rlefaive joined #evergreen
07:11 rlefaive joined #evergreen
07:41 krvmga joined #evergreen
07:55 ericar joined #evergreen
07:58 JBoyer joined #evergreen
08:25 collum joined #evergreen
08:38 Dyrcona joined #evergreen
09:10 Dyrcona Someone might want to looking into build/tools/relator_map.
09:10 Dyrcona It spit out some errors and left Open-ILS/src/templates/opac/parts/relators.tt2 empty.
09:10 Dyrcona I might Launchpad that later if I think of it.
09:11 afterl joined #evergreen
09:12 tsbere Dyrcona: I think that has more to do with issues with loc.gov than the script
09:12 Dyrcona It very well might.
09:13 Dyrcona I didn't check to see if they were up or anything.
09:13 * tsbere tried to look up a MARC value and the page isn't loading
09:18 mmorgan joined #evergreen
09:19 Dyrcona tsbere figured it out. Need to use https
09:21 tsbere Dyrcona: Or, you know, wait for them to fix the non-https version. Which apparently happened in the past minute or two.
09:21 Dyrcona Yep. Worked this time and there were no changes anyway. ;)
09:30 Dyrcona Yay for git clean -x -f -d
09:33 yboston joined #evergreen
09:43 jeff Dyrcona: that's the git equivalent of "if you don't clean up this room i'm coming back here with a big box and i'm going to throw everything out!"
09:44 Dyrcona Yes, indeed.
09:47 mmorgan joined #evergreen
09:51 krvmga Dyrcona: :)
09:51 krvmga jeff++
09:51 mceraso I did an OS upgrade to all of our servers last night. Reports are running, but the output is not showing anything
09:52 jeff "i want all of these files either committed or stashed, or they're all GONE!"
09:52 mceraso a df -h on the application servers shows all the correct shares
09:52 mceraso I'm not sure what else I may be missing
09:52 Dyrcona mceraso: enough free space to store the reports?
09:53 mceraso Dyrcona: I believe so
09:54 csharp mceraso: when you say the output is not showing anything, do you mean that they get a 404?
09:54 csharp or that the output screen is available and they aren't seeing results for some reason
09:55 mceraso csharp: They are directed to a URL that gives them a message "to try viewing through the new library catalog."
09:55 kmlussier That's special.
09:57 csharp okay, let's rule out NFS... can you go to the directory where the report share is mounted on an app server and 'ls' its contents?
09:58 mceraso Sure thing
09:58 kmlussier berick: In bug 1474566, you mentioned that you were going to test with live data and report back. Did you ever get the chance to test with live data?
09:58 pinesol_green Launchpad bug 1474566 in Evergreen "Credit card API amount paid rounding error." [Undecided,New] https://launchpad.net/bugs/1474566
09:58 * kmlussier is looking at credit card bugs today
09:59 csharp mceraso: note that *I* don't need to see the contents, I'm just curious if it shows the dir listing :-)
10:04 mceraso csharp: It does show the dir listing
10:05 csharp mceraso: ok - and that's true on all (both?) app server?
10:05 csharp servers*
10:05 JBoyer mceraso: You don't have any special apache redirects for old or odd urls do you? The message you quoted doesn't look familiar to me at all.
10:05 csharp JBoyer: that was my next thought
10:05 tsbere csharp / mceraso: I can think of two things. One: Can the reporter *write* to NFS? Two: Is there a bad rewriterule in the apache config.
10:06 mceraso Not sure if this is at all related, but I had trouble restarting the server after I did a reboot
10:07 mceraso I did the OS updates and then rebooted
10:07 csharp mceraso: well, let's try to rule out apache next, then we can see about the server reboot issue
10:07 berick kmlussier: re bug 1474566 -- yes, we're running that code in production
10:07 pinesol_green Launchpad bug 1474566 in Evergreen "Credit card API amount paid rounding error." [Undecided,New] https://launchpad.net/bugs/1474566
10:07 mceraso csharp: Fair enough :)
10:07 mceraso JBoyer: We typically get that message when all the drives were mounted on the app servers
10:08 kmlussier berick: It's ready for a pullrequest tag?
10:08 tsbere JBoyer: For your reference, my reaction to the message was "they configured a 404 document for the entire server and it is including the reports output dir"
10:08 berick kmlussier: yes.  added comment and tag now...
10:08 mceraso I tried mounting all of the drives again, and the problem still exists. Usually a sudo mount -a will remedy the issue
10:08 csharp mceraso: okay, then not NFS
10:09 csharp or at least it doesn't appear to be
10:09 tsbere mceraso: Before we dig too much, is it just *new* reports not showing output, or do pre-update reports also do the same thing?
10:09 kmlussier berick: Great!
10:10 kmlussier Of course, I have no idea how to test it. Not sure how you get to 9.999999 in fines.
10:10 tsbere kmlussier: Floating point weirdness.
10:10 mceraso tsbere: Checking on that now
10:11 mceraso tsbere: I can confirm it is just new reports
10:11 tsbere mceraso: Then your NFS *and* apache are fine.
10:12 tsbere mceraso: I would suggest looking at "Can the machine running clark kent write to NFS" and/or "Are the reports being written with the correct permissions"
10:13 tsbere mceraso: To clarify nfs/apache being fine: On the machine(s) running apache. NFS may *not* be fine where clark is running.
10:13 mceraso tsbere: Gotcha
10:13 csharp might be a file owner mismatch between the NFS share server and the app server
10:14 mceraso Could this have happened with just a server OS upgrade?
10:14 csharp i.e., opensrf on server 1 has a different UID than opensrf on server2 and NFS goes by UID
10:15 berick kmlussier: i had to create a scenario with 40+ transactions each for small amounts (.10,.20, etc.).  it was a pain.  in this case, i think as long as nothing is broken, we're good.
10:15 csharp mceraso: was it an in-place upgrade or a resintallation?
10:15 tsbere mceraso: Could be that the NFS never mounted on the machine running clark, so it is outputting files to the mountpoint instead of to NFS itself.
10:15 kmlussier berick: ok, thanks!
10:15 csharp tsbere: right
10:16 mceraso csharp: just a sudo apt-get update ; sudo apt-get dist-upgrade and a reboot
10:16 mceraso Nothing more
10:16 mceraso But there was a Port 22 Connection Failure when I tried to access the server again
10:16 csharp mceraso: oh, so not a version upgrade then, ok
10:17 csharp mceraso: while it was coming back up?
10:17 mceraso csharp: Yes
10:18 csharp mceraso: you'll get a connection refused if the server is not ready for SSH connections - if that was the only issue you saw, I would probably not worry about it
10:18 Bmagic Would anyone be able to help me turn up a bug report for "item checkin voids all bills but the circulation remains in the other/special"
10:19 csharp Bmagic: I think that's an OU setting
10:20 Bmagic csharp: oh sure, it is, but in the case a library doesn't charge overdue fines, and the item is returned, and the bill goes to $0 - there is a bug where the transaction doesn't get closed
10:20 tsbere Bmagic: The other/special is probably partially tied to the stop fines reason, so checking that may be a clue
10:21 Bmagic I could have swore I or someone else reported this bug
10:23 tsbere Bmagic: An old one relating to LOST items is this: https://bugs.launchpad.net/evergreen/+bug/758982
10:23 pinesol_green Launchpad bug 758982 in Evergreen "2.0, 1.6, Transactions not closing for LOST items once xact has been re-opened for modified billings. " [High,Fix released]
10:24 tsbere Bmagic: Maybe this? https://bugs.launchpad.net/evergreen/+bug/965656
10:24 pinesol_green Launchpad bug 965656 in Evergreen "Close transactions after fines are voided, " [Medium,Fix released]
10:25 Bmagic that one was pushed to master ~2.5
10:26 berick hmm, who can I bug to update the downloads page to point to 2.8.6 ?
10:26 Dyrcona berick: I believe I can do it.
10:26 berick Dyrcona++
10:26 Dyrcona I am releasing 2.9.2 today.
10:36 JBoyer kmlussier: The short answer to how to get to 9.99999 in fines is that you have to do a lot of addition. That bug may explain the 3 cent difference I saw on a CC payment recently. First time I've seen any issues and BAM, 3 issues at once, heh.
10:44 Dyrcona Computers are bad at floating point math, and Perl is even worse.
10:45 Christineb joined #evergreen
10:48 jvwoolf1 joined #evergreen
11:08 vlewis joined #evergreen
11:37 bmills joined #evergreen
11:45 Dyrcona OK. 2.9.2 is uploaded. I'll see if I do have permission to change the downloads page.
11:47 kmlussier Dyrcona: If you don't, let me know. I can do it.
11:47 Dyrcona I don't think I can. What is it I'm supposed to see when I can change it?
11:48 kmlussier You need to be logged into Wordpress. If your login expired, you won't see it. https://evergreen-ils.org/wp-admin
11:49 Dyrcona I am logged in. I thought I was supposed to get some kind of edit option or an additional row on the right column.
11:49 Dyrcona Looks like all I can do is add blog posts.
11:50 kmlussier In that case, yes, you should see an Edit Page option in a black bar along the top.
11:50 kmlussier You also need to make sure you're using https
11:50 Dyrcona Nope. Not there. Guess I don't have the permissions.
11:51 Dyrcona I've got the black bar, no edit page option.
11:51 kmlussier Dyrcona: Yeah, I just looked it up. Looks like you're an "author."
11:52 Dyrcona I thought I was gonna get promoted after last time, but whatevs. :)
11:53 kmlussier I'll update the page, but I'm also thinking that anyone elected as RM should get more powers as part of the job. :)
11:54 kmlussier While I'm there, 2.7 is security only, right?
11:54 Dyrcona Yes, I think so.
11:54 kmlussier A little bird just told me it's beyond security
11:54 kmlussier So I'll remove it.
11:55 kmlussier Or maybe I'll wait until beta comes out, the little bird says.
11:55 mmorgan joined #evergreen
12:06 pinesol_green [evergreen|Jason Stephenson] Forward port 2.9.1 to 2.9.2 upgrade script and server upgrade instructions. - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=39e197a>
12:09 kmlussier Downloads page is updated. I think I checked all the links, but somebody may want to double check.
12:22 mllewellyn joined #evergreen
12:28 jihpringle joined #evergreen
12:29 kmlussier Looks like the upgrade docs need to be updated too. http://docs.evergreen-ils.org/2.9/_upgr​ade_the_evergreen_database_schema.html I don't think I can get to it now, but if nobody else gets to it, I might be able do it later this afternoon.
12:29 kmlussier I think bshum used to take responsibility for that.
12:31 kmlussier Of course, it's something that can be done by anyone who has commit privileges for docs
12:36 kmlussier Calling 0956 and 0957
12:38 pinesol_green [evergreen|Bill Erickson] Porting 2.8.4-2.8.5 / 2.8.5-2.8.6 DB upgrade scripts - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=24cf938>
12:46 pinesol_green [evergreen|Bill Erickson] LP#1474051 Drop unneeded CC payment fields - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=4718bf4>
12:46 pinesol_green [evergreen|Bill Erickson] LP#1474051 PGTAP test checking dropped columns - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=c23f145>
12:46 pinesol_green [evergreen|Bill Erickson] LP#1474051 CRON script for clearing CC numbers - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=a8cb9b9>
12:46 pinesol_green [evergreen|Bill Erickson] LP#1474051 Release/Upgrade Notes on dropped CC fields - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=9247180>
12:46 pinesol_green [evergreen|Kathy Lussier] LP#1474051 Stamping upgrade script for drop-cc-data - <http://git.evergreen-ils.org/?p=​Evergreen.git;a=commit;h=4ae3fa1>
12:46 csharp berick++ kmlussier++
12:49 berick kmlussier++ indeed
13:25 jonadab joined #evergreen
13:25 egbuilder joined #evergreen
13:26 RBecker joined #evergreen
13:26 dkyle joined #evergreen
13:26 pastebot joined #evergreen
13:27 neady joined #evergreen
13:27 ldw_ joined #evergreen
13:27 phasefx joined #evergreen
13:27 tsbere joined #evergreen
13:29 Dyrcona Net split!
13:50 collum joined #evergreen
13:52 collum joined #evergreen
13:54 collum joined #evergreen
13:59 collum joined #evergreen
14:01 yboston joined #evergreen
14:03 collum joined #evergreen
14:25 collum joined #evergreen
14:25 kmlussier @dessert
14:25 * pinesol_green grabs some of yboston's cheese flan for kmlussier
14:25 * kmlussier is still waiting for her cheese flan from yboston. ;)
14:27 Dyrcona heh.
14:27 Dyrcona @brew
14:27 pinesol_green Dyrcona: http://www.firstpersontetris.com/
14:28 Dyrcona @bartender
14:28 * pinesol_green fills a pint glass with Wachusett Blueberry Ale, and sends it sliding down the bar to Dyrcona (http://beeradvocate.com/beer/profile/20/2190/)
14:28 Dyrcona Ooh. Something local...but not quite my favorite.
14:28 jeff not a huge fan of fruit beers.
14:29 jeff but i had a surprisingly tasty beer with nutmeg in it the other day: http://www.beeradvocate.com​/beer/profile/22564/79840/
14:30 * Dyrcona neither. Thought I do like an occasional raspberry lambic.
14:31 Dyrcona I get one with basil sometimes. I think it's called Emotional Rescue.
14:33 Dyrcona https://untappd.com/b/smuttynose-bre​wing-co-emotional-rescue-2-0/1034464
14:33 Dyrcona Apparently, brewed exclusively for my favorite place to get pizza.
14:36 jihpringle joined #evergreen
14:58 * csharp used to have a strong preference for IPAs
15:16 rjackson_isl joined #evergreen
15:28 bmills joined #evergreen
15:31 berick hmm, trouble connecting xul client to a server w/ tls 1.0 disabled (1.1 and 1.2 enabled)
15:33 * berick sees xulrunner 14 supports the old-style tls pref security.enable_tls, but not the new style security.tls.version.max
15:34 jvwoolf1 left #evergreen
15:34 berick .. which may mean 1.1 and 1.2 are not even supported by version 14
15:35 Dyrcona Could be, and so another reason to get away from it.
15:39 * dbs sighs
15:39 berick and too bad tls 1.0 is considered unsufficient for pci compliance
15:39 * berick knew this day would come, hoped it would come later
15:40 berick i'll open a bug so we have this documented
15:42 kmlussier :(
15:45 JBoyer I'm not trying to be negative, but accepting CCs through the client always skeeved me out. The way I read the PCI requirements the library network where the number is entered would "count" as far as compliance, and if you're part of a consortium with widely varying IT staff... D:
15:48 Dyrcona JBoyer: Yeah, we've pretty much told our members' staff that you do not want to touch credit cards at the circ. desk.
15:50 berick I mainly note the PCI aspect as further proof (to those what have funding power) that we need to continue aggressively killing xulrunner
15:51 berick fortuantely, we don't do CC payments via the staff client either.
15:52 kmlussier I wonder if anyone is doing CC payments through the client. The consortia here are not, but I know there has been talk at one library about using a card swipe to take payments at the desk.
15:52 Dyrcona It looks like XulRunner has been discontinued as of last October/November.
15:54 Dyrcona We're not doing CC payments at all. Some members want to, some don't seem to care, others have issues with the town and revenue.
15:54 jlitrell joined #evergreen
15:59 collum joined #evergreen
16:03 Dyrcona sandbergja2: Have you been getting my emails?
16:06 kmlussier OK, it's been more than three months since the hack-a-way, and I'm now just getting to my action item to start a larger community discussion on search.
16:06 kmlussier Actually, I noticed that action item didn't make it to the hack-a-way notes, but I remember it in my head, so I'm going with it.
16:07 kmlussier I was thinking I might start by having 1 or 2 focus-group-type discussions here in IRC. Would there be any objections to that?
16:08 berick certainly no objection from me
16:09 Dyrcona None here, either.
16:09 kmlussier As a reminder, the discussion will ultimately lead to identifying what the community would like to see in search and maybe help answer some questions of what we mean when we say we want "relevant results."
16:10 kmlussier My notes also said I was supposed to fiinish this by the end of the 2015 calendar year, so I'm obviously running behind schedule.
16:10 * kmlussier goes about setting up Doodle polls based on the lack of objection from two people paying attention at 4 p.m. on a Friday. :)
16:11 kmlussier But, first, a Doodle poll for Bug Squashing Day.
16:11 _bott_ Since someone mentioned CCs...  Our City recently switched to a company called Point & Pay for swipes and manual entries.  We may have to replace our Authorize.Net gateway with their processing.  Anyone out there ever worked with the Point & Pay API?
16:12 Dyrcona Schedules? We don't need no steenkin' schedules.....
16:13 kmlussier Speaking of schedules, the 2.10 release is scheduled for my birthday.
16:13 berick ah, yes, so much power granted to those who show up
16:13 kmlussier @eightball will I get a perfect Evergreen system that serves me coffee for my birthday?
16:13 pinesol_green kmlussier: I doubt it very much.
16:14 kmlussier berick: I always time my big "need feedback" questions for Friday afternoons.
16:14 berick kmlussier: you should work in politics
16:14 berick oh wait, i guess you kind of do
16:15 kmlussier _bott_: I've only worked with PayPal.
16:15 kmlussier berick: Yeah, I was just about to say that. :)
16:15 csharp politics--
16:15 _bott_ I'd never heard of Point & Pay.  Ah, the power of the lowest bidder...
16:16 csharp heh
16:26 dbwells kmlussier: Your b-day is March 17?
16:26 kmlussier dbwells: yup
16:27 dbwells kmlussier: me too :D  Always nice to have so many people celebrate and make rivers green and stuff.
16:27 Dyrcona _bott_: I almost read it as "Point and Pray."
16:27 dbs Punt and Pray
16:27 Dyrcona dbs++
16:27 berick hah
16:28 afterl left #evergreen
16:28 kmlussier dbwells: Really? Glad to share a bday with you!
16:29 dbwells kmlussier: thanks, and to you as well!
16:55 abowling1 dbwells: kmlussier: was just reading on hash collisions today, and the article alluded to the "birthday problem." are you familiar with it?
16:56 abowling1 it's an intriguing counterintuitive math problem
16:57 kmlussier abowling1: No, I've never heard of it.
16:57 * kmlussier consults Wikipedia
16:58 abowling1 in a room of 23 randomly selected people, the odds that any two of those people share are birthday is 50%. in a room of 50, it's 97%
16:58 berick it is very mind bendy
16:58 abowling1 in a room of 70, it's 99.9%
16:58 abowling1 berick: indeed
16:58 abowling1 far from what you'd expect, but i had a slight "a-ha" after seeing the math
16:59 dbwells The best trick I know when dealing with compound odds is to realize there is often only one way for something to "not" happen.  Then it becomes much plainer to see, IMHO.
17:00 abowling1 the birthday conversation just seemed too much a happy coincidence, as i had just read it not an hour ago
17:02 dbwells (ignoring leap years) So the odds of two people *not* sharing a birthday are 364/365.  Now we have two days covered, so the odds of a third not hitting those are 364/365 * 363/365.  Keep going, and its easy to see how the odds of it not happening diminish rapidly. :)
17:02 kmlussier What are the odds of learning about the birthday problem just an hour before then witnessing two mutual acquaintances discover that they have the same birthday? ;)
17:02 dbwells hah
17:03 kmlussier I can't look at match at 5 p.m. on a Friday
17:03 kmlussier math, even
17:05 abowling1 kmlussier: 1.) or not learning the odds, to dbwells point? :) 2.) i'm in the same friday mode, which is why i typed it
17:06 kmlussier heh
18:18 rlefaive joined #evergreen
18:39 * bshum whistles a happy tune to himself as he moves LP targets about
18:39 bshum berick++ Dyrcona++ # releases!
18:44 csharp kmlussier: FYI - the Monday - Wednesday offered for the Bug Squashing Day conflict with Code4Lib - Terran and I are going and I know of at least one other EG person who is planning to go - just FYI - hopefully the Friday before will work
18:47 bshum Added new milestones for 2.9.3 and 2.8.7 (with target for next march 16, but who's counting...)
18:50 rlefaive_ joined #evergreen
19:44 jeffdavis Is anyone actually preparing a patch for bug 1517298? It's listed on the 2.10 roadmap, but the ticket is unassigned and I don't see a branch for it in the working repo.
19:44 pinesol_green Launchpad bug 1517298 in Evergreen " Catalogue should support Piwik, a privacy-sensitive alternative to Google Analytics" [Wishlist,New] https://launchpad.net/bugs/1517298
23:39 kmlussier csharp: Yeah, Terran emailed me about that. Friday's looking good so far.
23:39 kmlussier Of course, we've only gotten 4 responses.
23:41 bshum jeffdavis: I wondered if your example would help us along with that bug, but yeah seems quiet for the moment.
23:42 bshum jeffdavis: I figured it would not be hard to add rigging for variables for piwik config and include the right snippet example with instructions to grab what you need and paste in from a piwik install.
23:42 bshum Just a matter of header or footer, maybe? :)
23:43 bshum "The higher, the fewer!"

| Channels | #evergreen index | Today | | Search | Google Search | Plain-Text | summary | Join Webchat