| Time |
Nick |
Message |
| 07:14 |
|
collum joined #evergreen |
| 08:39 |
|
mmorgan joined #evergreen |
| 08:56 |
|
dguarrac joined #evergreen |
| 09:08 |
Bmagic |
can anyone else confirm that installing Evergreen using OpenSRF main brainch fails at autogen? |
| 09:35 |
|
redavis joined #evergreen |
| 09:36 |
|
redavis_reloaded joined #evergreen |
| 09:40 |
redavis_reloaded |
sleary, if it could be in a separate branch, that'd be cool. |
| 09:40 |
|
redavis joined #evergreen |
| 09:58 |
Bmagic |
nevermind on the OpenSRF main brainch, my problem was something else |
| 10:06 |
redavis |
Just throwing a note in here. Feature freeze for 3.14 is this coming Tuesday. There are currently 14 LP tickets with signoffs waiting to be committed. I appreciate the flurry of activity since slush started and am hopeful to get at least a significant chunk of the remaining signoffs into main. They're listed here - https://docs.google.com/spreadsheets/d/1nVuknGBCOSUxBFId0bORVrJYw4Un0mD0kqr67d_AJ84/edit?usp=sharing |
| 10:07 |
sleary |
redavis++ |
| 10:08 |
redavis |
Additionally, there are 23 tickets that have pull requests but no signoffs. If anyone is of a mood and place in their life where they want to get and of those branches up on public test servers, I (and likely others) are around to test for sign offs. |
| 10:15 |
|
mmorgan joined #evergreen |
| 10:16 |
redavis |
(but I might go for a bikeride...just saying...but definitely available for testing before or after) |
| 10:35 |
Bmagic |
here's my issue: |
| 10:35 |
Bmagic |
psql:reporter-schema.sql:579: ERROR: type anycompatible does not exist |
| 10:40 |
Bmagic |
looks like maybe* the reporter security patch is missing some sql pieces to define the type for PG ? |
| 10:41 |
JBoyer |
any* types belong to Pg, likely the spelling of something changed between dev and the version you're using. (anycompatible sounds vaguely familiar, this or something like it may have come up recently) |
| 10:43 |
JBoyer |
There's more of these than I realized: https://www.postgresql.org/docs/current/datatype-pseudo.html |
| 10:44 |
Bmagic |
JBoyer++ # if you use that documentation page and change the version to Postgres 10, you'll find that "anycompatible" is not listed |
| 10:44 |
JBoyer |
yup. |
| 10:44 |
Bmagic |
starts on PG 13 |
| 10:44 |
Bmagic |
Are we raising the miniumum PG for EG 3.14? |
| 10:45 |
redavis |
Was just about to ask that |
| 10:46 |
redavis |
Bmagic JBoyer, should that be something sent to the devs list since not everyone is in here at the same time? |
| 10:46 |
JBoyer |
Probably not intentionally, no. That one may require some more study to see if it's easy enough to make work with 10+ or if there's need to have separate scripts by version or something. |
| 10:46 |
JBoyer |
yeah. Bmagic since you've got eyes on the issue could you put your findings on the list? |
| 10:50 |
|
Christineb joined #evergreen |
| 10:50 |
redavis |
PG10 is out of support (not sure if that actually matters a lot), so it seems like, if not with 3.14, then with next major release there should be a discussion to "force" PG upgrade to some recommended newer version. |
| 10:54 |
Bmagic |
personally, I think it's reasonable to remove support for PG10 after version X of Evergreen, as we've done before. Plus PG10/11 is not supported by the PG community. It will start to fade into history as newer OS's move on. |
| 10:54 |
Bmagic |
I'll write the email :) |
| 10:54 |
redavis |
Bmagic++ JBoyer++ |
| 10:57 |
JBoyer |
Certainly, but we probably need to actually do that on purpose and *before* things break because old. |
| 10:57 |
JBoyer |
Bmagic++ |
| 11:00 |
redavis |
From my perspective, it'd be lovely to say "upgrade PG" with this release, but I'm not sure how many sys admins would lose their minds. IIUC, the reports security patch causes an issue on PG10 because it's invoking that pseudo-type that doesn't exist until pg13? |
| 11:10 |
Bmagic |
redavis: yes, that seems to be the case |
| 11:10 |
redavis |
Bmagic++ |
| 11:21 |
pinesol |
News from commits: LP#1828456: use open-ils.auth.login for AuthProxy native login <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=d74e3139d9a88492d03f0ce34cccb7e02bd42ee0> |
| 11:22 |
redavis |
Noice! |
| 11:24 |
|
sandbergja joined #evergreen |
| 11:29 |
sandbergja |
Is the live_t/lp1468422_passwd_storage.pg pgtap test failing for anybody else on main? It's failing for me on postgres 13, 14, and 15. Haven't checked 11 and 12, since as pointed out earlier, those have bigger problems. |
| 11:36 |
|
jihpringle joined #evergreen |
| 11:36 |
sandbergja |
also failing on postgres 16 |
| 11:39 |
sandbergja |
Also, if I may be so gauche as to plug a launchpad bug: bug 2055796 would add automatic checks on versions 11, 12, 13, 14, 15, and 16 of postgres, so that we could catch bugs and incompatibilities earlier. |
| 11:39 |
pinesol |
Launchpad bug 2055796 in Evergreen "Have github actions run pgtap tests for us" [Undecided,New] https://launchpad.net/bugs/2055796 |
| 11:40 |
sandbergja |
If you log in to github, you can see sample output here: https://github.com/sandbergja/Evergreen/actions/runs/10962113742/job/30440600203. If you click on, say, "Run PgTap tests (11)", you can see that it couldn't find the reporter schema due to the incompatibility with postgres 11, for example. |
| 11:41 |
|
Dyrcona joined #evergreen |
| 11:42 |
sandbergja |
I don't know if we should add it when we have problems on literally every postgres version though... |
| 11:45 |
* Dyrcona |
just checked the logs to see what was missed. |
| 11:46 |
Dyrcona |
I'm OK with bumping the minimum version of PostgreSQL to 13, even 14. We should endeavor to upgrade our aging infrastructure. |
| 11:47 |
Dyrcona |
I also agree that this should go to the development list, since not everyone is in here at the same time. |
| 11:55 |
* Dyrcona |
jumped on that like a live hand grenade. |
| 11:55 |
Dyrcona |
'cause that's what this conversation is. |
| 11:58 |
Dyrcona |
/me considered starting to look at Pg 17 the other day. It should be released soonish, and it's available for testing. |
| 12:08 |
* sleary |
whispers "we're due to upgrade Angular again, too" and then flees |
| 12:10 |
Dyrcona |
Running to stand still is what it feels like sometimes. |
| 12:12 |
redavis |
lol, sleary, bless you. |
| 12:14 |
Dyrcona |
Yeah, we should do that, too, but I suspect it's too late for 3.14? |
| 12:14 |
redavis |
sandbergja++ #gauche is a great color on you. |
| 12:15 |
redavis |
Dyrcona, I think it is too late for 3.14 unless there's a champion free "right now." |
| 12:27 |
Dyrcona |
Login is broken on my test/training system again. Same message about auth.provisional not being a function. I did nothing with since I last got it working in here. |
| 12:39 |
sleary |
definitely not suggesting we do anything with Angular versions for 3.14, lol, no |
| 12:49 |
|
jvwoolf joined #evergreen |
| 12:56 |
Dyrcona |
Dunno if there's a bug on this already, but when trying to register a patron with the experimental Angular circulation module on Evergreen 3.13, I get this console message: Unable to check <input pattern=\u2018/[0-9|A-Z]{3,12}[-| ]?[0-9|A-Z]{0,4}/\u2019> because \u2018//[0-9|A-Z]{3,12}[-| ]?[0-9|A-Z]{0,4}//v\u2019 is not a valid regexp: invalid character in class in regular expression. |
| 12:56 |
Dyrcona |
The save button is never enabled either. |
| 13:05 |
Dyrcona |
Lp 2081327 |
| 13:05 |
pinesol |
Launchpad bug 2081327 in Evergreen "Angular Circulation cannot save a patron" [Undecided,New] https://launchpad.net/bugs/2081327 |
| 13:14 |
eeevil |
grabbing upgrade version number 1439 ... and then I'll be looking at the anycompatible issue re report security, doh :( |
| 13:15 |
Dyrcona |
snap-- ubuntu-- canonical-- |
| 13:22 |
pinesol |
News from commits: Stamping upgrade script <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=a01af01a4f55093ccb1c54036d7498ad8e885bd4> |
| 13:22 |
pinesol |
News from commits: LP924952: Make acq import resilient against empty subfields <https://git.evergreen-ils.org/?p=Evergreen.git;a=commitdiff;h=450de74660a7e79165e76ef14641a4d78bd12e0b> |
| 13:38 |
redavis |
sandbergja++ |
| 13:42 |
berick |
sleary: neato about Angular @let |
| 13:54 |
redavis |
Completely irrelevant to Evergreen but this must be announced somewhere. I've finally taken the time to set up my laser cutter after moving it from IN to VA. It does need a new exhaust system, and I haven't confirmed cutting yet. But did also install air assist which was much less painful than I'd supposed, but also hasn't been tested yet. |
| 13:56 |
Dyrcona |
redavis++ |
| 13:56 |
redavis |
But prepping a little test now. **delight** |
| 13:58 |
berick |
as long as you cut the word "evergreen" it's relevant. and tax-deducitble. |
| 13:59 |
redavis |
lol, I WILL cut the word "evergreen"! |
| 14:10 |
redavis |
Well, it's definitely burning stuff. We'll see the quality of the burn soon. |
| 14:16 |
jeff |
large (giant?) corporation's knowledgebase articles redirect to their support homepage on expiry of some arbitrary "session", even if you're not even logged in to begin with. back button brings you back where you were, but still... |
| 14:17 |
jeff |
(relevant since we're fixing/improving our own similar redirect behavior) |
| 14:24 |
redavis |
jeff++ |
| 14:57 |
|
jvwoolf joined #evergreen |
| 15:06 |
sleary |
jeff we should file an accessibility bug with them :D |
| 15:48 |
|
Dyrcona joined #evergreen |
| 16:24 |
eeevil |
Dyrcona: I just happened to peek at your bug 2081327 and before digging further I wanted to see if you know whether the "smart" quotes noted in the console message were just an artifact of the browser, or if the template actually might have \u2018 and \u2019 (the actual "smart" quote chars, not the escape sequences, obv) in it... oh, or it might just be the char classes. seems those should both be [0-9A-Z] instead of [0-9|A-Z] with a pipe |
| 16:24 |
pinesol |
Launchpad bug 2081327 in Evergreen "Angular Circulation cannot save a patron" [High,Confirmed] https://launchpad.net/bugs/2081327 |
| 16:27 |
|
sandbergja joined #evergreen |
| 16:28 |
eeevil |
I wonder if that's a YAOUS ... I don't see the string '0-9|A-Z' in the codebase |
| 16:30 |
Dyrcona |
eeevil: IT could be, but I wonder where the unicode comes from? I'll check my db. |
| 16:31 |
eeevil |
well, the error message actually mentions the character classes ... I skimmed right past that! I think that's it |
| 16:32 |
Dyrcona |
eeevil: Looks like it is complaing about our post code regex, but I'm not sure where the smart quotes are coming from. |
| 16:33 |
* Dyrcona |
wonders who put this in and how long ago? |
| 16:34 |
* eeevil |
suspects the unicode is just the render being cute when reconstructing HTML from in-mem DOM, and feeding the "pretty" output to a toJSON thing. |
| 16:34 |
* Dyrcona |
has to go again in a few minutes. |
| 16:35 |
Dyrcona |
Well, I'll fix the regex in the database and update the bug with the results. |
| 16:36 |
* Dyrcona |
has to poof out. |
| 16:36 |
eeevil |
+1 |
| 17:04 |
|
jvwoolf joined #evergreen |
| 17:06 |
Bmagic |
is there a bug in Evergreen that I can't turn up in LP for searching "starts with" requiring the word "the" when searching for "The Hunter". In other words, "hunter" doesn't turn up the title when using "starts with" |
| 17:08 |
|
mmorgan left #evergreen |
| 17:09 |
Bmagic |
seems to only be an issue on the staff client search |
| 17:27 |
jeff |
Evergreen rejects a refund of a credit card payment, even when you're just recording an "externally processed" transaction. That should probably change. I'll write up a bug. |
| 17:29 |
jeff |
Bmagic: not sure I follow. you're searching for "starts with hunter" and you're expecting something that does not start with "hunter" to show up in the results? |
| 17:29 |
jeff |
(not literally searching for the string "starts with hunter", I know) |
| 17:29 |
jeff |
I was trying to express the concept and not the syntax. :-) |
| 17:47 |
Bmagic |
jeff: using the dropdown menu on the staff search called "starts with", then type "hunter" into the search box |
| 17:51 |
|
jihpringle joined #evergreen |
