← Previous day | Today | Next day → | Search | Index
All times shown according to UTC.
Time | Nick | Message |
---|---|---|
00:13 | tuxayo | +1 , major OSes having still old releases supported doesn't mean programs still have to support that. Things are less strict. |
00:13 | dcook | They're also "supported" in a way that I think isn't accessible to most users anyway. |
00:14 | I think it comes with added fees so I don't know if people actually buy that extra support or just say it's supported because there's a future date published on the internet.. | |
00:15 | Docker is available from RHEL 7+ anyway, so hopefully koha-docker will mature and then it's less of an issue after that | |
00:16 | These days, if you're not running Debian/Ubuntu and want Koha, then I'd say the way to do it would be via Docker if you don't want to switch OSes | |
00:16 | Easier said than done at the moment, but I think Tomas and folk are working on it | |
00:17 | tuxayo | ""supported" in a way that I think isn't accessible to most users anyway. " |
00:17 | Aren't libraries or any public institutions have some chance of being Red Hat customers for stuff like Extended Life-cycle Support (ELS) ? | |
00:17 | Still should be rare though. | |
00:20 | "Docker is available from RHEL 7+ anyway, so hopefully koha-docker will mature and then it's less of an issue after that" | |
00:20 | Indeed, it feel it would be easier to get sysadmins to manage a container of a recent OS than upgrade the OS running on the bare metal. | |
00:20 | Even though it's still an OS to manage, update look for security issue. So I'm not sure if objectively there is much difference. | |
00:22 | "if you're not running Debian/Ubuntu and want Koha, then I'd say the way to do it would be via Docker if you don't want to switch OSes" | |
00:22 | Ah I see, it's objectively simpler to use containers if your bare metal OS it already running other stuff. | |
00:24 | > Easier said than done at the moment, but I think Tomas and folk are working on it | |
00:24 | dcook doesn't it work well to spin a Debian or Ubuntu container and follow the install instructions? | |
00:24 | Unless it's about using docker like a packaging format instead of just light VMs | |
00:27 | dcook | tuxayo: You have to be careful with your volumes if you don't want things to disappear when it comes time to update your container image |
00:28 | Managing containers comes with its own challenges I reckon. It's scarier to do updates since you're updating the whole thing instead of just a few files | |
00:32 | tuxayo | That lessens the potential of containers helping for old OS installs. And convincing to use containers instead of switching the OS. Still a net gain I think. |
00:33 | dcook | Well if their sysadmins are doing the work. If someone experienced with Koha is doing it, I think it'll be fine (once the Koha community builds up its skills too) |
00:33 | But I do wonder... | |
00:33 | It might be wise for Docker images to take over completely, so that everyone is deploying it the same way | |
00:34 | I think the images would still be using Debian or Ubuntu and we'd still end up with custom Deb packages on custom images | |
00:34 | But if that base infrastructure choice were shared across the board then it would be easier for everyone to support and collaborate... | |
00:48 | tuxayo | > It might be wise for Docker images to take over completely |
00:48 | Hopefully there will be a viable fork if Microsoft does something predatory. In the vein of embrace, extend, and extinguish. | |
00:48 | If they are subtile enough it will never be problematic enough to create the incentive of a fork to try to keep monopoly. And do things boiling frog style. | |
00:49 | Hopefully there will never be a "business opportunity" to monetize docker and the sales at Azure will be enough as well as having devs not switching to Linux anymore due to having now most nice stuff on Windows. | |
00:50 | dcook | There are already business opportunities for monetizing Docker and so far so good I think |
00:50 | Most of the core things should be open as well | |
00:51 | Funny enough Red Hat's Podman is a drop-in replacement for Docker | |
00:51 | I think Podman is actually supposed to be better than Docker. There have been times I've thought of trying it out | |
00:59 | tuxayo | > Most of the core things should be open as well |
00:59 | There are non-core things that can become very prevalent and an obstacle if needing to migrate. Hopefully nothing like that get will prevalent in the industry. | |
00:59 | Given there didn't seems to be mass migration from mongoDB and elasticsearch when they effectively became non libre as a whole, that looks like a real risk. | |
00:59 | Good to know about Podman, thanks. | |
00:59 | dcook | That's true. That is a worry that I have too |
01:00 | I think it's one reason why I haven't pushed to use Docker in prod beyond times where it's been the only choice | |
01:02 | I'd forgotten that Kubernetes stopped using Docker a while ago. It can use Docker images which are just OCI images I think.. | |
01:02 | tuxayo | Maybe just using it as a light VM and for the propositioning using Ansible would make the thing pretty safe since it could be swapped with even LXD |
01:03 | Ah great no need to even give up on Kubernetes | |
01:03 | dcook | Yeah lots of different tools for working with containers |
01:04 | tuxayo | (Assuming LXD and podman comply with OCI) |
01:04 | dcook | Can't speak to LXD but Podman does |
01:05 | What we call Docker images are really OCI images, and they're widely supported | |
01:05 | So I think overall it'd be OK | |
01:06 | Looks like LXD doesn't though | |
01:06 | But I think if we need to jump ship then we do that when the time comes. Rolling with the punches... getting with the times... etc | |
01:09 | tuxayo | > and for the propositioning using Ansible would make the thing pretty safe |
01:09 | Oh and that also avoids the issue of prebuild images. I'm not sure most production deployments of apps rebuild their images. It's another version of downloading and running binaries from the internet. Only without touching the host OS. | |
01:10 | > Looks like LXD doesn't though | |
01:10 | Oh :( | |
01:14 | > if we need to jump ship then we do that when the time comes | |
01:14 | It's not the same effort if having many hours of provisioning work coded in docker instead of Ansible. That's a choice for every project. | |
01:14 | And on a wider scale, Podman has to exist and Kubernetes to support it. There needed to be enough people using (and continuing to use) and making these alternatives so everyone has those lifeboats if needed. | |
01:16 | dcook | Yeah we got to keep our eyes and ears open |
01:16 | The work involved in making things work with Docker is generally useful work anyway | |
01:17 | tuxayo | > we got to keep our eyes and ears open |
01:17 | speaking of that, I found this: https://katacontainers.io/ | |
01:18 | dcook | Yeah the name is familiar but I can't remember why |
01:20 | So I have this system that I run in a VM in prod, but locally I run it on Docker | |
01:20 | And I've developed services for it that locally I'll run in separate containers | |
01:20 | But they can easily run on the same VM too | |
01:21 | tuxayo | > The work involved in making things work with Docker is generally useful work anyway |
01:21 | Indeed there is a lot. Though I wonder how hard it's to empty an existing dockerfile and to migrate the stuff to ansible, salt or something else. | |
01:21 | dcook | By architecting it to work in containers, I just made it easy to manage its config and dependencies |
01:21 | tuxayo | nice :D |
01:21 | dcook | Ideally your Dockerfile shouldn't be that big |
01:21 | If it is, it should probably have more things put into scripts | |
01:22 | tuxayo | Isn't declarative syntax better for a fair share of cases compared to scripts? |
01:23 | dcook | Depends on what you're trying/needing to do |
01:23 | https://gitlab.com/koha-commun[…]le?ref_type=heads | |
01:23 | tuxayo | Wait, dockefile are more just like scripts than declarative stuff? |
01:24 | dcook | Ansible and Docker aren't really trying to solve the same problems |
01:25 | tuxayo | Right, it's more like scripts. I don't know why I had an idea closer to a declarative paradigm. Maybe because of the logic of destroying and building from scratch a clean thing that docker and ansible allow to do. |
01:27 | dcook: QA question, should we prefer to have default values in the constructor rather than in the function? When possible, like seemingly here: | |
01:27 | https://bugs.koha-community.or[…]54680&action=diff | |
01:27 | vs part of the changes could be here: | |
01:27 | https://git.koha-community.org[…]Patroncard.pm#L74 | |
01:27 | I don't know enough about which Perl good practices we should try to care about beyond our mandatory guidelines. | |
01:29 | dcook | Yeah always a tough one eh |
01:30 | In theory, defaults in the constructor, but then Perl lets you mess with things, so sometimes it feels safer doing it in the function where no one else could mess with it | |
01:31 | tcohen[m] | Constructor. And check+exceptions for bad ones |
01:38 | tuxayo | > In theory, defaults in the constructor |
01:38 | Good, same a other langs :) | |
01:38 | *as | |
01:39 | tcohen: hola :) | |
01:39 | > Constructor | |
01:39 | So in this case I should do a follow-up a continue the review. | |
01:39 | tcohen[m] | We need to short circuit early if possible. On such deeply nested code as Koha's, tracking down where things come from is challenging. |
01:39 | tuxayo | "for bad ones" what do you mean? |
01:40 | s/a/and/ | |
01:40 | tcohen[m] | I should be sleeping so lacking some context. But yeah. Default where it makes sense, explode where it makes sense, so devs have tools to understand what they need to do |
01:40 | * dcook | thumbs up |
01:40 | dcook | Why are you're you sleeping, tcohen[m]? |
01:41 | aren't* | |
01:41 | "Why are you not sleeping"* | |
01:41 | Jeez maybe I should go back to bed... | |
01:41 | tcohen[m] | Hi both. Interested in reading the containers talk tomorrow morning |
01:41 | * dcook | is always down to talk about containers |
01:42 | tcohen[m] | Good night! |
01:42 | dcook | night! |
01:44 | tuxayo | tcohen good night, thanks for passing by :) |
01:54 | dcook: thanks for the update on the world of containers ^^ And the QA advice. I'll make a follow-up for setting the default in the constructor. The rest of the changes seems to be using intermediate variables for calculations in various steps instead of a long one. + reuse of a value. So good stuff. | |
02:19 | see you another day all :) | |
06:25 | marcelr joined #koha | |
06:25 | marcelr | hi #koha |
06:26 | lmstrand joined #koha | |
06:37 | cait joined #koha | |
06:39 | cait joined #koha | |
06:54 | magnuse | \o/ |
06:54 | marcelr | o/ |
07:03 | dcook | hola hola |
07:03 | and ciao ciao | |
07:08 | * cait | waves |
07:17 | magnuse | ciao dcook |
08:04 | thomas63[m] | hi there... (full message at <https://matrix.org/_matrix/med[…]mdHjhHSHDgOebIeJu>) |
08:09 | Joubu | bug 34683 |
08:09 | huginn | 04Bug https://bugs.koha-community.or[…]_bug.cgi?id=34683 enhancement, P5 - low, ---, jonathan.druart+koha, In Discussion , The DB upgrade process is not tested |
08:10 | Joubu | thomas63[m]: the selectors you want are .navbar and #header_search |
08:10 | (if you are running a recent version of Koha) | |
08:17 | PedroAmorim[m] | morning o/ |
08:24 | aroussos | Good morning PedroAmorim[m] |
08:25 | PedroAmorim[m]: Thank you for the speedy SO on Bug 31964 | |
08:25 | huginn | 04Bug https://bugs.koha-community.or[…]_bug.cgi?id=31964 normal, P5 - low, ---, a.roussos, Signed Off , Missing manpage for koha-z3950-responder |
08:25 | aroussos | PedroAmorim[m]++ |
08:25 | thomas63[m] | Joubu: thanks a lot, works! |
08:28 | marcelr | hi PedroAmorim[m] |
08:28 | * cait | waves some more |
08:33 | marcelr | hi cait |
08:34 | thomas63[m] | Joubu: one more keyword please… for the breadcrumbs menu |
08:40 | paulderscheid[m] | @later tell tcohen I'd like to test your ktd testing branch just by working, should I expect anything catastrophic? :D |
08:40 | huginn | paulderscheid[m]: The operation succeeded. |
08:40 | Joubu | thomas63[m]: in your browser you can open the developper tools and inspect the structure of the page |
08:41 | that will help you find the class/id you want | |
08:41 | The breadcrumbs is contained into the element with id="sub-header" | |
08:53 | cait | Joubu++ |
09:34 | PedroAmorim[m] | Easy bugs to test that need QA:... (full message at <https://matrix.org/_matrix/med[…]OlbrwnsMZUfTGLDoA>) |
09:48 | marcelr | not found.. |
09:49 | but we have 160 to go | |
09:49 | that is 30 lower than this morning btw | |
09:50 | cait | marcelr++ |
09:50 | stuck on wriing some specification here .( | |
09:51 | marcelr | Joubu++ |
09:52 | Joubu | only picking the easy ones to lower the number.. |
09:52 | marcelr | cheating :) |
09:53 | i have only enh in the so queue and they dont seem to move much nowadays | |
09:53 | not talking about getting a so | |
09:54 | see you later #koha | |
10:31 | khall joined #koha | |
10:35 | oleonard | Hi #koha |
10:53 | khall_ joined #koha | |
11:13 | PedroAmorim[m] | \o |
11:14 | tcohen[m] | hola #koha o/ |
11:22 | cait | can someone tell if we have an API route for old checkouts? |
11:22 | as in... am I missing it or is it not there yet? | |
11:25 | Joubu | cait: there is a checked_in flag you can pass |
11:25 | cait | hm to which route? |
11:25 | Joubu | checkouts |
11:25 | tcohen[m] | Exactly |
11:25 | cait | aha! |
11:26 | thx :) | |
11:26 | and then I an use x-query to limit to a patron_id - right? | |
11:26 | Joubu | https://api.koha-community.org[…]stBiblioCheckouts |
11:26 | checked_in | |
11:26 | boolean | |
11:26 | By default, current checkouts are returned, when this is true then checked in checkouts are returned as result. | |
11:26 | cait | i think then I am onyl missing self registration and address change requests |
11:31 | because both are moderated, verification email etc.... need their own thing | |
11:37 | have a great weekend all! | |
11:37 | and Joubu++ | |
11:37 | cait left #koha | |
11:37 | tcohen[m] | an admin endpoint for doing it from a third party service? |
11:38 | oleonard | Joubu++ |
11:38 | Thanks for the QA work! | |
11:47 | Joubu | oleonard: thank you for the nice patches! |
11:47 | tcohen[m]: seen bug 34683? | |
11:47 | huginn | 04Bug https://bugs.koha-community.or[…]_bug.cgi?id=34683 enhancement, P5 - low, ---, jonathan.druart+koha, In Discussion , The DB upgrade process is not tested |
12:35 | oleonard | tcohen: Your issue_396 ktd branch didn't make Elasticsearch magically work on my Intel Mac ;) |
12:47 | tcohen[m] | without it, it works, right? |
12:47 | I have an intel mac and hove no issues | |
12:47 | oleonard | I don't think I've ever gotten ES to work on my Mac |
12:48 | tcohen[m] | buy an M2 |
12:48 | haha | |
12:48 | let me take a look at my intel one | |
12:49 | * oleonard | will wait as long as possible to replace his 27" iMac now that they're not making them anymore :( |
12:49 | tcohen[m] | keep it forever! |
12:49 | Dyrcona joined #koha | |
12:57 | oleonard | Joubu: Regarding Bug 32890, is it only the "delivered today" tab that is incorrect? |
12:57 | huginn | 04Bug https://bugs.koha-community.or[…]_bug.cgi?id=32890 enhancement, P5 - low, ---, oleonard, Signed Off , Add DataTables to curbside pickups |
12:59 | tcohen[m] | oleonard: have you done a ktd --es7 pull recently |
13:00 | Joubu | oleonard: I don't know, it's Friday afternoon :D I remember that this was a nightmare to make it ordered correctly. |
13:00 | caroline is the one who knows! :D | |
13:01 | oleonard | Okay XD |
13:02 | caroline | I know everything!! lol |
13:02 | I can't even remember what I did yesterday... | |
13:04 | oleonard | I'd completely forgotten I'd written a patch for Bug 34135 until Joubu QA'd it |
13:04 | huginn | 04Bug https://bugs.koha-community.or[…]_bug.cgi?id=34135 enhancement, P5 - low, ---, oleonard, Passed QA , Show the icons for selected tab to the left of the search bar in the staff interface |
13:05 | oleonard | In fact during one of the KohaCon sessions someone pointed out a bug that was bothering them and I thought, "I'll bet we can fix that" Turns out I'd already submitted a patch! |
13:05 | tcohen[m] | PQA push fest starting once I take a screenshot |
13:06 | * tcohen[m] | uploaded an image: (1164KiB) < https://matrix.org/_matrix/med[…]RZixrXH/image.png > |
13:06 | uploaded an image: (793KiB) < https://matrix.org/_matrix/med[…]euhKfkk/image.png > | |
13:07 | uploaded an image: (199KiB) < https://matrix.org/_matrix/med[…]hXZzFgZ/image.png > | |
13:18 | caroline | re bug 32890, it seems I had already asked that it be a datatable with a default order in bug 32883... at least I'm consistent ;) |
13:18 | huginn | 04Bug https://bugs.koha-community.or[…]_bug.cgi?id=32890 enhancement, P5 - low, ---, oleonard, Signed Off , Add DataTables to curbside pickups |
13:18 | 04Bug https://bugs.koha-community.or[…]_bug.cgi?id=32883 normal, P5 - low, ---, jonathan.druart+koha, RESOLVED FIXED, Curbside pickups - Order "To be staged" by date and time of scheduled pickup | |
13:25 | khall joined #koha | |
13:33 | Joubu | yes, caroline, that's why you opened 32890 ;) |
14:50 | aude_c[m] | caroline++ for her help with Documentation management tasks |
14:58 | KodoKorkalo[m] | help! bulkmarcimport messes up my character encodings! |
14:59 | it used to work fine on the previous version of Koha but in 22.11 my i get mess with characters like å, ä, ö | |
14:59 | s/my// | |
15:00 | the nice thing is that we're on our production conversion here 😄 | |
15:00 | jalway joined #koha | |
15:01 | KodoKorkalo[m] | it seems that the characters on the first record are correct, but starting from the second they get all messed up |
15:28 | oleonard | KodoKorkalo[m]: I don't use bulkmarcimport but when I get that problem with a file I've imported via the GUI it usually means the encoding I picked didn't match the encoding of the file |
15:28 | KodoKorkalo[m]: MARC8 vs. UTF-8 for instance | |
15:28 | KodoKorkalo[m] | nope, the file is utf-8 |
15:29 | strange, it works on my dev-koha 😄 | |
15:30 | same koha ja same data, different result | |
15:30 | ja=and 😄 | |
15:34 | jalway | KodoKorkalo[m]: Have you run it through the MARC Validator in MARCEdit? |
15:36 | KodoKorkalo[m] | jalway nope, but I have validated the XML with libxml |
15:37 | jalway in addition we have a monsterous script that does all kinds of things to MARC data prior to import (including different checks and validation), but that doesn't seem to be the problem either and it works on my dev machine afterall 😄 | |
15:38 | jalway | Are your dev machine and production machines on the same version? |
15:40 | KodoKorkalo[m] | jalway yep, they seem to be |
15:41 | jalway | Are you using Unimarc? |
15:41 | KodoKorkalo[m] | jalway nope, MARC21 |
15:42 | hmm... i think it might be one of out own patches that breaks it | |
15:43 | the version on my dev-machine is pretty much clean community-koha without the "local" patches | |
15:45 | jalway | That sounds like a place to start. I'm not seeing any bugs relating to bulkmarcimport in koha's bugzilla. Doesn't mean there isn't, but as you say it works in your dev environment. |
15:45 | KodoKorkalo[m] | yep, it must be one of our own patches, i'll talk to my collagues on monday about this |
15:46 | thanks for your help and time jalway 🙂 | |
15:46 | and oleonard too | |
15:48 | jalway | KodoKorkalo[m]: My pleasure! |
16:05 | PedroAmorim[m] | filed bug 34694 I think this is a critical bug if some more people can also take a quick look at |
16:06 | huginn | 04Bug https://bugs.koha-community.or[…]_bug.cgi?id=34694 critical, P5 - low, ---, oleonard, NEW , OPAC bib record blows up with error 500 |
17:57 | khall joined #koha | |
18:18 | bag joined #koha | |
22:25 | Laura joined #koha | |
22:26 | Guest1481 | hola a todos, soy Laura de Argentina y estoy preparando un trabajo sobre Koha |
23:45 | tcohen[m] | Buen fin de semana, Guest1481 |
← Previous day | Today | Next day → | Search | Index