| # | Time | Nick | Message |
|---|
| # | 01:04:46 | edoceo has quit IRC |
| # | 09:15:08 | Dyrcona has joined #evergreen |
| # | 09:19:09 | Dyrcona | yipe! |
| # | 09:19:23 | Dyrcona | lots of conflicts in a pull on my dev branch. |
| # | 09:21:59 | Dyrcona | Looks like an upgrade script was committed with the XXXX.* name. |
| # | 09:27:12 | tsbere blames denials |
| # | 09:27:15 | tsbere | <_< |
| # | 09:28:48 | tsbere grabs 0682 to fix it himself anyway |
| # | 09:33:01 | tsbere | Dyrcona: Pushed a fix for the XXXX script. Probably doesn't help with your conflicts, as I think you had pulled a pile of stuff that went into master after you pulled them ;) |
| # | 09:33:42 | Dyrcona | Yeah. Looks like copy location groups went in and maybe some others. |
| # | 09:33:57 | Dyrcona | I'll just make a new branch tomorrow morning. |
| # | 09:39:12 | dbs_bah has joined #evergreen |
| # | 09:39:31 | dbs_bah | tsbere++ # thanks for the save on my fail to update the upgrade. duh. |
| # | 09:40:40 | dbs_bah is trapped; rebooted into the native Windows side of his laptop for the first time in months and ohhhh the updates. |
| # | 09:40:53 | tsbere | heh |
| # | 09:41:56 | dbs_bah | tsbere: I guess I should hand-update the point-to-point version upgrade too? Or are you taking one more swing at that for alpha3? |
| # | 09:42:42 | tsbere | dbs_bah: I will be taking a swing at upgrading the previous one I had there (that I don't think anyone pushed to master yet either) to cover things that have happened since. I will then push it to master *before* attempting the alpha3 cut ;) |
| # | 09:43:06 | dbs_bah | sweet |
| # | 09:43:40 | tsbere | dbs_bah: You have enough to deal with on OpenSRF cutting anyway, I think. |
| # | 09:44:01 | dbs_bah | ah, that's easy, just "git tag" and roll :) |
| # | 09:44:13 | tsbere | heh |
| # | 09:44:16 | dbs_bah | (not quite, but in comparison to Evergreen it's a walk in the park) |
| # | 09:59:47 | dbs_bah has quit IRC |
| # | 10:01:59 | Callender_ has joined #evergreen |
| # | 10:02:12 | mtate has quit IRC |
| # | 10:02:13 | tater has joined #evergreen |
| # | 10:02:38 | Callender has quit IRC |
| # | 10:02:52 | Callender_ is now known as Callender |
| # | 10:03:05 | tfaile_ has quit IRC |
| # | 10:19:46 | dbs has joined #evergreen |
| # | 10:19:46 | dbs has joined #evergreen |
| # | 10:21:23 | dbs | Dyrcona: sorry for subjecting you to pull conflict failure :( |
| # | 10:22:01 | dbs is about to put a proper upgrade script together for the unapi-subobject branch, then move on to opensrf cutting |
| # | 10:23:33 | dbs | having a 21" widescreen helps with these 227-character lines a bit |
| # | 10:27:52 | tsbere | heh |
| # | 10:28:10 | tsbere | Sometimes I avoid long lines, other times I just go with wrapping |
| # | 10:28:25 | dbs | ah, there's a 456-character wide line |
| # | 10:29:27 | dbs | I figure shorter lines makes it easier to figure out what changed on the line when looking at diffs. |
| # | 10:29:29 | dbs | <-- lazy |
| # | 10:29:38 | dbs continues scriptery |
| # | 10:38:39 | Dyrcona | dbs: I figured I should have waited at least until to day to make my branch. |
| # | 10:39:49 | dbs | Dyrcona: the upgrade script for the unapi_subobject branch is just a copy of 990.schema.unapi.sql |
| # | 10:43:11 | Dyrcona | dbs: I wan't planning on including that branch originally, but guess I will give it a go this time around. |
| # | 10:47:02 | dbs | Dyrcona: user/dbs/unapi_improve_limit now has an unwrapped upgrade script, if you want to give it a go |
| # | 10:47:47 | dbs | bshum likes it, I'm sure you guys will too - although the library ranking might need a small tweak for MVLC's preferences |
| # | 10:49:15 | dbs | (to wit: the library ranking puts copies in order of closeness to the preferred library - perfect match at the top, first depth children next ordered by lib name, second depth children next; and then the same for libs outside of pref lib scope |
| # | 10:50:02 | dbs | and IIRC what MVLC wanted was "if pref_lib or children, put first & order by name; then anything else order by name" |
| # | 10:50:26 | dbs | then again, for MVLC's 1 branch-per-system org tree, there might not be any visible difference :) |
| # | 10:52:28 | dbs | Do you want me to rebase it against current master? |
| # | 10:57:38 | dbs | looks like it merges cleanly |
| # | 11:16:52 | tsbere | dbs: We do have a few 2 branch-per-system entries. Otherwise we wouldn't have bothered with the system level at all ;) |
| # | 11:30:08 | dbs | Hmm. wonder why my attempt to push tags failed with "DENIED by fallthru" |
| # | 11:37:31 | dbs | maybe related to gitolite config when it was tightened up to prevent errant pushes to origin that were meant for working? |
| # | 11:38:27 | dbs will proceed without the tag for now |
| # | 11:47:20 | dbs | http://evergreen-ils.org/downloads/OpenSRF-ChangeLog-2.1.0-alpha1 is up, as are http://evergreen-ils.org/downloads/opensrf-2.1.0-alpha1.tar.gz and http://evergreen-ils.org/downloads/opensrf-2.1.0-alpha1.tar.gz.md5 |
| # | 12:23:25 | dbs | Dear apache: a .tar.gz.md5 file is not an application/x-gzip file |
| # | 12:25:27 | dbs | @later tell moodaepo maybe a dash of https://issues.apache.org/jira/browse/INFRA-2177 would help teach apache not to treat .tar.gz.md5 as the wrong mime type? |
| # | 12:25:27 | pinesol_green | dbs: The operation succeeded. |
| # | 12:27:30 | bshum | dbs: Done. |
| # | 12:37:00 | dbs | bshum++ |
| # | 12:37:49 | bshum | Updating the OpenSRF download page with links |
| # | 12:38:04 | bshum | Just noticed something though, I think we're supposed to put alpha stuff in downloads/previews ? |
| # | 12:38:26 | dbs | damn, probably |
| # | 12:38:34 | bshum | Not a big deal, I guess we can move it around afterwards when we retire it. |
| # | 12:38:47 | dbs | Or move them there now and create symlinks |
| # | 12:38:50 | dbs will do so |
| # | 12:38:55 | bshum | Okay. |
| # | 12:40:39 | dbs | done. I guess I should update the OpenSRF release document accordingly |
| # | 12:42:31 | dbs | http://evergreen-ils.org/dokuwiki/doku.php?id=dev:release_process:opensrf:2.0 updated |
| # | 12:47:56 | bshum | dbs++ |
| # | 13:11:42 | tsbere | dbs: I can mess with the gitolite config |
| # | 13:12:19 | dbs | Thanks, I haven't pushed a tag since September so that seemed like the likely place to poke :) |
| # | 13:12:50 | tsbere | What did you name the tag, out of curiosity? |
| # | 13:13:44 | dbs | osrf_rel_2_1_0-alpha1 |
| # | 13:13:50 | dbs | is it the hyphen? |
| # | 13:14:24 | tsbere | Ahh, I have tags locked down to just rel_ on OpenSRF. Should I add osrf_ to the front of that? |
| # | 13:15:15 | tsbere | meh, just going to add that as an additional allow line |
| # | 13:15:39 | dbs | http://git.evergreen-ils.org/?p=OpenSRF.git;a=tags says osrf_* was allowed at one point :) |
| # | 13:15:59 | tsbere | Yea, back when I had the rule as "you can put anything in the repo" ;) |
| # | 13:16:44 | tsbere | ok, should be clear for osrf_rel_whatever tags on OpenSRF now |
| # | 13:17:38 | dbs | sweet, will try |
| # | 13:18:52 | dbs | remote: error: hook declined to update refs/tags/osrf_rel_2_1_0-alpha1 |
| # | 13:19:11 | tsbere may need to dig out his docs |
| # | 13:20:12 | dbs doesn't want to get in the way of the 2.2 alpha3 release, please worry about this later :) |
| # | 13:21:19 | tsbere | AHA, figured it out. By default it assumes all rules are in refs/heads. I need to tell it that the tags are, well, tags. |
| # | 13:21:28 | dbs | dang tags |
| # | 13:22:05 | tsbere | dbs: Try it again? |
| # | 13:22:24 | dbs | Sweet success |
| # | 13:22:28 | dbs | tsbere++ |
| # | 13:22:48 | tsbere | Commit message for that change: "OpenSRF uses *real* tags, tell the config that" |
| # | 13:24:58 | tsbere | dbs: In addition to alpha3 I get to worry about upgrading MVLC's production servers tonight too. <_< |
| # | 13:25:34 | tsbere | On a different note, should we update the Evergreen side README file to point at the 2.1 alpha instead of 2.0 on the OpenSRF front? |
| # | 13:25:52 | dbs | tsbere: I think so, because 2.0 will fail for alpha3 users right? |
| # | 13:26:04 | tsbere | Yes. Or master users, really ;) |
| # | 13:26:43 | dbs | tsbere: you should commit the unapi branch to master soon, then, so that you can have an awesome alpha3 & awesome production servers |
| # | 13:26:49 | dbs | <wink> |
| # | 13:26:59 | tsbere | Of course, it isn't like the system is 100% unusable or anything. You just can't do anything that requires authentication without a later OpenSRF. ;) |
| # | 13:27:34 | tsbere has been avoiding the unapi branch due to a complete lack of unapi understanding......which is also why he ignores most ACQ branches too. |
| # | 13:28:23 | dbs | unapi I have a bit of a handle on; acq I'm still flailing after all these years |
| # | 13:29:14 | tsbere | I blame my focus on improving things like circulation. I have caught myself quoting files and functions at people that ask about certain things that happen in circ (and holds). Without having to look them up. |
| # | 13:29:52 | dbs | (I'm still amused that we convert little bits of data into full-blown XML, JSON-ify it to send it as an XML message stream, de-JSONify it, and then XPath the results to get back to the data bits |
| # | 13:30:17 | dbs | you have a scary deep understanding of circ. Don't go anywhere :) |
| # | 13:31:21 | tsbere | I have seen data->xml->json->php serialized array->XMPP and back again in the past. <_< And I wrote two or three pieces of the chain at the time. |
| # | 13:37:04 | dbs needs to look at osrf_utf8.c:510:29: warning: ‘utf8_char’ may be used uninitialized in this function [-Wuninitialized] sometime |
| # | 13:44:01 | tsbere is mentally going over ideas for how to tech C and Perl to load errors from the database but also not have to wait for the DB frequently when doing so.......so far he has "cache the first code/language request we get for each code" |
| # | 13:44:45 | tsbere | I find myself wondering if memcache would be a good candidate for some of that |
| # | 13:45:41 | dbs | Seems like a good possibility, with one key per event code? |
| # | 13:46:35 | dbs | ils_event.code.en-US.SUCCESS or something like that? |
| # | 13:51:13 | Dyrcona | dbs: doing unsigned long utf8_char = 0; on line 349 of osrf_utf8.c should make that warning go away harmlessly. |
| # | 13:52:23 | Dyrcona | you want I should submit a branch? |
| # | 13:52:39 | tsbere | dbs: I would probably go ils_Event.code.SUCCESS.en-US, but that is personal preference ;) |
| # | 13:52:51 | dbs | with C I have a long standing habit of initializing the variable separately from declaration time |
| # | 13:53:00 | tsbere likes keeping language at the end for these things, makes things group more nicely |
| # | 13:53:45 | dbs | tsbere: go for it, man |
| # | 13:54:06 | tsbere is going to ask for more opinions on doing that, in general, on Monday. If he has time. |
| # | 13:54:19 | tsbere | Today is more of a "get our production systems updated and build alpha3" day ;) |
| # | 13:54:46 | dbs _thinks_ that hearkens back to initializations in libraries where an initialization at declaration time could have unexpected results the next time the function is called |
| # | 13:55:21 | dbs also hasn't done much C coding in about 10 years |
| # | 13:56:02 | dbs | Dyrcona: maybe make it a bitesize bug for a student to pick up, actually? |
| # | 13:56:49 | dbs | Feels weird to deliberately leave rough edges, oh well. We'll know in a week if we will be GSoC participants or not |
| # | 13:57:12 | tsbere | heh |
| # | 13:57:46 | tsbere | *leaving* rough edges for a bit is one thing. Putting them in just to make people fix them is another. |
| # | 13:58:35 | Dyrcona | Makes me wonder why osrf_utf8.c even exists, instead of using icu, glib, or iconv.... |
| # | 13:59:03 | Dyrcona | but anyway.... |
| # | 13:59:15 | tsbere | NIH syndrome? |
| # | 13:59:15 | Dyrcona | yeah, a bitesize bug would be a good idea. |
| # | 14:00:00 | Dyrcona is guilty of inventing new ways of doing things, but it isn't NIH, it's ignorance that a different implementation already existed. |
| # | 14:00:01 | dbs | Not quite NIH, I think it's specifically about encoding/decoding Unicode in JSON |
| # | 14:01:21 | dbs | http://markmail.org/message/h6szi3sydailotbd <-- pertinent thread |
| # | 14:01:24 | Dyrcona | sometimes, it is NIH, in the sense of, "Gee. I don't like their interface or their implementation. I want something different." |
| # | 14:02:57 | Dyrcona | Ah, yes....standards. *cough, cough* :) |
| # | 14:06:52 | tsbere | http://www.xkcd.com/927/ |
| # | 14:08:48 | dbs is sure there's an alternative xkcd interpretation of that standard xkcd reference |
| # | 14:09:32 | Dyrcona | To paraphrase RFC 4627: use utf8, but don't use actual utf8....use some bastardized normalization that doesn't exist in any published Unicode standard. |
| # | 14:10:45 | Dyrcona | Would have been soooo much simpler to say, use NFC or NFD...... |
| # | 14:10:53 | Dyrcona | Thank you, Douglas Crockford. |
| # | 14:11:44 | Dyrcona stops bitching. |
| # | 14:14:51 | dbs thinks limitations of EcmaScript at the time of JSON's specification might have been the root of the problem |
| # | 14:15:18 | dbs | http://wiki.ecmascript.org/doku.php?id=proposals:update_unicode - yep, looks like it |
| # | 14:16:16 | Dyrcona | Heh... More "standards." :) |
| # | 14:16:26 | Dyrcona | Standards are good, right? |
| # | 14:19:33 | Dyrcona makes a note to avoid JSON for his own projects in the future. |
| # | 14:20:05 | dbs | Yeah. I try to assume that people writing standards are way smarter than I am and that they're trying to solve problems that cascade into horrendous difficulty when you cross boundaries of projects |
| # | 14:31:22 | Dyrcona | Sounds like ECMAScript uses a half-baked implementation of unicode. |
| # | 14:34:45 | tsbere | dbs: I try to assume that of *some* people writing standards. If it involves libraries, though, I change to "they're vendors trying to ensure that they can lock you into using their products with no escape route" |
| # | 14:34:54 | dbs | tsbere++ |
| # | 14:35:20 | dbs | touché |
| # | 14:36:04 | tsbere wonders if that shouldn't just be the overall "default" for any and all standards: "Assume that someone is trying to ensure you have to use their software to make this work" |
| # | 14:37:31 | dbs | having worked with people that were on the ISO SQL standards committee, that's a safe assumption for some problem spaces :) |
| # | 14:57:26 | dbs may propose adding the XML pretty printer to our core set of functions for the convenience & sanity of working with unapi: http://people.planetpostgresql.org/andrew/index.php?/archives/61-XSLT-working-properly-in-Postgres-at-last.html |
| # | 15:17:45 | dbs | heh, small bug in adding the copy table so as to sort by copy library - located URIs don't have copies, thus no URIs. sheesh. |
| # | 15:24:53 | tsbere | Good thing I didn't try to grab that branch yet then? ;) |
| # | 15:25:17 | dbs | No, it's fine. A small regression compared to all the forward movement. |
| # | 15:33:03 | tsbere likely won't hit that today anyway |
| # | 15:34:05 | dbs | no problemo |
| # | 15:40:38 | Dyrcona has quit IRC |
| # | 15:49:24 | dbs | aaaaaand the fix |
| # | 15:50:45 | bjwebb has quit IRC |
| # | 15:57:25 | dbs | pushed onto user/dbs/unapi_improve_limit for the interested |
| # | 16:25:05 | dbs has quit IRC |
| # | 17:28:18 | bjwebb has joined #evergreen |
| # | 17:52:44 | phasefx_ has quit IRC |
| # | 17:57:44 | edoceo has joined #evergreen |
| # | 18:33:42 | Dyrcona has joined #evergreen |
| # | 19:28:51 | Dyrcona has quit IRC |
| # | 19:58:59 | Dyrcona has joined #evergreen |
| # | 20:18:28 | Dyrcona has quit IRC |
| # | 20:24:22 | tsbere has quit IRC |
| # | 20:24:22 | tsbere_ has joined #evergreen |
| # | 20:55:32 | tsbere_ is now known as tsbere |
| # | 21:15:04 | bjwebb has quit IRC |
April 10-13