| *** Daisy has joined #openstack-i18n | 01:33 | |
| *** steveg_afk has joined #openstack-i18n | 01:34 | |
| *** Daisy has quit IRC | 01:38 | |
| *** Daisy has joined #openstack-i18n | 01:39 | |
| *** Daisy has quit IRC | 01:42 | |
| *** Daisy has joined #openstack-i18n | 01:42 | |
| *** Daisy has quit IRC | 02:28 | |
| *** ykatabam has quit IRC | 04:00 | |
| *** ykatabam has joined #openstack-i18n | 04:01 | |
| *** Daisy has joined #openstack-i18n | 04:17 | |
| *** ykatabam1 has joined #openstack-i18n | 04:24 | |
| *** ykatabam has quit IRC | 04:25 | |
| *** Daisy has quit IRC | 04:28 | |
| *** Daisy has joined #openstack-i18n | 04:29 | |
| *** Daisy has quit IRC | 04:33 | |
| *** amotoki has joined #openstack-i18n | 04:38 | |
| *** iranzo has joined #openstack-i18n | 05:52 | |
| *** iranzo has joined #openstack-i18n | 05:52 | |
| *** ykatabam1 has quit IRC | 05:57 | |
| *** ykatabam has joined #openstack-i18n | 07:04 | |
| *** ygbo has joined #openstack-i18n | 07:29 | |
| *** ykatabam has quit IRC | 07:31 | |
| *** macjack has joined #openstack-i18n | 07:54 | |
| *** rsimai_away is now known as rsimai | 07:57 | |
| *** Fabrizio1 has joined #openstack-i18n | 08:56 | |
| *** Fabrizio1 has left #openstack-i18n | 08:56 | |
| *** iranzo has quit IRC | 09:04 | |
| *** adiantum has joined #openstack-i18n | 09:11 | |
| *** iranzo has joined #openstack-i18n | 09:39 | |
| *** Daisy has joined #openstack-i18n | 10:41 | |
| *** Daisy has quit IRC | 10:54 | |
| *** Daisy has joined #openstack-i18n | 10:55 | |
| *** iranzo has quit IRC | 10:57 | |
| *** iranzo has joined #openstack-i18n | 10:57 | |
| *** Daisy has quit IRC | 10:59 | |
| *** amotoki has quit IRC | 11:24 | |
| *** adiantum has quit IRC | 12:09 | |
| *** Daisy has joined #openstack-i18n | 12:31 | |
| *** steveg_afk has joined #openstack-i18n | 12:32 | |
| *** steveg_afk has quit IRC | 12:32 | |
| *** amotoki has joined #openstack-i18n | 12:35 | |
| amotoki | hi, I noticed strings in POT files of openstack-manuals (at least with RST format) are sorted in the alphabetical order of strings | 12:37 |
|---|---|---|
| amotoki | I think it makes translating difficult. | 12:38 |
| amotoki | translators cannot understand the context they are translating.... | 12:38 |
| amotoki | is it better to file a bug in openstack-manuals? | 12:39 |
| *** macjack has quit IRC | 12:41 | |
| clarkb | yes I think the extraction code lives in openstack-manuals | 12:42 |
| amotoki | clarkb: thanks. I just noticed it and join #-doc channel. | 12:43 |
| clarkb | https://git.openstack.org/cgit/openstack/openstack-manuals/tree/tools/generatepot-rst.sh that is the script I think | 12:43 |
| clarkb | uses msgcat --sort-output | 12:44 |
| amotoki | hmm... --sort-by-file looks better. | 12:45 |
| clarkb | ya I think that may be the fix | 12:46 |
| amotoki | I will check how things are changed. | 12:47 |
| *** Daisy has quit IRC | 12:58 | |
| *** amotoki has quit IRC | 13:06 | |
| *** amotoki has joined #openstack-i18n | 13:29 | |
| *** amotoki has quit IRC | 13:33 | |
| *** amotoki has joined #openstack-i18n | 13:39 | |
| *** Daisy has joined #openstack-i18n | 13:48 | |
| *** amotoki has quit IRC | 13:49 | |
| *** amotoki has joined #openstack-i18n | 14:00 | |
| Daisy | doug-fish: ping | 14:11 |
| *** Daisy has quit IRC | 14:17 | |
| *** Daisy has joined #openstack-i18n | 14:18 | |
| *** Daisy has quit IRC | 14:19 | |
| *** Daisy has joined #openstack-i18n | 14:20 | |
| doug-fish | Daisy: hi | 14:22 |
| Daisy | Hi | 14:25 |
| Daisy | I want to understand why newline mismatch will cause a gate check fail in Gerrit. | 14:26 |
| doug-fish | It's because the newline mismatch will cause Horizon | 14:27 |
| doug-fish | 's run_tests.sh --compilemessages script to fail | 14:27 |
| Daisy | what's the root cause ? only Horizon ? | 14:28 |
| doug-fish | So the devstack installation will fail | 14:28 |
| Daisy | do you know ? | 14:28 |
| doug-fish | hmm | 14:28 |
| doug-fish | I'd expect this to be true for every project. But I care the most about Horizon. | 14:28 |
| doug-fish | I can't say for sure on the others | 14:28 |
| Daisy | I see it was the command of msgfmt in clarkb's email | 14:28 |
| doug-fish | maybe we could try it out | 14:28 |
| clarkb | I think cinder hit it too | 14:29 |
| doug-fish | yeah I'm not surprised - I think Babel is doing this checking | 14:29 |
| clarkb | https://git.openstack.org/cgit/openstack/cinder/tree/tox.ini#n124 that line | 14:29 |
| doug-fish | it's used directly by most projects and indirectly by Horizon | 14:29 |
| clarkb | or maybe its Babel | 14:29 |
| Daisy | I read the document of msgfmt, I cannot find this error description in its document. | 14:30 |
| doug-fish | I don't know msgfmt | 14:30 |
| doug-fish | is it doing a lint check? or actually creating compiled mos? | 14:30 |
| doug-fish | nevermind | 14:31 |
| doug-fish | I got it | 14:31 |
| doug-fish | that's doing compiling | 14:31 |
| Daisy | From a translator's perspective, it is not reasonable. I might misunderstand something. | 14:31 |
| Daisy | msgfmt --check-format | 14:31 |
| doug-fish | so <wild speculation> maybe Babel uses msgfmt, and it's the ultimate source of this requirment </wild speculation> | 14:31 |
| clarkb | Daisy: basically the newlines must match, I am confused why that is a problem for translators | 14:32 |
| clarkb | its the same as the python string interpolation formatters must match | 14:32 |
| clarkb | they are just sanity checks enforced by the system because otherwise it does not work as expected | 14:32 |
| amotoki | clarkb: most newlines in horizon translations are meaningless. It comes from newlines in HTML templates. | 14:33 |
| clarkb | amotoki: but why can't you preserve them? | 14:33 |
| amotoki | clarkb: it depends on sitautions. | 14:33 |
| clarkb | regardless of whether or not they are useless, preserving them should be trivial right? | 14:33 |
| doug-fish | I'm having a hard time understanding why they would be hard to preserve as well | 14:34 |
| amotoki | during translations, I inserted 10 newlines after completing a translation just to remove warnigs. | 14:34 |
| Daisy | The length of different languages are different. | 14:34 |
| amotoki | during reviews, I inserted more than 100 newlines just to approve translations. | 14:34 |
| clarkb | amotoki: so it sounds like the real fix would be to clean up horizons strings | 14:34 |
| doug-fish | I think maybe we are talking about 2 different conditions | 14:34 |
| clarkb | (the source strings) | 14:35 |
| doug-fish | amotoki: are you talking about adding newlines to make the line count match for source and target strings? | 14:35 |
| amotoki | doug-fish: exactly | 14:35 |
| doug-fish | amotoki: that's a different issue | 14:35 |
| doug-fish | that will _not_ cause a compilation failure | 14:35 |
| amotoki | doug-fish: sorry. it semes i misunderstood the sitaution. | 14:36 |
| doug-fish | no worries - we've been talking about both at the same time | 14:36 |
| Daisy | What will cause a compilation failure, then ? | 14:36 |
| clarkb | doug-fish: right its just leading and trailing newlines iirc | 14:36 |
| clarkb | which is all that the zanata checker should be checking I thought | 14:37 |
| doug-fish | If a source string has a leading or trailing newline the target string must have exactly the same | 14:37 |
| doug-fish | I think Zanata flags both of these with warnings ... right? | 14:37 |
| -amotoki- will be back and read logs after neutron meeting is over | 14:37 | |
| clarkb | doug-fish: no we switched it to error after several projects ended up broken | 14:37 |
| clarkb | doug-fish: because translations were coming in with the wrong leading and trailing whitespace preventing any translations for that project from merging into git | 14:38 |
| Daisy | Check for consistent leading and trailing newline (\n). | 14:38 |
| Daisy | from the UI of Zanata, it is said it will check: | 14:38 |
| doug-fish | clarkb: yep, understood | 14:38 |
| Daisy | But from the complains from translators, it sounds like it check the exactly same line numbers ? | 14:38 |
| doug-fish | clarkb: and enforcing that leading/trailing newline check is important because it will result in failures (like you said) | 14:39 |
| clarkb | Daisy: it is possible there is a bug in the check | 14:39 |
| Daisy | I will do a test. | 14:39 |
| Daisy | Daisy is doing a test now | 14:39 |
| clarkb | we should test ya and submit a bug if necessary | 14:39 |
| clarkb | cool | 14:39 |
| Daisy | The error reported is : Not enough lines in translation (expected 2, found 1) | 14:45 |
| doug-fish | oh that helps clarify - I don't think that should be an error | 14:46 |
| Daisy | There is another kind of error msg: Unexpected leading newline (\n) | 14:46 |
| doug-fish | That error message is important | 14:47 |
| Daisy | and "Unexpected trailing newline (\n)" | 14:47 |
| doug-fish | it's a real bug | 14:47 |
| doug-fish | yep - that's important too | 14:47 |
| Daisy | I wonder if Zanata need to check the same line numbers in the source and in the translation. | 14:47 |
| doug-fish | maybe this is a bug/enhancement needed for Zanata? (like clarkb said). We need to have teh trailing/leading newline checks, but the same line count check is troublesome and unneeded | 14:49 |
| *** daemontool_ has quit IRC | 14:49 | |
| Daisy | ianychoi and amotoki: if you are there, could you give an example that you have to insert empty line to make sure the line numbers of source and translation are same. | 14:51 |
| Daisy | I want to understand the situation that line number of your translation is different from the line number of the source string. | 14:51 |
| Daisy | In Chinese, we usually use the same format. | 14:52 |
| Daisy | In Chinese, we usually keep the same format. | 14:52 |
| doug-fish | Daisy: I see a couple of examples in the German translation | 14:57 |
| doug-fish | like https://github.com/openstack/horizon/blob/master/openstack_dashboard/locale/de/LC_MESSAGES/django.po#L56 | 14:57 |
| doug-fish | here's a smaller example: https://github.com/openstack/horizon/blob/master/openstack_dashboard/locale/de/LC_MESSAGES/django.po#L101 | 14:58 |
| Daisy | doug-fish: I think this kind of new line is meanless. | 14:59 |
| Daisy | http://git.openstack.org/cgit/openstack/horizon/tree/openstack_dashboard/locale/django.pot#n2676 | 14:59 |
| doug-fish | Daisy: this format isn't really specified directly in Horizon - it's the format chosen by the message extractor | 15:01 |
| doug-fish | https://github.com/openstack/horizon/blob/3a72687ab3ed46e2640a007ca415aeb058abbf52/openstack_dashboard/contrib/sahara/content/data_processing/jobs/templates/data_processing.jobs/_create_job_help.html#L25 | 15:01 |
| doug-fish | that's the source code where the message originates | 15:01 |
| Daisy | doug-fish: German translations have such errors in Zanata. So German translations only keep the leading/tailing newlines. German translations don't keep the same line number. Zanata will report such errors but those errors won't break the gate tests. | 15:01 |
| Daisy | doug-fish: I don't think it's caused by message extractor. I guess there must be a new line character in this paragraph. | 15:03 |
| amotoki | Daisy: hi. one example is resource-id b887813e028d1eb225b90a2fc9a1cbe1. | 15:04 |
| doug-fish | Daisy: oh that's true - but the message is wrapped into 3 lines | 15:04 |
| Daisy | doug-fish: source message is wrapped into 3 lines, but there is only one newline character in this paragraph. I could read the newline character from Zanata editor. | 15:05 |
| *** david-ly_ is now known as david-lyle | 15:06 | |
| Daisy | if the source string is like: http://git.openstack.org/cgit/openstack/horizon/tree/openstack_dashboard/locale/django.pot#n3486 | 15:06 |
| Daisy | I don't think it is nessary to keep the same line number. | 15:07 |
| doug-fish | absolutely agreed | 15:07 |
| Daisy | so I think, it is needed to report a bug to Zanata. | 15:07 |
| Daisy | Thank you, amotoki. So we will report a bug to Zanata. We don't want to make sure the line numbers to be exactly same. We just want to check the heading and tailing newlines. | 15:11 |
| amotoki | Daisy: there are two problems: Too many lines in translation (expected 9, found 10) and Unexpected trailing newline (\n) | 15:11 |
| amotoki | Daisy: the first one is not a big matter. | 15:11 |
| Daisy | Exactly. | 15:12 |
| amotoki | Daisy: the second one is tricky. | 15:12 |
| Daisy | It looks like msgfmt will check the second one. | 15:12 |
| Daisy | the heading and tailing newline (\n) | 15:12 |
| amotoki | we need to insert space at the last line, or we need to insert an extra newlines in the middle of the content. | 15:12 |
| Daisy | amotoki: msgfmt --check-format will check the format to make sure the same number of heading newlines and tailing newlines. If it fails, it will cause the failure of Gate tests. So we want the heading and tailing newline checking in Zanata. | 15:14 |
| Daisy | I think, currently Zanata will check: (1) the same headling newlines (2) the same tailing newlines, (3) the same line number. We don't need (3). | 15:15 |
| amotoki | Daisy: I know it. it is reasonable from gettext tools and PO management. | 15:17 |
| doug-fish | Daisy: amotoki does this mean there are translation updates still needed for Liberty that haven't been proposed to the projects yet? | 15:22 |
| Daisy | doug-fish: what do you mean ? | 15:23 |
| david-lyle | Daisy, horizon RC-2 is open. ttx would like to close it tomorrow, do you know of additional pending translations? | 15:24 |
| doug-fish | we've been chatting about problems with Zanata related to newlines + number of lines ... I wonder if that's blocking any translation updates | 15:24 |
| Daisy | I don't think it would, doug-fish | 15:24 |
| doug-fish | great. Thanks Daisy! | 15:24 |
| Daisy | unless the different heading newlines and tailing newlines break the testing. If there are errors, I could fix them. | 15:25 |
| doug-fish | gate tests are passing | 15:25 |
| doug-fish | so newlines must all be correct at this point | 15:25 |
| amotoki | doug-fish: agree with Daisy. I don't think we need more work in translation side. | 15:26 |
| doug-fish | amotoki: that's great news. It seems our work for Liberty may be almost done! | 15:26 |
| doug-fish | well except for deploying | 15:27 |
| doug-fish | and supporting | 15:27 |
| doug-fish | and describing | 15:27 |
| doug-fish | hmm | 15:27 |
| doug-fish | I'm tired now. | 15:27 |
| -amotoki- is reading logs | 15:30 | |
| *** rsimai is now known as rsimai_away | 15:46 | |
| *** iranzo has quit IRC | 15:49 | |
| *** amotoki has quit IRC | 16:20 | |
| ianychoi | Daisy, are you here? | 16:23 |
| Daisy | ianychoi: yes | 16:23 |
| ianychoi | Sorry, I couldn't follow all of conversations. You said that I would show some examples of different number of lines? | 16:23 |
| Daisy | It's OK now, ianychoi | 16:23 |
| Daisy | I have found enough samples. | 16:24 |
| ianychoi | Aha... thanks :) | 16:24 |
| Daisy | Thank you. | 16:24 |
| ianychoi | I think in Korean, some Korean & English strings have different number of lines and others have the same number of lines... | 16:24 |
| ianychoi | it would be corrected and unified later.. but I am not sure whether me and ujuc would fix them before Liberty RC2 release. | 16:25 |
| Daisy | ianychoi: the errors about different line numbers will not affect the translation importing to repo. | 16:31 |
| Daisy | So don't worry about it. | 16:31 |
| Daisy | I think, we don't regard different line numbers as errors in openstack translations, ianychoi. | 16:32 |
| *** Daisy has quit IRC | 16:36 | |
| *** Daisy has joined #openstack-i18n | 16:36 | |
| *** Daisy has quit IRC | 16:41 | |
| *** ygbo has quit IRC | 16:46 | |
| *** Daisy has joined #openstack-i18n | 17:16 | |
| *** david-ly_ has joined #openstack-i18n | 17:21 | |
| *** david-lyle has quit IRC | 17:21 | |
| *** david-ly_ is now known as david-lyle | 17:22 | |
| *** iranzo has joined #openstack-i18n | 18:45 | |
| *** Daisy has quit IRC | 19:01 | |
| *** Daisy has joined #openstack-i18n | 19:02 | |
| *** Daisy has quit IRC | 19:06 | |
| *** iranzo has quit IRC | 19:12 | |
| *** aeng has joined #openstack-i18n | 20:17 | |
| *** DuncanT has quit IRC | 20:43 | |
| *** krotscheck has quit IRC | 20:43 | |
| *** neillc has quit IRC | 20:43 | |
| *** DuncanT has joined #openstack-i18n | 20:46 | |
| *** krotscheck has joined #openstack-i18n | 20:46 | |
| *** neillc has joined #openstack-i18n | 20:46 | |
| *** pleia2 has quit IRC | 20:47 | |
| *** pleia2 has joined #openstack-i18n | 20:56 | |
| *** ykatabam has joined #openstack-i18n | 21:27 | |
| *** aeng has quit IRC | 21:39 | |
| *** aeng has joined #openstack-i18n | 21:55 | |
| *** Adri2000 has quit IRC | 22:04 | |
| *** Daisy has joined #openstack-i18n | 23:46 | |
| *** ykatabam1 has joined #openstack-i18n | 23:52 | |
| *** Daisy has quit IRC | 23:53 | |
| *** ykatabam has quit IRC | 23:53 | |
| *** Daisy has joined #openstack-i18n | 23:54 | |
| *** Daisy has quit IRC | 23:58 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!