13:00:23 #startmeeting OpenStack I18n Meeting 13:00:24 Meeting started Thu Nov 19 13:00:23 2015 UTC and is due to finish in 60 minutes. The chair is Daisy. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:00:25 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 13:00:27 The meeting name has been set to 'openstack_i18n_meeting' 13:00:43 Hi, katomo, ianychoi and ujuc 13:00:46 Nice to see you. 13:00:53 Hello, katomo and Daisy :) 13:00:58 Hi 13:01:02 Frank here, hello! 13:01:10 Hello :) 13:01:12 Hi Daisy, ianychoi, ujuc 13:01:22 Hi, eumel8. 13:01:25 Let's start. 13:01:29 hi eumel8 13:01:45 #topic Status update and issues collection of the work items 13:02:25 I'm going to review the progress of the work items we planned in Tokyo summit, and collected issues if there are. 13:03:01 1. Statistic in stackalytics 13:03:46 Ilya owns this. The lastest status from him is: I'm working on data acquisition (with workaround for translators list) as we discussed in Tokyo and waiting for https://zanata.atlassian.net/browse/ZNTA-742 (and watching what's happening =) ) . 13:04:46 I think it's an easy function, I will check the progress with Zanata team after meeting. 13:05:03 2. Enable translation check website (Daisy) 13:05:11 +1 13:05:34 Got it for 1. 13:05:42 I have studied the two repos from Akihiro. I initiated the discussion with infrasture team. 13:06:15 Hope next time I could make some progress. 13:06:32 3. Have training guide in translation website (KATO) 13:06:45 It is done, right, katomo ? 13:06:48 going 13:06:50 none 13:07:02 how is the progress ? 13:07:03 #link https://review.openstack.org/#/c/245615/ 13:07:14 almost there :) 13:07:20 1 +2 13:07:31 OK. Thank you. 13:07:34 :) 13:07:41 4. Translate Installation guide (All) 13:07:59 Installation guide is branched. The stable liberty version is created. 13:08:07 Let me check the progress. :) 13:08:10 1 13:08:41 katomo: Must we translate "common" file ? 13:08:57 yes 13:09:09 all of the common ? 13:09:11 some contents are in common-rst. 13:09:23 yes, all. 13:09:42 I reduce common-rst to the minimum for install guides. 13:09:51 OK. Got it. 13:09:57 Oh.... install-guide + common-rst.. I thought only install-guide is targeted for next January. 13:10:01 Then I will write it down to the home page. 13:10:06 Okay. 13:10:14 Yes. only installation guide. 13:10:34 I see Japanese team have finished, and Korean and Chinese team have solid progress. 13:11:22 Got it. I now understand that 'install-guide for next January' and 'common-rst' is not for next January but important. 13:11:23 to clarify, translate common-rst at stable/liberty branch for Liberty install guides. 13:11:40 good progress. nice work. 13:11:58 ianychoi: I think, common-rst is for next January too. 13:12:11 Okay..... I see! 13:12:14 because part of installation guide contents are in common-rst 13:12:15 yeah, no need so harry :) 13:12:53 Yes, but I thought only install-guide and no common-rst is targeted for next January. 13:13:02 Now I got it. 13:13:07 Next January is a target we set. No target, no pressure. :) 13:13:26 :) 13:13:28 Yep! Thanks, Daisy :) 13:13:40 ianychoi: I will update the home page the clarify this. 13:13:56 Thank you for that clarification :) 13:14:26 5. Translate openstack dashboard 13:14:39 it's not started. We will wait for the feature freeze. :) 13:14:44 6. Set up i18n repo, add terminology files, enable the sync if possible. (Daisy) 13:15:04 Thank katomo to get the repo structure ready. 13:15:19 you are welcome 13:15:20 Thanks to Akihiro and katomo to work on the documents. 13:15:42 almost Akihiro :) 13:15:43 The term and translations are merged. 13:16:22 but I met with an issue while I'm trying to set up the infrastructure. 13:17:00 i18n repo is different from other projects, because i18n.pot is not extracted. It's manually created. 13:17:44 hm 13:17:56 The current logic of the uploading and downloading scripts will use a script to extract strings and generate new pot files. Then use "git status" to check if pot files are updated. 13:18:29 If no pot files get updated, don't upload to Zanata. If pot files are updated, upload to Zanata. 13:19:28 In our case, it's different. Let's imaging I add a new term to i18n.pot, and commit to i18n repo. The uploading script will be invoked. 13:19:47 How can the upload script tell whether a new pot file generates ? 13:20:39 So I'm thinking about two opinions. One is to follow other projects logic to generate i18n.pot, the other is to change the uploading script to handle the special case of i18n repo. 13:20:57 I haven't decided yet. 13:21:09 any comments ? 13:21:27 the latter sound reasonable. 13:21:49 the latter can use a free name :) 13:21:50 katomo: it might be the easier one. 13:22:02 Yes, I don't like i18n.pot. I like term.pot. 13:22:45 Maybe I work on the laster option firstly. 13:22:48 Just my idea... we can forcefully upload everytime 13:23:08 Yes, we can. 13:23:24 It means that rather other other projects, we need to only manually upload pot files without downloading from zanata. 13:23:28 Am I understanding right? 13:23:47 yes, ianychoi . 13:24:01 Let's move to next item. 13:24:09 7. Run scripts to manage release process (Daisy, Alex ) 13:24:10 Yes, then I prefer the 2nd options. Okay 13:24:11 Not started. 13:24:25 8. Set readonly to all projects in Transifex and handle 84 join request. May delete it in next summit. 13:24:39 I handled part of the requests , just before the meeting. 13:25:08 thanks 13:25:20 I have rejected almost a half of the requests. Send messages to notify them we have moved to a new website. 13:25:37 I will complete before next meeting. 13:26:01 ok. we have gone through all the items. Anything I missed ? 13:26:30 if no, let's go to next topic. 13:26:44 no 13:26:49 no 13:26:49 ok. Thank you, katomo 13:26:54 #topic Discussion: how to manage I18n related bug reports ? 13:27:18 hm, amotoki is not here... 13:27:36 I see there are discussions in Gerrit. 13:28:14 #link https://review.openstack.org/#/c/246256/ 13:28:42 amotoki has summarized three kinds of bugs: 13:28:55 * bugs during translations 13:28:56 * bugs in original strings 13:28:56 * bugs in tool chains 13:29:22 I think they are reasonable. Any other bug types you can think out ? 13:29:57 sounds good 13:30:01 i think good 13:30:02 Reasonable, but 'bugs during translations' means that bugs during translation period? 13:30:08 +1 13:30:26 ianychoi: I think it means bad translations. 13:30:33 or missed translations. 13:30:42 Aha.. got it! 13:30:58 Typo translation? 13:31:26 ujuc: maybe. It could be a bad translation too. Translation is not correct, but no typo. 13:31:45 Yes ok 13:31:55 Either users cannot understand, or users don't agree with the translation 13:32:17 okay 13:32:45 I have an idea. We could figure out who will report those types of bugs, and we separate the types from role perspectives. 13:33:01 An end user might not be able to tell whether a bug is a tranlsation bug or a original string bug. 13:33:47 i think so 13:34:09 My thought is to ask end users to report translation bugs to i18n project in Lanchpad. And we, the i18n team, could figure out the root cause, either we direct the bug to development team, or we fix in Zanata. 13:34:27 This was also my question, who is reporting bugs, and this is very useful, I think. 13:34:47 s/direct/redirect 13:35:02 Good idea. 13:35:35 so let's say: end users may report "bugs during translations", let's call it "translation bugs" 13:35:45 +1 13:35:48 +1 13:35:55 +1 13:35:59 if an end user figure a "translation bugs", he/she goes to Lanchpad. 13:36:03 +1 13:36:35 If a translator ( who has joined the team and have the right to update Zanata) finds a bad translation, he/she could go to Zanata directly 13:37:16 We should think over the translator behaviour when he/she finds a translation bug. 13:37:54 As a coordinator, katomo, ianychoi, ujuc, eumel8, do you like a translator of your team goes to Zanata and directly change the translation ? 13:38:13 sure 13:38:20 :) 13:38:21 absolutely 13:38:21 Yes 13:38:37 yes, but is Launchpad usable for non-english speaking end-users? 13:38:47 we can check them from review status :) 13:39:00 katomo: good point. 13:40:12 eumel8: Launchpad should support that. Non-english speaking end-users could report bugs in their language. The question is do we accept non-English described translation bugs ? 13:40:14 eumel8: i think we can accept non-english report 13:40:45 #link https://bugs.launchpad.net/openstack-manuals/+bug/1474452 13:40:45 Launchpad bug 1474452 in openstack-manuals "Document sidebartoc typo: "한국의" > "한국어"" [Undecided,Fix released] - Assigned to Ian Y. Choi (ianychoi) 13:40:45 katomo: I remember I saw a bug repot in Chinese to a translated document before. 13:40:50 A day ago, we got a chinese bug report. 13:41:12 Thank you for the example, katomo and ianychoi. 13:41:13 Daisy: yes 13:41:18 This is an example of reporting Korean translation error bug report. (Although it could not be solved in Zanata) 13:41:48 If we could add button in Horizon, to report translation bugs, it will be wonderful. 13:42:04 okay, this is good, if the end-user can interact with the tool in his language and, of course, can report his bug in native lang 13:42:06 I don't know if Horizon has bug report buttons. 13:42:32 this was also my question in gerrit 13:42:48 In case we can not understand what language is written, it's difficult :) 13:43:06 katomo: what is difficult ? 13:43:39 Daisy: we can not understand who should treat that bug. 13:43:49 @katomo: I think the right language team can grab the bug ;) 13:44:05 Translated launchpad would be much better but hmm.. it is another problem .. :) 13:44:06 eumel8: i hope so 13:44:25 I hope users could type their languages in English. 13:44:41 type their language name in English. Then we will figure out who could treat the bug. 13:45:02 +1 for 'typing language name in English' 13:45:06 yeah 13:45:23 try to typing 13:45:28 :) 13:46:00 and then, who will find bugs in original strings and bugs in tool chain ? 13:46:40 I imagine, a translation tester may find a original string is untranslatable. 13:46:40 translators 13:46:41 developer 13:47:10 Translators... 13:47:14 language researcher? 13:47:29 ianychoi: what is language researcher ? 13:48:02 I just have thought an unordinary case.. for a linguist. 13:48:18 while after the translation check website is ready, we will get a new role: translation tester 13:48:26 Or a guy who want to extract po file text .. 13:48:34 Japanese team sometimes find : the original strings is difficult to translate, can not control word order(untranslatable), or strange:) 13:48:44 ianychoi: yeah, they will find bugs of the tool chain. 13:49:54 let's say: while a translator is translating or testing, he/she may find bugs of original strings. 13:50:42 In german we have a lots of proper names from english - so it's also not translatable. Sometimes I have also to think about the sense. 13:51:22 let's say: while a tool user is using our tools, he/she may find bugs of the tool chains. 13:52:24 Hope amotoki could figure out the contents he needs from our chat history. 13:52:57 anyway, we will continue the discussion in Gerrit. It's not in a hurry. It's a new item and very useful. 13:53:07 yes 13:53:20 yeah 13:53:34 So I will leave the remain time for open discussions. 13:53:38 #topic open discussions 13:53:53 o/ 13:54:02 http://docs.openstack.org/contributor-guide/rst-conv/inline-markups.html 13:54:25 :program:`RabbitMQ` is useful to translate ? 13:54:37 No 13:55:05 I want to change :program:` ` to `` `` to simplify 13:55:08 ujuc, then just ``RabbitMQ`` ? 13:55:19 ujuc, thanks 13:55:43 I'm back. 13:55:46 As a translator, i think so 13:56:01 Daisy: hello :) 13:56:01 Yes, programs is no translation. 13:56:23 katomo: we should create a translator guidance. 13:56:43 guidance? 13:57:05 we need a translator guidance to tell them: which should be translated and which should not; which should be careful and which is called a "bad translation" 13:57:28 Translaion guide 13:57:30 Translation guide 13:57:30 Aha.. +1 because some translators would translate 'program' word, which generates wrong html outputs. 13:57:32 yeah 13:57:50 `` `` is not sufficient? 13:58:03 we could document these down in the "i18n contribution guide", right ? 13:58:13 +1 for the guide 13:58:22 the doc Akihiro and you are working on, I cannot remember the exact name. 13:58:25 I mean some users translate :program:`RabbitMQ` ==> :[aaaa]:`[bbbb]` 13:58:29 Daisy: sure 13:58:36 aaaa and bbbb are local languages 13:58:48 so I agree to just use ``RabbitMQ`` 13:58:54 wow 13:58:54 ianychoi: yes, the new translator will make this mistake. 13:59:10 so,,, `` `` is better :) 13:59:34 `` can not translate. 13:59:42 Yes 13:59:51 Yes :) 13:59:55 thanks. good example. 13:59:55 katomo: we don't have time in the meeting. But we could go to i18n channel to continue the discussion. 14:00:04 oops 14:00:09 okay 14:00:13 Thanks for all the guys! 14:00:16 katomo: I have to close the meeting now. Thank you for attending. 14:00:17 Ok 14:00:19 #endmeeting