Final plans before translations switch to Zanata
Hi everyone, As I mentioned at the infrastructure meeting yesterday[0], we are almost ready to launch the production Zanata server (translate.openstack.org). I have a week long trip starting this Friday night so I'm sharing my plans here so clarkb, StevenK and others can continue the work while I'm gone. I'm Cc:ing the translations team on this message so they are kept in the loop. First of all, timeline-wise we're aiming to get this in the hands of the translators by the week of September 7th so they have time to do their work before the Liberty release. We have a few improvements to installation that we still need to work out: 1. Switch to archive module for server installation (change here: https://review.openstack.org/#/c/192901/) 2. Switch to archive module for client installation (I'll get that change proposed today) 3. Figure out why Wildfly 9.1 isn't being installed and used (instead of 8.1 that we began with), in spite of changing the download link (it's being downloaded) and upgrading the puppet-wildfly module (Steve: you've doven into the latest changes in the puppet-wildfly module, perhaps you're in a good place to look at this?) This should all be finished in the next couple days and we can then launch translate.openstack.org Beyond raw server launching, there are a few other things that we need to make sure happen: 1. Work with Carlos to get instructions for properly importing the user database to the new server (I've Cc:ed him here, we need this ASAP) 2. Work with Daisy to make sure there are cut-over instructions from Transifex to Zanata, and if there's anything we need to do (we should make a plan, even if the exact date is not clear) 3. Generally make sure everything on our script side of things is going well (StevenK has done an amazing job with these so I'm not worried) 4. Make sure we have imported everything we need. AJaeger will be traveling some too, but had the following to say: "First priority would be that the translators can translate. So, this means pushing source strings to Zanata (at least once initially) and getting them working is most important. Then the regular update after each change (which should be working fine already). Importing in the projects can wait a few days, there's no harm if that one stops for a week or two since we do not lose time this way. For translation the most important project is horizon with it's friends" I think that's it. I'll do my best to check in and reply to email and IRC pings while I'm traveling, but some things may need to wait until I'm back at my hotel in the [US east coast] evening. I'll be back the [US west coast] night of Sunday, September 6th. The 7th is a holiday in the US, but I'll be available if we have unexpected problems to work through. Thanks everyone. [0] Starting 19:19: http://eavesdrop.openstack.org/meetings/infra/2015/infra.2015-08-25-19.00.lo... -- Elizabeth Krumbach Joseph || Lyz || pleia2
Hi Elizabeth, I will get working on some instructions for the migration of the user data. Just to be clear, I'm assuming that what you want migrated is: - List of languages - List of users and open id's - User's memberships to language teams - User's permissions and roles (this includes admins and such) No other data will be migrated. Please let me know if there's something I'm missing. Regards, Carlos. Carlos A. Munoz Software Engineering Supervisor Engineering - Internationalization Red Hat On 08/27/2015 02:34 AM, Elizabeth K. Joseph wrote:
Hi everyone,
As I mentioned at the infrastructure meeting yesterday[0], we are almost ready to launch the production Zanata server (translate.openstack.org). I have a week long trip starting this Friday night so I'm sharing my plans here so clarkb, StevenK and others can continue the work while I'm gone. I'm Cc:ing the translations team on this message so they are kept in the loop.
First of all, timeline-wise we're aiming to get this in the hands of the translators by the week of September 7th so they have time to do their work before the Liberty release.
We have a few improvements to installation that we still need to work out:
1. Switch to archive module for server installation (change here: https://review.openstack.org/#/c/192901/)
2. Switch to archive module for client installation (I'll get that change proposed today)
3. Figure out why Wildfly 9.1 isn't being installed and used (instead of 8.1 that we began with), in spite of changing the download link (it's being downloaded) and upgrading the puppet-wildfly module (Steve: you've doven into the latest changes in the puppet-wildfly module, perhaps you're in a good place to look at this?)
This should all be finished in the next couple days and we can then launch translate.openstack.org
Beyond raw server launching, there are a few other things that we need to make sure happen:
1. Work with Carlos to get instructions for properly importing the user database to the new server (I've Cc:ed him here, we need this ASAP)
2. Work with Daisy to make sure there are cut-over instructions from Transifex to Zanata, and if there's anything we need to do (we should make a plan, even if the exact date is not clear)
3. Generally make sure everything on our script side of things is going well (StevenK has done an amazing job with these so I'm not worried)
4. Make sure we have imported everything we need. AJaeger will be traveling some too, but had the following to say:
"First priority would be that the translators can translate. So, this means pushing source strings to Zanata (at least once initially) and getting them working is most important. Then the regular update after each change (which should be working fine already). Importing in the projects can wait a few days, there's no harm if that one stops for a week or two since we do not lose time this way.
For translation the most important project is horizon with it's friends"
I think that's it. I'll do my best to check in and reply to email and IRC pings while I'm traveling, but some things may need to wait until I'm back at my hotel in the [US east coast] evening. I'll be back the [US west coast] night of Sunday, September 6th. The 7th is a holiday in the US, but I'll be available if we have unexpected problems to work through.
Thanks everyone.
[0] Starting 19:19: http://eavesdrop.openstack.org/meetings/infra/2015/infra.2015-08-25-19.00.lo...
On Wed, Aug 26, 2015 at 3:51 PM, Carlos Munoz <camunoz@redhat.com> wrote:
Hi Elizabeth, I will get working on some instructions for the migration of the user data.
Just to be clear, I'm assuming that what you want migrated is: - List of languages
I understand that this one is tied to everything else, but we may need to do some editing of this list for languages that were added by admins and couldn't be deleted from the interface (like duplicates that resulted from things being case-sensitive, dash vs underscore, etc). These should be easy to delete from our import because they shouldn't be associated with any users or projects. Daisy and StevenK should have a list of what languages we do want imported so we can clean this up.
- List of users and open id's - User's memberships to language teams - User's permissions and roles (this includes admins and such)
Right, along with who are the specific language coordinators, we don't have one for every language at this stage but I think Daisy set some of them up.
No other data will be migrated. Please let me know if there's something I'm missing.
Yes, I think that's correct. Thank you! -- Elizabeth Krumbach Joseph || Lyz || pleia2
Ok. Got it. We can trim the list of languages from our import script. Carlos A. Munoz Software Engineering Supervisor Engineering - Internationalization Red Hat On 08/27/2015 08:59 AM, Elizabeth K. Joseph wrote:
On Wed, Aug 26, 2015 at 3:51 PM, Carlos Munoz <camunoz@redhat.com> wrote:
Hi Elizabeth, I will get working on some instructions for the migration of the user data.
Just to be clear, I'm assuming that what you want migrated is: - List of languages
I understand that this one is tied to everything else, but we may need to do some editing of this list for languages that were added by admins and couldn't be deleted from the interface (like duplicates that resulted from things being case-sensitive, dash vs underscore, etc). These should be easy to delete from our import because they shouldn't be associated with any users or projects. Daisy and StevenK should have a list of what languages we do want imported so we can clean this up.
- List of users and open id's - User's memberships to language teams - User's permissions and roles (this includes admins and such)
Right, along with who are the specific language coordinators, we don't have one for every language at this stage but I think Daisy set some of them up.
No other data will be migrated. Please let me know if there's something I'm missing.
Yes, I think that's correct.
Thank you!
On 27/08/15 02:34, Elizabeth K. Joseph wrote:
3. Figure out why Wildfly 9.1 isn't being installed and used (instead of 8.1 that we began with), in spite of changing the download link (it's being downloaded) and upgrading the puppet-wildfly module (Steve: you've doven into the latest changes in the puppet-wildfly module, perhaps you're in a good place to look at this?)
With a fresh install, I'd be very surprised if 8.1 was still being installed. For a puppet run with 8.1 already deployed, the archive module will not download and install from the URL if the file specified in creates => exists, so you have to remove /opt/wildfly/jboss-modules.jar and re-run puppet, which sadly breaks the site while you're trying to upgrade it. :-(
3. Generally make sure everything on our script side of things is going well (StevenK has done an amazing job with these so I'm not worried)
The only thing I'm concerned about here is there is currently no support for pulling translations for Horizon or the OpenStack manuals projects, due to having multiple directories with overlapping locales. For example, horizon master has: openstack_dashboard/locale/{as,bn_IN,brx,cs,de,en_AU,...} horizon/locale/{ca,cs,de,en_AU,...} If we perform what we do for the simple case of looking for locales and pulling from the server, we end up pulling down Catalan translations for openstack_dashboard, which isn't intended. I have previously had discussions with Carlos about this issue, and per-document download would help -- or an option to zanata-cli to only freshen local documents when pulling.
I think that's it. I'll do my best to check in and reply to email and IRC pings while I'm traveling, but some things may need to wait until I'm back at my hotel in the [US east coast] evening. I'll be back the [US west coast] night of Sunday, September 6th. The 7th is a holiday in the US, but I'll be available if we have unexpected problems to work through.
I also have some traveling coming up -- I'll be in New Zealand from the 4th of September until the 7th -- most of that is a weekend so we should be fine, but most of the Friday and Monday will be lost dealing with airports and airplanes. Thanks, -- Steve "I'm a doctor, not a doorstop!" - EMH, USS Enterprise
On 2015-08-27 04:58, Steve Kowalik wrote:
[...]
The only thing I'm concerned about here is there is currently no support for pulling translations for Horizon or the OpenStack manuals projects, due to having multiple directories with overlapping locales.
For example, horizon master has: openstack_dashboard/locale/{as,bn_IN,brx,cs,de,en_AU,...} horizon/locale/{ca,cs,de,en_AU,...}
If we perform what we do for the simple case of looking for locales and pulling from the server, we end up pulling down Catalan translations for openstack_dashboard, which isn't intended.
I have previously had discussions with Carlos about this issue, and per-document download would help -- or an option to zanata-cli to only freshen local documents when pulling.
What about postprocessing on our side? I mean something like: 1. Download all languages 2. Go through all newly added files and filter out those that we do not want We do some kind of postprocessing already where we only push if there are real string updates - this could be enhanced. Andreas -- Andreas Jaeger aj@{suse.com,opensuse.org} Twitter/Identica: jaegerandi SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany GF: Felix Imendörffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nürnberg) GPG fingerprint = 93A3 365E CE47 B889 DF7F FED1 389A 563C C272 A126
Could we also revise languages and language codes? I think we should prefer Country-independent languages over country ones. So, we should prefer German ("de") over German in Germany ("de_DE") and German in Austria ("de_AT"). In case where we have country specific ones - like pl_PL -, can we rename them to country independent ones - like pl? When asking for new languages, we should then get the independent one in first and only after consolidation with the existing team create a specific one. What do you think? Andreas -- Andreas Jaeger aj@{suse.com,opensuse.org} Twitter/Identica: jaegerandi SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany GF: Felix Imendörffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nürnberg) GPG fingerprint = 93A3 365E CE47 B889 DF7F FED1 389A 563C C272 A126
On Thu, Aug 27, 2015 at 3:17 AM, Andreas Jaeger <aj@suse.com> wrote:
Could we also revise languages and language codes?
I think we should prefer Country-independent languages over country ones.
So, we should prefer German ("de") over German in Germany ("de_DE") and German in Austria ("de_AT").
In case where we have country specific ones - like pl_PL -, can we rename them to country independent ones - like pl?
When asking for new languages, we should then get the independent one in first and only after consolidation with the existing team create a specific one.
What do you think?
I can't directly speak to the conversation that led to this list, but StevenK and Daisy came up with the following list, which seems to go by what you suggest: ar as bg-BG bn-IN brx ca cs de el en-AU en-GB es es-MX fi-FI fil fr gu he hi hu id it ja ka-GE kn ko-KR kok ks mai mni mr ne nl-NL pa-IN pl-PL pt pt-BR ru sl-SI sr ta te-IN tr-TR ur vi-VN zh-CN zh-TW Copied from this pastebin that StevenK shared last night, in case it expires: http://paste.openstack.org/show/429137/ -- Elizabeth Krumbach Joseph || Lyz || pleia2
looks good but there is not US (en_US) in that list. Remo
On Aug 27, 2015, at 09:10, Elizabeth K. Joseph <lyz@princessleia.com> wrote:
On Thu, Aug 27, 2015 at 3:17 AM, Andreas Jaeger <aj@suse.com> wrote:
Could we also revise languages and language codes?
I think we should prefer Country-independent languages over country ones.
So, we should prefer German ("de") over German in Germany ("de_DE") and German in Austria ("de_AT").
In case where we have country specific ones - like pl_PL -, can we rename them to country independent ones - like pl?
When asking for new languages, we should then get the independent one in first and only after consolidation with the existing team create a specific one.
What do you think?
I can't directly speak to the conversation that led to this list, but StevenK and Daisy came up with the following list, which seems to go by what you suggest:
ar as bg-BG bn-IN brx ca cs de el en-AU en-GB es es-MX fi-FI fil fr gu he hi hu id it ja ka-GE kn ko-KR kok ks mai mni mr ne nl-NL pa-IN pl-PL pt pt-BR ru sl-SI sr ta te-IN tr-TR ur vi-VN zh-CN zh-TW
Copied from this pastebin that StevenK shared last night, in case it expires: http://paste.openstack.org/show/429137/
-- Elizabeth Krumbach Joseph || Lyz || pleia2
_______________________________________________ OpenStack-Infra mailing list OpenStack-Infra@lists.openstack.org http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-infra
!DSPAM:1,55df36d4185551148732578!
On 08/27/2015 06:13 PM, Remo Mattei wrote:
looks good but there is not US (en_US) in that list.
US English is the source language, we do not need it translated We have en in the list as source and that is by convention en_US, Andreas -- Andreas Jaeger aj@{suse.com,opensuse.org} Twitter/Identica: jaegerandi SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany GF: Felix Imendörffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nürnberg) GPG fingerprint = 93A3 365E CE47 B889 DF7F FED1 389A 563C C272 A126
On 2015-08-27 09:13:54 -0700 (-0700), Remo Mattei wrote:
looks good but there is not US (en_US) in that list.
Since it's the source language, it's not configured as a translation target. -- Jeremy Stanley
On 08/27/2015 06:10 PM, Elizabeth K. Joseph wrote:
On Thu, Aug 27, 2015 at 3:17 AM, Andreas Jaeger <aj@suse.com> wrote:
Could we also revise languages and language codes?
I think we should prefer Country-independent languages over country ones.
So, we should prefer German ("de") over German in Germany ("de_DE") and German in Austria ("de_AT").
In case where we have country specific ones - like pl_PL -, can we rename them to country independent ones - like pl?
When asking for new languages, we should then get the independent one in first and only after consolidation with the existing team create a specific one.
What do you think?
I can't directly speak to the conversation that led to this list, but StevenK and Daisy came up with the following list, which seems to go by what you suggest:
ar as bg-BG
Rename to bg?
bn-IN
Rename to bn?
brx ca cs de el en-AU en-GB es es-MX fi-FI
Rename to fi?
fil fr gu he hi hu id it ja ka-GE
Rename to ka?
kn ko-KR
Can we rename to ko, please?
kok ks mai mni mr ne nl-NL
Rename to nl?
pa-IN
Rename to pa?
pl-PL
Can we rename the above to pl, please?
pt pt-BR ru sl-SI
Rename to sl?
sr ta te-IN
Rename to te?
tr-TR
Rename to tr?
ur vi-VN
Rename to vi?
zh-CN zh-TW
Copied from this pastebin that StevenK shared last night, in case it expires: http://paste.openstack.org/show/429137/
Background: We get *lots* of request on transifex for new languages, for example today for pl. There's right now no reason to have both pl and pl-PL. I'd like us to standardize here... Andreas -- Andreas Jaeger aj@{suse.com,opensuse.org} Twitter/Identica: jaegerandi SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany GF: Felix Imendörffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nürnberg) GPG fingerprint = 93A3 365E CE47 B889 DF7F FED1 389A 563C C272 A126
I copied the language list exactly from Transifex, including language teams which have made any progress in translations. I don't suggest to make renaming. Because the names of the language teams will be the folder names in repo after uploading translations to Git repo. If we make renaming here, there will be a new folder added, other than overwriting the existing one. There will be two folders for these renaming languages, for example, folder "bg" and folder "bg-BG". Best regards Ying Chun Guo (Daisy) Andreas Jaeger <aj@suse.com> wrote on 08/28/2015 01:48:25 AM:
From: Andreas Jaeger <aj@suse.com> To: "Elizabeth K. Joseph" <lyz@princessleia.com> Cc: "openstack-i18n@lists.openstack.org" <openstack- i18n@lists.openstack.org>, OpenStack Infra <openstack- infra@lists.openstack.org> Date: 08/28/2015 01:53 AM Subject: Re: [OpenStack-Infra] [Openstack-i18n] Final plans before translations switch to Zanata
On 08/27/2015 06:10 PM, Elizabeth K. Joseph wrote:
On Thu, Aug 27, 2015 at 3:17 AM, Andreas Jaeger <aj@suse.com> wrote:
Could we also revise languages and language codes?
I think we should prefer Country-independent languages over country ones.
So, we should prefer German ("de") over German in Germany ("de_DE") and German in Austria ("de_AT").
In case where we have country specific ones - like pl_PL -, can we rename them to country independent ones - like pl?
When asking for new languages, we should then get the independent one in first and only after consolidation with the existing team create a specific one.
What do you think?
I can't directly speak to the conversation that led to this list, but StevenK and Daisy came up with the following list, which seems to go by what you suggest:
ar as bg-BG
Rename to bg?
bn-IN
Rename to bn?
brx ca cs de el en-AU en-GB es es-MX fi-FI
Rename to fi?
fil fr gu he hi hu id it ja ka-GE
Rename to ka?
kn ko-KR
Can we rename to ko, please?
kok ks mai mni mr ne nl-NL
Rename to nl?
pa-IN
Rename to pa?
pl-PL
Can we rename the above to pl, please?
pt pt-BR ru sl-SI
Rename to sl?
sr ta te-IN
Rename to te?
tr-TR
Rename to tr?
ur vi-VN
Rename to vi?
zh-CN zh-TW
Copied from this pastebin that StevenK shared last night, in case it expires: http://paste.openstack.org/show/429137/
Background: We get *lots* of request on transifex for new languages, for
example today for pl. There's right now no reason to have both pl and pl-PL. I'd like us to standardize here...
Andreas -- Andreas Jaeger aj@{suse.com,opensuse.org} Twitter/Identica: jaegerandi SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany GF: Felix Imendörffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nürnberg) GPG fingerprint = 93A3 365E CE47 B889 DF7F FED1 389A 563C C272 A126
_______________________________________________ OpenStack-Infra mailing list OpenStack-Infra@lists.openstack.org http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-infra
On 08/28/2015 09:28 AM, Ying Chun Guo wrote:
I copied the language list exactly from Transifex, including language teams which have made any progress in translations. I don't suggest to make renaming. Because the names of the language teams will be the folder names in repo after uploading translations to Git repo. If we make renaming here, there will be a new folder added, other than overwriting the existing one. There will be two folders for these renaming languages, for example, folder "bg" and folder "bg-BG".
You're right with the renaming problems in the projects. What about at least defining guidelines for the future? Andreas
Best regards Ying Chun Guo (Daisy)
Andreas Jaeger <aj@suse.com> wrote on 08/28/2015 01:48:25 AM:
From: Andreas Jaeger <aj@suse.com> To: "Elizabeth K. Joseph" <lyz@princessleia.com> Cc: "openstack-i18n@lists.openstack.org" <openstack- i18n@lists.openstack.org>, OpenStack Infra <openstack- infra@lists.openstack.org> Date: 08/28/2015 01:53 AM Subject: Re: [OpenStack-Infra] [Openstack-i18n] Final plans before translations switch to Zanata
On 08/27/2015 06:10 PM, Elizabeth K. Joseph wrote:
On Thu, Aug 27, 2015 at 3:17 AM, Andreas Jaeger <aj@suse.com> wrote:
Could we also revise languages and language codes?
I think we should prefer Country-independent languages over country ones.
So, we should prefer German ("de") over German in Germany ("de_DE") and German in Austria ("de_AT").
In case where we have country specific ones - like pl_PL -, can we rename them to country independent ones - like pl?
When asking for new languages, we should then get the independent one in first and only after consolidation with the existing team create a specific one.
What do you think?
I can't directly speak to the conversation that led to this list, but StevenK and Daisy came up with the following list, which seems to go by what you suggest:
ar as bg-BG
Rename to bg?
bn-IN
Rename to bn?
brx ca cs de el en-AU en-GB es es-MX fi-FI
Rename to fi?
fil fr gu he hi hu id it ja ka-GE
Rename to ka?
kn ko-KR
Can we rename to ko, please?
kok ks mai mni mr ne nl-NL
Rename to nl?
pa-IN
Rename to pa?
pl-PL
Can we rename the above to pl, please?
pt pt-BR ru sl-SI
Rename to sl?
sr ta te-IN
Rename to te?
tr-TR
Rename to tr?
ur vi-VN
Rename to vi?
zh-CN zh-TW
Copied from this pastebin that StevenK shared last night, in case it expires: http://paste.openstack.org/show/429137/
Background: We get *lots* of request on transifex for new languages, for example today for pl. There's right now no reason to have both pl and pl-PL. I'd like us to standardize here...
Andreas -- Andreas Jaeger aj@{suse.com,opensuse.org} Twitter/Identica: jaegerandi SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany GF: Felix Imendörffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nürnberg) GPG fingerprint = 93A3 365E CE47 B889 DF7F FED1 389A 563C C272 A126
_______________________________________________ OpenStack-Infra mailing list OpenStack-Infra@lists.openstack.org http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-infra
-- Andreas Jaeger aj@{suse.com,opensuse.org} Twitter/Identica: jaegerandi SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany GF: Felix Imendörffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nürnberg) GPG fingerprint = 93A3 365E CE47 B889 DF7F FED1 389A 563C C272 A126
It is better to me to keep the current language names during the migration to Zanata. Regarding renaming itself, I think it is good to use shorter names as possible as we can. Using shorter names will be able to cover more areas for a language. On the other hand, they may be various reasons to use longer names which only such language users know). We can delegate decisions to language coordinators (who I think understand the situation most). Akihiro 2015-08-29 2:55 GMT+09:00 Andreas Jaeger <aj@suse.com>:
On 08/28/2015 09:28 AM, Ying Chun Guo wrote:
I copied the language list exactly from Transifex, including language teams which have made any progress in translations. I don't suggest to make renaming. Because the names of the language teams will be the folder names in repo after uploading translations to Git repo. If we make renaming here, there will be a new folder added, other than overwriting the existing one. There will be two folders for these renaming languages, for example, folder "bg" and folder "bg-BG".
You're right with the renaming problems in the projects. What about at least defining guidelines for the future?
Andreas
Best regards
Ying Chun Guo (Daisy)
Andreas Jaeger <aj@suse.com> wrote on 08/28/2015 01:48:25 AM:
From: Andreas Jaeger <aj@suse.com> To: "Elizabeth K. Joseph" <lyz@princessleia.com> Cc: "openstack-i18n@lists.openstack.org" <openstack- i18n@lists.openstack.org>, OpenStack Infra <openstack- infra@lists.openstack.org> Date: 08/28/2015 01:53 AM Subject: Re: [OpenStack-Infra] [Openstack-i18n] Final plans before translations switch to Zanata
On 08/27/2015 06:10 PM, Elizabeth K. Joseph wrote:
On Thu, Aug 27, 2015 at 3:17 AM, Andreas Jaeger <aj@suse.com> wrote:
Could we also revise languages and language codes?
I think we should prefer Country-independent languages over country ones.
So, we should prefer German ("de") over German in Germany ("de_DE") and German in Austria ("de_AT").
In case where we have country specific ones - like pl_PL -, can we rename them to country independent ones - like pl?
When asking for new languages, we should then get the independent one in first and only after consolidation with the existing team create a specific one.
What do you think?
I can't directly speak to the conversation that led to this list, but StevenK and Daisy came up with the following list, which seems to go by what you suggest:
ar as bg-BG
Rename to bg?
bn-IN
Rename to bn?
brx ca cs de el en-AU en-GB es es-MX fi-FI
Rename to fi?
fil fr gu he hi hu id it ja ka-GE
Rename to ka?
kn ko-KR
Can we rename to ko, please?
kok ks mai mni mr ne nl-NL
Rename to nl?
pa-IN
Rename to pa?
pl-PL
Can we rename the above to pl, please?
pt pt-BR ru sl-SI
Rename to sl?
sr ta te-IN
Rename to te?
tr-TR
Rename to tr?
ur vi-VN
Rename to vi?
zh-CN zh-TW
Copied from this pastebin that StevenK shared last night, in case it expires: http://paste.openstack.org/show/429137/
Background: We get *lots* of request on transifex for new languages, for example today for pl. There's right now no reason to have both pl and pl-PL. I'd like us to standardize here...
Andreas -- Andreas Jaeger aj@{suse.com,opensuse.org} Twitter/Identica: jaegerandi SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany GF: Felix Imendörffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nürnberg) GPG fingerprint = 93A3 365E CE47 B889 DF7F FED1 389A 563C C272 A126
_______________________________________________ OpenStack-Infra mailing list OpenStack-Infra@lists.openstack.org http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-infra
-- Andreas Jaeger aj@{suse.com,opensuse.org} Twitter/Identica: jaegerandi SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany GF: Felix Imendörffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nürnberg) GPG fingerprint = 93A3 365E CE47 B889 DF7F FED1 389A 563C C272 A126
_______________________________________________ Openstack-i18n mailing list Openstack-i18n@lists.openstack.org http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-i18n
It's good to have this guideline for the future: If the language code is same with country code, just use the language code. Best regards Ying Chun Guo (Daisy) Andreas Jaeger <aj@suse.com> wrote on 08/29/2015 01:55:03 AM:
From: Andreas Jaeger <aj@suse.com> To: Ying Chun Guo/China/IBM@IBMCN Cc: "Elizabeth K. Joseph" <lyz@princessleia.com>, "openstack- i18n@lists.openstack.org" <openstack-i18n@lists.openstack.org>, OpenStack Infra <openstack-infra@lists.openstack.org> Date: 08/29/2015 01:55 AM Subject: Re: [OpenStack-Infra] [Openstack-i18n] Final plans before translations switch to Zanata
On 08/28/2015 09:28 AM, Ying Chun Guo wrote:
I copied the language list exactly from Transifex, including language teams which have made any progress in translations. I don't suggest to make renaming. Because the names of the language teams will be the folder names in repo after uploading translations to Git repo. If we make renaming here, there will be a new folder added, other than overwriting the existing one. There will be two folders for these renaming languages, for example, folder "bg" and folder "bg-BG".
You're right with the renaming problems in the projects. What about at least defining guidelines for the future?
Andreas
Best regards Ying Chun Guo (Daisy)
Andreas Jaeger <aj@suse.com> wrote on 08/28/2015 01:48:25 AM:
From: Andreas Jaeger <aj@suse.com> To: "Elizabeth K. Joseph" <lyz@princessleia.com> Cc: "openstack-i18n@lists.openstack.org" <openstack- i18n@lists.openstack.org>, OpenStack Infra <openstack- infra@lists.openstack.org> Date: 08/28/2015 01:53 AM Subject: Re: [OpenStack-Infra] [Openstack-i18n] Final plans before translations switch to Zanata
On 08/27/2015 06:10 PM, Elizabeth K. Joseph wrote:
On Thu, Aug 27, 2015 at 3:17 AM, Andreas Jaeger <aj@suse.com> wrote:
Could we also revise languages and language codes?
I think we should prefer Country-independent languages over country ones.
So, we should prefer German ("de") over German in Germany ("de_DE") and German in Austria ("de_AT").
In case where we have country specific ones - like pl_PL -, can we rename them to country independent ones - like pl?
When asking for new languages, we should then get the independent one in first and only after consolidation with the existing team create a specific one.
What do you think?
I can't directly speak to the conversation that led to this list, but StevenK and Daisy came up with the following list, which seems to go by what you suggest:
ar as bg-BG
Rename to bg?
bn-IN
Rename to bn?
brx ca cs de el en-AU en-GB es es-MX fi-FI
Rename to fi?
fil fr gu he hi hu id it ja ka-GE
Rename to ka?
kn ko-KR
Can we rename to ko, please?
kok ks mai mni mr ne nl-NL
Rename to nl?
pa-IN
Rename to pa?
pl-PL
Can we rename the above to pl, please?
pt pt-BR ru sl-SI
Rename to sl?
sr ta te-IN
Rename to te?
tr-TR
Rename to tr?
ur vi-VN
Rename to vi?
zh-CN zh-TW
Copied from this pastebin that StevenK shared last night, in case it expires: http://paste.openstack.org/show/429137/
Background: We get *lots* of request on transifex for new languages, for example today for pl. There's right now no reason to have both pl and pl-PL. I'd like us to standardize here...
Andreas -- Andreas Jaeger aj@{suse.com,opensuse.org} Twitter/Identica: jaegerandi SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany GF: Felix Imendörffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nürnberg) GPG fingerprint = 93A3 365E CE47 B889 DF7F FED1 389A 563CC272 A126
_______________________________________________ OpenStack-Infra mailing list OpenStack-Infra@lists.openstack.org http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-infra
-- Andreas Jaeger aj@{suse.com,opensuse.org} Twitter/Identica: jaegerandi SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany GF: Felix Imendörffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nürnberg) GPG fingerprint = 93A3 365E CE47 B889 DF7F FED1 389A 563C C272 A126
On Fri, Aug 28, 2015 at 10:55 AM, Andreas Jaeger <aj@suse.com> wrote:
On 08/28/2015 09:28 AM, Ying Chun Guo wrote:
I copied the language list exactly from Transifex, including language teams which have made any progress in translations. I don't suggest to make renaming. Because the names of the language teams will be the folder names in repo after uploading translations to Git repo. If we make renaming here, there will be a new folder added, other than overwriting the existing one. There will be two folders for these renaming languages, for example, folder "bg" and folder "bg-BG".
You're right with the renaming problems in the projects. What about at least defining guidelines for the future?
Just to confirm, because of the considerations mentioned above, we're keeping the language list as-is from what Steve and Daisy agreed upon and I posted here: http://lists.openstack.org/pipermail/openstack-i18n/2015-August/001293.html Which I believe is the list that Daisy exported from Transifex. We'll need this when we start doing the migration into production at 00:00 on September 9th to make sure we have all the correct languages, and only the correct languages, as the UI does not currently allow for deletions. Thanks. -- Elizabeth Krumbach Joseph || Lyz || pleia2
Please find my comments for bn and pa inline. Thanks Jimidar (Amandeep Singh) On Thu, Aug 27, 2015 at 10:48 AM, Andreas Jaeger <aj@suse.com> wrote:
On 08/27/2015 06:10 PM, Elizabeth K. Joseph wrote:
On Thu, Aug 27, 2015 at 3:17 AM, Andreas Jaeger <aj@suse.com> wrote:
Could we also revise languages and language codes?
I think we should prefer Country-independent languages over country ones.
So, we should prefer German ("de") over German in Germany ("de_DE") and German in Austria ("de_AT").
In case where we have country specific ones - like pl_PL -, can we rename them to country independent ones - like pl?
When asking for new languages, we should then get the independent one in first and only after consolidation with the existing team create a specific one.
What do you think?
I can't directly speak to the conversation that led to this list, but StevenK and Daisy came up with the following list, which seems to go by what you suggest:
ar as bg-BG
Rename to bg?
bn-IN
Rename to bn?
Bangla is used in two different countries. Bengali in India. and Bangla in Bangladesh. Can not generalize the term for it.
brx
ca cs de el en-AU en-GB es es-MX fi-FI
Rename to fi?
fil
fr gu he hi hu id it ja ka-GE
Rename to ka?
kn
ko-KR
Can we rename to ko, please?
kok
ks mai mni mr ne nl-NL
Rename to nl?
pa-IN
Rename to pa?
Punjabi (pa) has two scripts, one is Gurmukhi used in Indian side of Punjab and the other is Shahmukhi (Urdu) used in Pakistan's Punjab, so renaming to a generalized name wont do the justice to both entirely different scripts.
pl-PL
Can we rename the above to pl, please?
pt
pt-BR ru sl-SI
Rename to sl?
sr
ta te-IN
Rename to te?
tr-TR
Rename to tr?
ur
vi-VN
Rename to vi?
zh-CN
zh-TW
Copied from this pastebin that StevenK shared last night, in case it expires: http://paste.openstack.org/show/429137/
Background: We get *lots* of request on transifex for new languages, for example today for pl. There's right now no reason to have both pl and pl-PL. I'd like us to standardize here...
Andreas -- Andreas Jaeger aj@{suse.com,opensuse.org} Twitter/Identica: jaegerandi SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany GF: Felix Imendörffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nürnberg) GPG fingerprint = 93A3 365E CE47 B889 DF7F FED1 389A 563C C272 A126
_______________________________________________ Openstack-i18n mailing list Openstack-i18n@lists.openstack.org http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-i18n
Got your input, thank you. We will keep the country codes for bn and pa. Best regards Ying Chun Guo (Daisy) ਅਨਪੜ੍ਹ ਜਿਮੀਂਦਾਰ <jimidar@gmail.com> wrote on 09/05/2015 03:41:32 AM:
From: ਅਨਪੜ੍ਹ ਜਿਮੀਂਦਾਰ <jimidar@gmail.com> To: openstack-i18n@lists.openstack.org, openstack-infra@lists.openstack.org Date: 09/05/2015 03:42 AM Subject: Re: [Openstack-i18n] Final plans before translations switchto Zanata
Please find my comments for bn and pa inline.
Thanks Jimidar (Amandeep Singh)
On Thu, Aug 27, 2015 at 10:48 AM, Andreas Jaeger <aj@suse.com> wrote: On 08/27/2015 06:10 PM, Elizabeth K. Joseph wrote: On Thu, Aug 27, 2015 at 3:17 AM, Andreas Jaeger <aj@suse.com> wrote: Could we also revise languages and language codes?
I think we should prefer Country-independent languages over country ones.
So, we should prefer German ("de") over German in Germany ("de_DE") and German in Austria ("de_AT").
In case where we have country specific ones - like pl_PL -, can we rename them to country independent ones - like pl?
When asking for new languages, we should then get the independent one in first and only after consolidation with the existing team create a specific one.
What do you think?
I can't directly speak to the conversation that led to this list, but StevenK and Daisy came up with the following list, which seems to go by what you suggest:
ar as bg-BG
Rename to bg?
bn-IN
Rename to bn?
Bangla is used in two different countries. Bengali in India. and Bangla in Bangladesh. Can not generalize the term for it.
brx ca cs de el en-AU en-GB es es-MX fi-FI
Rename to fi?
fil fr gu he hi hu id it ja ka-GE
Rename to ka?
kn ko-KR
Can we rename to ko, please?
kok ks mai mni mr ne nl-NL
Rename to nl?
pa-IN
Rename to pa?
Punjabi (pa) has two scripts, one is Gurmukhi used in Indian side of Punjab and the other is Shahmukhi (Urdu) used in Pakistan's Punjab, so renaming to a generalized name wont do the justice to both entirely different scripts.
pl-PL
Can we rename the above to pl, please?
pt pt-BR ru sl-SI
Rename to sl?
sr ta te-IN
Rename to te?
tr-TR
Rename to tr?
ur vi-VN
Rename to vi?
zh-CN zh-TW
Copied from this pastebin that StevenK shared last night, in case it expires: http://paste.openstack.org/show/429137/
Background: We get *lots* of request on transifex for new languages, for example today for pl. There's right now no reason to have both pl and pl-PL. I'd like us to standardize here...
Andreas -- Andreas Jaeger aj@{suse.com,opensuse.org} Twitter/Identica: jaegerandi SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany GF: Felix Imendörffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nürnberg) GPG fingerprint = 93A3 365E CE47 B889 DF7F FED1 389A 563C C272 A126
_______________________________________________ Openstack-i18n mailing list Openstack-i18n@lists.openstack.org http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-i18n _______________________________________________ Openstack-i18n mailing list Openstack-i18n@lists.openstack.org http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-i18n
On 5 September 2015 at 05:41, ਅਨਪੜ੍ਹ ਜਿਮੀਂਦਾਰ <jimidar@gmail.com> wrote:
Please find my comments for bn and pa inline.
Thanks Jimidar (Amandeep Singh)
On Thu, Aug 27, 2015 at 10:48 AM, Andreas Jaeger <aj@suse.com> wrote:
pa-IN
Rename to pa?
Punjabi (pa) has two scripts, one is Gurmukhi used in Indian side of Punjab and the other is Shahmukhi (Urdu) used in Pakistan's Punjab, so renaming to a generalized name wont do the justice to both entirely different scripts.
Would "pa-Guru" work as a more general locale for Punjabi Gurmukhi? With "pa-Arab" for Punjabi Shahmukhi? -- Sean Flanigan Principal Software Engineer Globalisation Tools Engineering Red Hat
participants (10)
-
Akihiro Motoki
-
Andreas Jaeger
-
Carlos Munoz
-
Elizabeth K. Joseph
-
Jeremy Stanley
-
Remo Mattei
-
Sean Flanigan
-
Steve Kowalik
-
Ying Chun Guo
-
ਅਨਪੜ੍ਹ ਜਿਮੀਂਦਾਰ