Re: [openstack-dev] [cinder][globalization] Need input on how to proceed .
Thanks for the input Duncan. The removal of the debug logs is really a separate issue. I was just hoping to reduce the number of patches that would touch a large number of files. As we are thinking through this though, it really is a separate change so it is best to do separate patches. On a related note, we shouldn't be taking changes that have debug messages translated if we are moving forward with removing translation of debug messages. Jay On Thu, 2014-05-01 at 16:41 +0100, Duncan Thomas wrote: That sounds like a sensible way forward, yes. If the dependency is not need, then great, makes review and merge even easier. Thanks On 28 April 2014 17:03, Jay S. Bryant jsbry...@electronicjungle.net wrote: Duncan, Thanks for the response. Have some additional thoughts, in-line, below: On Mon, 2014-04-28 at 12:15 +0100, Duncan Thomas wrote: Two separate patches, or even two chains of separate patches, will make reviewing and more importantly (hopefully temporary) backouts easier. It will also reduce the number of merge conflicts, which are still likely to be substantial. True, I suppose we need to keep in mind the fact that we might want to make this be easy to back-out in the future. Hopefully it isn't an issue this time around though. There's no benefit at all to all of this being done in one patch, and substantial costs. Doing the conversion by sections seems like the way forward. So, let me propose a different process here. Handling the i18n and removal of debug separately instead. First, propose one patch that will add the explicit import of '_' to all files. There will be a lot of files touched, but they all will be 1 liners. Then make the patch for the re-enablement of lazy tanslation a second patch that is dependent upon the first patch. Then handle removal of _() from DEBUG logs as a separate issue once the one above has merged. For that change do it in multiple patches divided by section. Make the sections be the top level directories under cinder/ ? Does that sound like a reasonable plan? Doing both around the same time (maybe as dependant patches) seems reasonable As I think about it, I don't know that the debug translation removal needs to be dependent, but we could work it out that way if you feel that is important. Let me know what you think. Thanks! On 27 April 2014 00:20, Jay S. Bryant jsbry...@electronicjungle.net wrote: All, I am looking for feedback on how to complete implementation of i18n support for Cinder. I need to open a new BluePrint for Juno as soon as the cinder-specs process is available. In the mean time I would like to start working on this and need feedback on the scope I should undertake with this. First, the majority of the code for i18n support went in with Icehouse. There is just a small change that is needed to actually enable Lazy Translation again. I want to get this enabled as soon as possible to get plenty of runtime on the code for Icehouse. The second change is to add an explicit export for '_' to all of our files to be consistent with other projects. [1] This is also the safer way to implement i18n. My plan is to integrate the change as part of the i18n work. Unfortunately this will touch many of the files in Cinder. Given that fact, this brings me to the item I need feedback upon. It appears that Nova is moving forward with the plan to remove translation of debug messages as there was a recent patch submitted to enable a check for translated DEBUG messages. Given that fact, would it be an appropriate time, while adding the explicit import of '_' to also remove translation of debug messages. It is going to make the commit for enabling Lazy Translation much bigger, but it would also take out several work items that need to be addressed at once. I am willing to undertake the effort if I have support for the changes. Please let me know your thoughts. Thanks!2] Jay (jungleboyj on freenode) [1] https://bugs.launchpad.net/cinder/+bug/1306275 ___ OpenStack-dev mailing list OpenStack-dev@lists.openstack.org http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev ___ OpenStack-dev mailing list OpenStack-dev@lists.openstack.org http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev ___ OpenStack-dev mailing list OpenStack-dev@lists.openstack.org http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
Re: [openstack-dev] [cinder][globalization] Need input on how to proceed .
That sounds like a sensible way forward, yes. If the dependency is not need, then great, makes review and merge even easier. Thanks On 28 April 2014 17:03, Jay S. Bryant jsbry...@electronicjungle.net wrote: Duncan, Thanks for the response. Have some additional thoughts, in-line, below: On Mon, 2014-04-28 at 12:15 +0100, Duncan Thomas wrote: Two separate patches, or even two chains of separate patches, will make reviewing and more importantly (hopefully temporary) backouts easier. It will also reduce the number of merge conflicts, which are still likely to be substantial. True, I suppose we need to keep in mind the fact that we might want to make this be easy to back-out in the future. Hopefully it isn't an issue this time around though. There's no benefit at all to all of this being done in one patch, and substantial costs. Doing the conversion by sections seems like the way forward. So, let me propose a different process here. Handling the i18n and removal of debug separately instead. First, propose one patch that will add the explicit import of '_' to all files. There will be a lot of files touched, but they all will be 1 liners. Then make the patch for the re-enablement of lazy tanslation a second patch that is dependent upon the first patch. Then handle removal of _() from DEBUG logs as a separate issue once the one above has merged. For that change do it in multiple patches divided by section. Make the sections be the top level directories under cinder/ ? Does that sound like a reasonable plan? Doing both around the same time (maybe as dependant patches) seems reasonable As I think about it, I don't know that the debug translation removal needs to be dependent, but we could work it out that way if you feel that is important. Let me know what you think. Thanks! On 27 April 2014 00:20, Jay S. Bryant jsbry...@electronicjungle.net wrote: All, I am looking for feedback on how to complete implementation of i18n support for Cinder. I need to open a new BluePrint for Juno as soon as the cinder-specs process is available. In the mean time I would like to start working on this and need feedback on the scope I should undertake with this. First, the majority of the code for i18n support went in with Icehouse. There is just a small change that is needed to actually enable Lazy Translation again. I want to get this enabled as soon as possible to get plenty of runtime on the code for Icehouse. The second change is to add an explicit export for '_' to all of our files to be consistent with other projects. [1] This is also the safer way to implement i18n. My plan is to integrate the change as part of the i18n work. Unfortunately this will touch many of the files in Cinder. Given that fact, this brings me to the item I need feedback upon. It appears that Nova is moving forward with the plan to remove translation of debug messages as there was a recent patch submitted to enable a check for translated DEBUG messages. Given that fact, would it be an appropriate time, while adding the explicit import of '_' to also remove translation of debug messages. It is going to make the commit for enabling Lazy Translation much bigger, but it would also take out several work items that need to be addressed at once. I am willing to undertake the effort if I have support for the changes. Please let me know your thoughts. Thanks!2] Jay (jungleboyj on freenode) [1] https://bugs.launchpad.net/cinder/+bug/1306275 ___ OpenStack-dev mailing list OpenStack-dev@lists.openstack.org http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev ___ OpenStack-dev mailing list OpenStack-dev@lists.openstack.org http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev -- Duncan Thomas ___ OpenStack-dev mailing list OpenStack-dev@lists.openstack.org http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
Re: [openstack-dev] [cinder][globalization] Need input on how to proceed .
Two separate patches, or even two chains of separate patches, will make reviewing and more importantly (hopefully temporary) backouts easier. It will also reduce the number of merge conflicts, which are still likely to be substantial. There's no benefit at all to all of this being done in one patch, and substantial costs. Doing the conversion by sections seems like the way forward. Doing both around the same time (maybe as dependant patches) seems reasonable On 27 April 2014 00:20, Jay S. Bryant jsbry...@electronicjungle.net wrote: All, I am looking for feedback on how to complete implementation of i18n support for Cinder. I need to open a new BluePrint for Juno as soon as the cinder-specs process is available. In the mean time I would like to start working on this and need feedback on the scope I should undertake with this. First, the majority of the code for i18n support went in with Icehouse. There is just a small change that is needed to actually enable Lazy Translation again. I want to get this enabled as soon as possible to get plenty of runtime on the code for Icehouse. The second change is to add an explicit export for '_' to all of our files to be consistent with other projects. [1] This is also the safer way to implement i18n. My plan is to integrate the change as part of the i18n work. Unfortunately this will touch many of the files in Cinder. Given that fact, this brings me to the item I need feedback upon. It appears that Nova is moving forward with the plan to remove translation of debug messages as there was a recent patch submitted to enable a check for translated DEBUG messages. Given that fact, would it be an appropriate time, while adding the explicit import of '_' to also remove translation of debug messages. It is going to make the commit for enabling Lazy Translation much bigger, but it would also take out several work items that need to be addressed at once. I am willing to undertake the effort if I have support for the changes. Please let me know your thoughts. Thanks!2] Jay (jungleboyj on freenode) [1] https://bugs.launchpad.net/cinder/+bug/1306275 ___ OpenStack-dev mailing list OpenStack-dev@lists.openstack.org http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev -- Duncan Thomas ___ OpenStack-dev mailing list OpenStack-dev@lists.openstack.org http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
Re: [openstack-dev] [cinder][globalization] Need input on how to proceed .
Duncan, Thanks for the response. Have some additional thoughts, in-line, below: On Mon, 2014-04-28 at 12:15 +0100, Duncan Thomas wrote: Two separate patches, or even two chains of separate patches, will make reviewing and more importantly (hopefully temporary) backouts easier. It will also reduce the number of merge conflicts, which are still likely to be substantial. True, I suppose we need to keep in mind the fact that we might want to make this be easy to back-out in the future. Hopefully it isn't an issue this time around though. There's no benefit at all to all of this being done in one patch, and substantial costs. Doing the conversion by sections seems like the way forward. So, let me propose a different process here. Handling the i18n and removal of debug separately instead. First, propose one patch that will add the explicit import of '_' to all files. There will be a lot of files touched, but they all will be 1 liners. Then make the patch for the re-enablement of lazy tanslation a second patch that is dependent upon the first patch. Then handle removal of _() from DEBUG logs as a separate issue once the one above has merged. For that change do it in multiple patches divided by section. Make the sections be the top level directories under cinder/ ? Does that sound like a reasonable plan? Doing both around the same time (maybe as dependant patches) seems reasonable As I think about it, I don't know that the debug translation removal needs to be dependent, but we could work it out that way if you feel that is important. Let me know what you think. Thanks! On 27 April 2014 00:20, Jay S. Bryant jsbry...@electronicjungle.net wrote: All, I am looking for feedback on how to complete implementation of i18n support for Cinder. I need to open a new BluePrint for Juno as soon as the cinder-specs process is available. In the mean time I would like to start working on this and need feedback on the scope I should undertake with this. First, the majority of the code for i18n support went in with Icehouse. There is just a small change that is needed to actually enable Lazy Translation again. I want to get this enabled as soon as possible to get plenty of runtime on the code for Icehouse. The second change is to add an explicit export for '_' to all of our files to be consistent with other projects. [1] This is also the safer way to implement i18n. My plan is to integrate the change as part of the i18n work. Unfortunately this will touch many of the files in Cinder. Given that fact, this brings me to the item I need feedback upon. It appears that Nova is moving forward with the plan to remove translation of debug messages as there was a recent patch submitted to enable a check for translated DEBUG messages. Given that fact, would it be an appropriate time, while adding the explicit import of '_' to also remove translation of debug messages. It is going to make the commit for enabling Lazy Translation much bigger, but it would also take out several work items that need to be addressed at once. I am willing to undertake the effort if I have support for the changes. Please let me know your thoughts. Thanks!2] Jay (jungleboyj on freenode) [1] https://bugs.launchpad.net/cinder/+bug/1306275 ___ OpenStack-dev mailing list OpenStack-dev@lists.openstack.org http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev ___ OpenStack-dev mailing list OpenStack-dev@lists.openstack.org http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
