)]}'
{"/PATCHSET_LEVEL":[{"author":{"_account_id":9816,"name":"Takashi Kajinami","email":"kajinamit@oss.nttdata.com","username":"kajinamit"},"change_message_id":"a5d198f38997c7a7bc7acc0857a4ca331953bcd0","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"186caebd_08937689","updated":"2023-03-03 05:14:04.000000000","message":"I understand this is annoying but I\u0027m not too sure we should backport this considering its interface impact. In case we get 14.0.1 released with this fix then upgrade from 14.0.1 to 14.2.0(Antelope) causes regression.\n\nI won\u0027t block this if the other oslo cores are fine with backporting this but just leaving my concern for records.","commit_id":"f4766cbf45fdb77318e372dc89572351026385c1"},{"author":{"_account_id":28619,"name":"Dmitriy Rabotyagov","email":"noonedeadpunk@gmail.com","username":"noonedeadpunk"},"change_message_id":"fe9176a9ad84056f165c35d01ee82fc36735702e","unresolved":false,"context_lines":[],"source_content_type":"","patch_set":1,"id":"6738cd29_7ccc009f","in_reply_to":"186caebd_08937689","updated":"2023-03-03 13:55:29.000000000","message":"I\u0027m not sure it\u0027s really have an impact other then alerting in logs that deprecated variable is being used.\n\nOr what kind of regression do you mean here? I can imagine only weird release note renderring, but I can probably drop or rename release note so it was identified properly.","commit_id":"f4766cbf45fdb77318e372dc89572351026385c1"},{"author":{"_account_id":9816,"name":"Takashi Kajinami","email":"kajinamit@oss.nttdata.com","username":"kajinamit"},"change_message_id":"ca87581631f37806a8dd4509d74b4ba54c5059c1","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":1,"id":"578b6118_c6d31139","in_reply_to":"49a93579_dac7e74f","updated":"2024-01-24 15:26:25.000000000","message":"I think it\u0027s ok to backport this to stable/2023.1, because stable/2023.2 already supports both the new and old parameters.\n\nRegarding stable/zed, we may be able to backport this, by the following steps.\n 1. We backport this to stable/2023.1\n 2. Create a new stable/2023.1 release (14.4.2)\n 3. Backport this and clearly mention that users have to use the version newer than 14.4.2 when the deployment is being upgrade","commit_id":"f4766cbf45fdb77318e372dc89572351026385c1"},{"author":{"_account_id":9816,"name":"Takashi Kajinami","email":"kajinamit@oss.nttdata.com","username":"kajinamit"},"change_message_id":"98ac1691f57065f1b1795d6fd9df3c386de3e904","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":1,"id":"86288fdd_e7999d4a","in_reply_to":"6738cd29_7ccc009f","updated":"2023-03-07 11:18:52.000000000","message":"There can be a deployment which has 14.0.1 (zed with this fix imported) and users might pull 14.2.0 (antelope) while they updates the deployment. This is the valid update path, based on the version numbers.\n\nThe problem here is, in case the user users the new parameter name in 14.0.1, then the option is suddenly ignored to a newer version, and this is the \"regression\" I\u0027m concerned with.","commit_id":"f4766cbf45fdb77318e372dc89572351026385c1"},{"author":{"_account_id":28522,"name":"Hervé Beraud","email":"herveberaud.pro@gmail.com","username":"hberaud"},"change_message_id":"d3ca8ffe3e4ec47b983600b8f71ef18725de5d30","unresolved":true,"context_lines":[],"source_content_type":"","patch_set":1,"id":"49a93579_dac7e74f","in_reply_to":"86288fdd_e7999d4a","updated":"2024-01-24 13:41:07.000000000","message":"Interesting situation...\n\nTechnically speaking, the original version of this patch was merged during the Antelope series, at February 22 [1].\nThis date is posterior to the lib freeze date of Antelope (February 17) [2].\nCore developers validations for this original patch were made around February 14 [1], which is correct and before the lib freeze [3].\n\nHowever, the gates was stucks and this original patch required a couple of recheck, this is why it was merged at February 22.\n\nAt this point the final release for oslo.messaging was already made and the 2023.1/antelope branch cut was made by refering to a release/sha that correspond to a parent commit of this patch. A parent commit made before the original commit of this backport. So the original commit was merged on a master that correspond 2023.2/bobcat.\n\nHence, Antelope do not includes this fix. Takashi is right. If we merge this patch right now on Zed, then 2023.1/Antelope/14.2.x won\u0027t be compatible with this change.\n\nFor now only 2023.2/bobcat is compatible with this backport.\n\nEither we propose to backport it to 2023.1/antelope first, and then on Zed, or we abandon this backport.\n\n@Takashi: Thoughts?\n\n\n[1] https://review.opendev.org/c/openstack/oslo.messaging/+/873630\n[2] https://review.opendev.org/c/openstack/releases/+/874452\n[3] https://releases.openstack.org/antelope/schedule.html","commit_id":"f4766cbf45fdb77318e372dc89572351026385c1"}]}