We currently have two severity systems in place: an old one based on keywords such as major, critical, etc., and a new one based on numbers (S1, S2, etc.).
We should convert the old style severity values to the new style, to complete the transition we started a few years ago.

As part of this bug, we should do:
severity==trivial or severity==minor -> S4
severity==normal -> S3
(severity==major or severity==critical or severity==blocker) and (crash signature is not empty or "crash" is in keywords) -> S2

In https://github.com/mozilla/relman-auto-nag/issues/1484 we will migrate the rest of the major+ (by letting teams retriage those).

Since it could be distracting to ask teams to retriage bugs in the "Testing" product, we could silently migrate the severity values for these bugs. Thus the last rule in comment 0 should change be as follows:

(severity==major or severity==critical or severity==blocker) and (crash signature is not empty or "crash" is in keywords or product==Testing) -> S2

Working on this now.

(In reply to Suhaib Mujahid [:suhaib] from comment #1)

Since it could be distracting to ask teams to retriage bugs in the "Testing" product, we could silently migrate the severity values for these bugs. Thus the last rule in comment 0 should change be as follows:

(severity==major or severity==critical or severity==blocker) and (crash signature is not empty or "crash" is in keywords or product==Testing) -> S2

Spoke with Suhaib yesterday and clarified that all bugs touched will be silently migrated and that no emails will be sent. I have the script ready to go to update the bugs but will be putting up for review first.

One more question, this is in affect for all products in BMO right? Not just Firefox related products? So the end result will be that we will just remove the old severities so they are no longer used?

I think we should do this only for Firefox-related products, and leave Thunderbird, etc., out as they might still be using the old severity levels.

In the context of https://github.com/mozilla/relman-auto-nag/issues/1484 , the following are the products that will be affected:

[
  "Core",
  "Developer Infrastructure",
  "DevTools",
  "External Software Affecting Firefox",
  "Fenix",
  "Firefox",
  "Firefox for iOS",
  "Firefox Build System",
  "GeckoView",
  "JSS",
  "NSPR",
  "NSS",
  "Remote Protocol",
  "Testing",
  "Toolkit",
  "Web Compatibility",
  "WebExtensions"
]

(In reply to Marco Castelluccio [:marco] from comment #4)

I think we should do this only for Firefox-related products, and leave Thunderbird, etc., out as they might still be using the old severity levels.

Thunderbird is not using old severity levels. So it is OK to run the script against products Thunderbird, Calendar, Chat Core, and Mailnews Core.

However, I'll take a quick spin through the potentially affected bugs - should only take an hour. (there is only 1,000 sev=minor bugs - ha!)

Oh thanks for letting us know. Suhaib then we can also update the autonag script to consider the products Wayne mentioned.

My review is done. Ready to go

(In reply to Marco Castelluccio [:marco] from comment #8)

Oh thanks for letting us know. Suhaib then we can also update the autonag script to consider the products Wayne mentioned.

Done in https://github.com/mozilla/relman-auto-nag/pull/1648

This will affect ~468 bugs, where teams will be asked to re-triage the bugs:
https://bugzilla.mozilla.org/buglist.cgi?f1=keywords&o1=nowords&v1=crash&product=Calendar&product=Chat%20Core&product=MailNews%20Core&product=Thunderbird&f2=cf_crash_signature&o2=isempty&bug_severity=blocker&bug_severity=critical&bug_severity=major&resolution=---

Shouldn't this be limited to open bugs?

(In reply to :glob ✱ from comment #11)

Shouldn't this be limited to open bugs?

Yes, I thought I mentioned it in comment 0 but it looks like I didn't :)
Thanks for catching that!

Updated the script with the new search criteria. Here are the counts.

https://bugzilla.mozilla.org/buglist.cgi?o3=equals&v3=Testing&v2=crash&o2=equals&o1=isnotempty&bug_severity=major&bug_severity=critical&bug_severity=blocker&product=Calendar&product=Chat+Core&product=Core&product=Developer+Infrastructure&product=DevTools&product=External+Software+Affecting+Firefox&product=Fenix&product=Firefox+Build+System&product=Firefox+for+iOS&product=Firefox&product=GeckoView&product=JSS&product=MailNews+Core&product=NSPR&product=NSS&product=Remote+Protocol&product=Testing&product=Thunderbird&product=Toolkit&product=Web+Compatibility&product=WebExtensions&resolution=---&f3=product&f2=keywords&j_top=OR&f1=cf_crash_signature

742 bugs found.

https://bugzilla.mozilla.org/buglist.cgi?resolution=---&product=Calendar&product=Chat+Core&product=Core&product=Developer+Infrastructure&product=DevTools&product=External+Software+Affecting+Firefox&product=Fenix&product=Firefox+Build+System&product=Firefox+for+iOS&product=Firefox&product=GeckoView&product=JSS&product=MailNews+Core&product=NSPR&product=NSS&product=Remote+Protocol&product=Testing&product=Thunderbird&product=Toolkit&product=Web+Compatibility&product=WebExtensions&bug_severity=normal

Over 10000 bugs found (buglist.cgi caps at 10000).

https://bugzilla.mozilla.org/buglist.cgi?bug_severity=minor&bug_severity=trivial&resolution=---&product=Calendar&product=Chat+Core&product=Core&product=Developer+Infrastructure&product=DevTools&product=External+Software+Affecting+Firefox&product=Fenix&product=Firefox+Build+System&product=Firefox+for+iOS&product=Firefox&product=GeckoView&product=JSS&product=MailNews+Core&product=NSPR&product=NSS&product=Remote+Protocol&product=Testing&product=Thunderbird&product=Toolkit&product=Web+Compatibility&product=WebExtensions

4159 bugs found.

This change should have preserved the last modified time. It makes "Sort result by Last Update" useless.

(In reply to Masatoshi Kimura [:emk] from comment #14)

This change should have preserved the last modified time. It makes "Sort result by Last Update" useless.

This is unfortunate in how BMO database is setup that if the update timestamp was not updated as part of normal bug update process, then we would have database discrepancies that could cause unknown problems.

If it is helpful, we did add a new field to the REST API recently that gives the last modified timestamp of a bug that was made by human (non-bot). The field could be used for sorting purposes if you are running a dashboard, etc. which will give you the actual sort order you need.

More info here:
https://bmo.readthedocs.io/en/latest/api/core/v1/bug.html#rest-single-bug
Search the page for last_change_time_non_bot

Example with all default fields plus the optional new field:
https://bugzilla.mozilla.org/rest/bug/35?include_fields=_default,last_change_time_non_bot

Let me know if you have questions

(In reply to David Lawrence [:dkl] from comment #13)

Updated the script with the new search criteria. Here are the counts.

All bugs have been updated so closing this bug.

(In reply to David Lawrence [:dkl] from comment #15)

Let me know if you have questions

I filed bug 1794888.