1135139 - Increment the versions in install.rdf when we sign the add-on

Status: RESOLVED FIXED

(addons.mozilla.org Graveyard :: Maintenance Scripts, defect)

Description

[Wil Clouser [:clouserw]]

This is only for the big "sign add-on" script in bug 1070227.  After that initial run newly submitted add-ons should have larger versions anyway.

So, for that script we need to increment the actual version value in the install.rdf file in the .zip.  Mossop assures me that simply tacking a .1 onto the end of the value will work, no matter what the existing value is, so we shouldn't need to parse it or try to figure out their versioning scheme.

Comment 1

[Mathieu Agopian [:magopian]]

PR: https://github.com/mozilla/olympia/pull/463

Comment 2

[Mathieu Agopian [:magopian]]

Here's the content of a comment on github (https://github.com/mozilla/olympia/pull/463#discussion_r26173902):

"""
This isn't going to work. See https://bugzil.la/1070227#c3

For version numbers which already have 4 components, this is going to make the new version number lower than the current version number. This turned out to be a significant issue in last round of SDK repacks.
"""

Kris also said "We already have code for dealing with toolkit versions in validator.version. I think the easiest thing to do would be to make these objects mutable. I wrote that module, so let me know if you want me to do that."

Dave, what do you think?

Comment 3

[Dave Townsend [:mossop]]

Bug 1070227 comment 3 contains some incorrect information. It claims that Firefox only recognizes 4-component version numbers. Unless you plan on supporting Firefox 1.0 this is not true. Firefox accepts an unlimited number of version components. Perhaps AMO has a limit on the number of version components?

Comment 4

[Mathieu Agopian [:magopian]]

Thanks Dave for the reply, Kris, do you have a clue about that?

Comment 5

[Kris Maglione [:kmag]]

Hm. You're right. I was going off what the toolkit versions MDN page used to say. But I actually ported the version comparator to Python with the correct behavior, so I guess I have no excuse.

I think we're good, then.

Comment 6

[Mathieu Agopian [:magopian]]

Fixed in https://github.com/mozilla/olympia/commit/78a206e15631d05fc8b531c081163e5ebef30e5c