311700 - Extension-manager should update disabled extensions to new extension-version even new extension is not compatible with browser-version

Status: RESOLVED WONTFIX

(Toolkit :: Add-ons Manager, enhancement)

Description

[Tobias B. Besemer [:BesTo] (QA)]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8b4) Gecko/20050908 Firefox/1.4
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8b4) Gecko/20050908 Firefox/1.4

Extension-manager should update to new extension-version even new extension is
not compatible with browser-version.
For example: istalled extension xyz 0.1 -> update avaible version 0.2 but not
compatible with installed browser version -> extension-manager should update to
new version even version ist not compatible with browser-version.
Browser should allways stay on top with new versions.
1. You can downgrade to an older browser version without updating extensions.
2. You will see that there is still development going on on extension.

Reproducible: Always

Comment 1

[Dave Townsend [:mossop]]

I would say this is a dupe of bug 294644. Reporter, please reopen if that is not
what you are looking for.


*** This bug has been marked as a duplicate of 294644 ***

Comment 2

[Tobias B. Besemer [:BesTo] (QA)]

Sorry, but Im not sure if this is really the same. -> Marked as depends on.
Will check it after the other bug is fixed.

Comment 3

[Robert Strong (they/them - no direct email)]

You're asking that we allow extensions to be updated to an incompatible version.
What about if the old extension is compatible... then the user would have an
extension that no longer works that used to work. What value is there really
with installing a newer version if it is disabled? Most updates provided by
extension authors are already compatible with the latest release so this would
be an edgecase and add significant complexity to the code while providing little
value for most users. Seeing that development is going on provides little solace
to a user that wants the extension to work. When downgrading it is just about as
easy to update after downgrading. As it stands this bug is wontfix - there are
obvious issues that providing this functionality would cause for most users that
have not been addressed.

Comment 4

[Tobias B. Besemer [:BesTo] (QA)]

Hope that makes it more clear. :-)

Comment 5

[Dave Townsend [:mossop]]

Do you mean app-disabled extension versions or those disabled by the user?

If I disable an extension I don't want it to then get updated to a newer version
that I can't then enable.

Comment 6

[Robert Strong (they/them - no direct email)]

None of the issues that providing this feature would cause have been answered
(see comment #3) and the value it would provide as stated in comment #0 is
minimal as I see it especially in relation to these issues. The value of showing
that development is going on for an extension is greatly reduced considering
that an when an extension author updates an extension the vast majority of the
time they update it to *be* compatible with the latest version. The value of
being able to update so a person can then install a previous release is easily
worked around by just updating the extension after they downgrade as well as
only providing value to those that downgrade.

With the effort to provide this, the additional complexity that would be needed
in the code, the value as stated, etc. makes this wontfix.

Comment 7

[Tobias B. Besemer [:BesTo] (QA)]

(In reply to comment #5)
> Do you mean app-disabled extension versions or those disabled by the user?
> 
> If I disable an extension I don't want it to then get updated to a newer version
> that I can't then enable.
Maybee there should be an question-dialog.
Course I disabled some extensions course the gave me errors although they should
be compatible. ... ???

Comment 8

[Tobias B. Besemer [:BesTo] (QA)]

(In reply to comment #6)
> None of the issues that providing this feature would cause have been answered
> (see comment #3) and the value it would provide as stated in comment #0 is
> minimal as I see it especially in relation to these issues. The value of showing
> that development is going on for an extension is greatly reduced considering
> that an when an extension author updates an extension the vast majority of the
> time they update it to *be* compatible with the latest version. The value of
> being able to update so a person can then install a previous release is easily
> worked around by just updating the extension after they downgrade as well as
> only providing value to those that downgrade.
> 
> With the effort to provide this, the additional complexity that would be needed
> in the code, the value as stated, etc. makes this wontfix.

Sorry, but my english seems to be to base to understand everything you write ...
sorry.
An example:
I use the sender verification extension in Thunderbird, but now Im upraded to
1.5 Beta 2. In my extension-manager there is the version 0.73 but Im know that
there is an 0.75 out. Im now try to use the tester-extension to activate the
incompatible extensions but (I havend tested the tool fully yet) it seems that
the extension isnt able to update them couse the new version is still
incompatible. I also will write the test-extension-author the problem but I
think this bug should be markted as "future" and "confirmed" until I can come
back end fix it by my self with an patch.

Comment 9

[Robert Strong (they/them - no direct email)]

According to your statement in comment #8 you want to install a new version of
an extension to verify that it works with the latest version of Firefox. So the
value now is to be able to install an incompatible version in order to verify
that it works with a newer version and no longer as you stated previously:
> 1. You can downgrade to an older browser version without updating extensions.
> 2. You will see that there is still development going on on extension.

This value statement is now for people testing nightly / beta builds. So far the
issues brought up haven't been addressed and in bug 294644 you were told about a
couple of methods where you would be able to install extensions which do work
though you had problems with one of them.

Since the target audience for this feature are people not using release builds
and for the other reasons cited here along with the other bugs that are similar
even if you do provide a patch chances are it would never be approved or landed.
This seems much more appropriate to provide as an extension and not as part of
the product itself but I'll think on this some more but as is it is still
wontfix. In many ways comment #8 makes this bug the same as bug 294644 with a
different approach possibly because you can then use the Nightly Tester Tools to
make it compatible if this bug was implemented. As is, the Nightly Tester Tools
will allow you to install incompatible extensions as others have stated
elsewhere even though you had problems doing so.

Comment 10

[Tobias B. Besemer [:BesTo] (QA)]

(In reply to comment #9)
Wrote an mail to the nightly-tester-tool-author:

Hey.

Can U please have a look at my bugs and comments (filled as
mozilla.bugilla.lV-crew.org)
https://bugzilla.mozilla.org/show_bug.cgi?id=311700
https://bugzilla.mozilla.org/show_bug.cgi?id=294644#c12

My problem is that I know there is an updated (even incompatible) version of an
extension, but I cant install them course the update says that it isnt compatible.
Can U add the overwrite-browser-version to the automated- and hand-maade-update ???
Plz contact me. Thx in advance ...

Greets slim.

Comment 11

[Dave Townsend [:mossop]]

So far as I can make out, your problem is that even with the Nightly Tester
Tools extension installed, you cant "update" an extension to a newer (but still
incompatible) version of that extension.

This isn't the place to discuss this. These kind of problems are issues with
that extension, not Firefox bugs. Further problems of this sort should be dealt
with by email to the extension author or on the support forums, not in bugzilla.

Changing this back to WONTFIX as per Rob's decision.