Did you stop updating the Firefox version ?

[I-I-IT]

I know There was a spat between @gorhill and Mozilla AMO. But I believe you should still self-distribute it here on Github. Because, if not, many people will miss out on updates for filter list. Some people focused about security might go to Chromium if uBol is now unmaintained on Firefox. Thank you

[hj-collab]

@gorhill Same thoughts as OP. Please revive the uBol. 🙏 uBol is godsent for people with higher focus on security and low powered devices.

[hj-collab]

Last time in your 2 issues you were using uBOL for Chromium: #23 + #62, have you switched to Firefox recently or you use it on both Chromium & Firefox.

Yes, I've recently switched to Firefox completely and now use Chromium solely for testing.

[alsoLut]

if the burden is to much for @gorhill surely someone else could pick this minor task, right? The ideea of building a open source project is to share it with as many people as possible.

[stephenhawk8054]

It's a GPLv3 License project. Anyone can fork and publish it as long as they comply with the license. Dev could not prevent the "clones", and he's not obliged to do what he does not want to. Preventing malicious add-ons should be the jobs of AMO, not him.

[alsoLut]

It's a GPLv3 License project. Anyone can fork and publish it as long as they comply with the license. Dev could not prevent the "clones", and he's not obliged to do what he does not want to. Preventing malicious add-ons should be the jobs of AMO, not him.

I did not imply anything you mentioned above.

[stephenhawk8054]

Then the clones are not the issue, as long as they comply with the license.

[alsoLut]

Then the clones are not the issue, as long as they comply with the license.

The idea is to discourage clones by offering an official extension with numerous positive reviews.

[stephenhawk8054]

Dev is not obliged to discourage them if they comply the license correctly. If anything, that just forces him to put more work and time to something that's already causing many frictions. When uBO was not on Safari or Legacy Edge, somebody else volunteered to pick up the work with the stores and communicated back clearly, reducing dev's work and time. You can do it too, anyone else can do it too.

[gorhill]

surely someone else could pick this minor task

What is needed for this to happen is to have volunteers contributing by:

• Building uBO Lite locally
• Using the locally built version to verify that all work as expected for each release
  • Verify that at first install the extension first-install page opens as expected
  • Verify that first-install page doesn't open on subsequent launch
  • Verify that all filtering modes work as expected on many sites, with different set of rulesets enabled
  • Verify that admin deployment options work as expected
  • etc.

On both Firefox and Firefox for Android.

If I have guarantees that it is well tested, I would feel confident to again publish a version on AMO.

[gorhill]

There are many factors that must be considered, like which websites were visited and for how long, how are the blockers configured (they should be as similarly configured as possible, which other extensions were present, and so on

[I-I-IT]

If there is some way to objectively test it, I'll see if I can try it. While I'm sure you are right about what Firefox says in terms of CPU profiling, it could be something specific to this family of devices (it was the same on the Pixel 5 and 7 Pro, the 5 used a Qualcomm SoC and the 7 and 8 use Google's own silicon)

Kiwi Browser and Microsoft Edge have Android support for extensions. It would help to compare within the same browser.

[kuro68k]

Kiwi Browser and Microsoft Edge have Android support for extensions. It would help to compare within the same browser.

Both are based on Chromium though, and the issue is likely related to Firefox. Chromium has a more efficient filtering engine for Manifest V3 extensions than Firefox does.

I have been testing it out, and I haven't checked the changelog since earlier this year when I last tested it, but it does seem to be substantially better now. I'm trying to come up with a good test plan, but it isn't easy. For the time being I've switched back to uBO for daily use.

[kuro68k]

I did the best tests I could without automation, and I would say that whatever the issue was it is now fixed. Any difference in energy performance is within the margin of error. I saved an old nightly from a year ago and re-confirmed the issue there, so whatever it is has been fixed between then and now.