Why is Firefox abandoning its strength in customizability?!

[billyswong]

Those who stay active in this forum or follow Firefox's development non-stop may nay-say and said users like me are just lazy, as announcement such as blah blah blah has been posted for xyz long beforehand. But sorry, a normal people don't actively read roadmaps and discussion of a software's future development regularly *unless* he or she find something in that software unsatisfactory and/or have room in improvement. So for many people like me, a long time user of Firefox, the kill-off plan of "legacy" extensions + no more UI customizability in new extension system comes as a huge shock. (And feels like the Earth that got killed in the beginning of "The Hitchhiker's Guide to the Galaxy".)

The Classic Theme Restorer extension say explicitly they can't port it to the new extension system because the new system dropped features. User Agent Switcher and Right Encoding has not been "updated" for a long time and they worry me too. (Right Encoding is for fixing mis-guessed text encoding for webpages in frames / iframes manually) More popular extensions like Greasemonkey and Flash and Video Download should survive this mess, or so I hope. Less famous ones' future like HLS Stream Detector and Dictionary Lookup Extension are gloomier.

I heard one of the reasons that Firefox developers decide to restrict extensions in the future are for "security". Come on, installing extensions are akin to installing native programs and of course some of them, especially if not posted in official Firefox add-on site, could be malware in disguise! Banning it all is extreme babysitting. And the another reason talking of old extensions stepping on each others' shoes... new system should improve that part or tell extension users to bite the bullet (and decide which combination of extensions one should install wisely), not chopping toes like in the current plan.

Sidenote: I am suspecting such decision of chopping toes may be partly due to those opt-in telemetry being misused for user-pattern analysis and heavily biased unfavourably to certain kind of users.

[allande]

If you want to run legacy extensions for a while, you can use Firefox 52 ESR until May / June 2018. Installing it may require building a new profile if you have run Firefox 55.

There are two issues here. Killing legacy extensions, and not allowing much UI customization in the new webextensions. Would you agree that Firefox needs to be a good multi-process browser? Ever since Chrome came out that was pretty clearly essential, and clock speeds have hit the wall and more cores is the way things are going. Some extension developers rewrote extensions to be compatible with multiprocess, but not many. If multi-process was fully turned on, full speed, no compatibility shims, that right there kills most legacy extensions. As a technology issue, it seems like that had to happen. If it was done years ago, Firefox might be in a better place now.

The second issue is future UI adjustment. Mozilla claims they will be willing to work with extension writers to extend the webextensions api and add some new features. However, features that affect the UI are usually not making the cut. Instead of being a technical issue, I believe this is a political one, I don't see why UI changes couldn't with security and performance be added to the API. IMO somebody decided they didn't want those type of extensions. But this is in a way good news - if it isn't technical and is just a choice, they can change their mind, which they may if they feel they are getting a bad response to the new browser. They'll just have to hope that some extension developers will come back and work with them, since all of these events must have been discouraging to extension writers.

[billyswong]

I am fine with Firefox killing legacy extensions some day. It is Classic Theme Restorer saying they can't port it that alerts me the change is not just about bringing technology forward.

Then CTR pointed me to this bugzilla page: https://bugzilla.mozilla.org/show_bug.cgi?id=1328244 Those !@#$! said WONTFIX because "extensions and themes won't be able to keep up". Such arrogant behaviour is astounding. Now I am more worrying than before. Mozilla may have been infected and trojaned by Google like what what happened to Nokia by Microsoft.

[the-edmeister]

Some of the features in CTR can be "ported" to WebExtensions, but the main feature for which CRT was created in the first place relies upon XUL and that will be gone from Firefox 57. If Aris would have "ported" CTR to WE the end result would be something that could be called CTR but wouldn't have the "retro" toolbar arrangement that started with Firefox 4.0 and ended with the Australis UI - the whole reason that CTR was originally written! What's the point in doing that?


.

[Sugoi]

Everything has its pros and cons. FF's strength is at the same time also its weakness. I personally embrace the change. If it turns out that it's not to my liking I'll simply look for an alternative.

...Some extension developers rewrote extensions to be compatible with multiprocess, but not many...

Some devs don't bother as the legacy extension will be completely useless as from FF57 anyways. So instead they put all the effort in making a WebExt port.

[billyswong]

I am not blaming Aris the author of CTR. I am complaining Firefox development team finally got arrogant enough (or hijacked enough) to kill UI customizability. According to another link from CTR, https://blog.mozilla.org/addons/2017/04 ... ent-223852 the product manager said "Some top add-ons can’t be completely ported (DownThemAll and Firebug come to mind) and we’re accepting that risk." I am not a add-on writer and I don't know how restrictive they want WebExtensions to keep as is. But from how lacking Google Chrome's extensions in features and how the Firefox team behave so far, the future of user freedom is doomed. Some not-really-UI-related extensions may probably be also un-portable. I have already clicked yes and updated to Fx55 before discovering this damn change, so I guess I may stay in Fx56 forever until this computer breaks.

(I use CTR for clear&square big reload-page button left of location bar, and space-efficient square tabs)

[allande]

The Photon theme coming with Firefox 57 has square tabs. The downgrade to 52 ESR isn't too difficult if you are willing to lose your history by using a bookmark backup instead of the places.sqlite file.

[billyswong]

The Photon theme coming with Firefox 57 has square tabs. The downgrade to 52 ESR isn't too difficult if you are willing to lose your history by using a bookmark backup instead of the places.sqlite file.

Good news for that. But concerning other "legacy" functionality extensions not porting yet (or ever), I will still postpone going Fx57 indefinitely.

[Brummelchen]

downgrade is no longer possible for you without loosing data - you are already on v55 and v55 had some major changes which are not downward compatible. eg favicons.

(ignored the rest of your rant, dont care)

[GHM113]

I think the answer is very simple. CTR has ~335k users, Firefox has ~80m users so only 0.41% of Firefox user base have CTR installed. In other words, nobody needs CTR except for a very small group of people. Same story with other legacy addons that often used to criticize WebExtensions: Tree Style Tab - 100k users, Tab Mix Plus - 600k users, DTA - 1m users, Session Manager - 200k users and so on.

[flaneurb]

I think the answer is very simple. ..., nobody needs CTR except for a very small group of people. Same story with other legacy addons that often used to criticize WebExtensions: ... and so on.

Awesome

[mightyglydd]

Nothing new about AMO diddling with Extension and Theme download/usage numbers ..why would you believe they're accurate now ?
If it suited the Mozilla agenda they'd say 99% of users use CTR.....
FYI: https://blog.mozilla.org/addons/2011/06 ... -on-users/

@There were some interesting threads about this here, but I can't find the (C.Nicks/Frank ?) related one I recall....

[therube]

(And not to be harsh, only taking a contrarian route...)

Would you agree that Firefox needs to be a good multi-process browser?

I guess you have to put me in the "show me" category.
Show me how MP improves my browsing experience.

As far as I can tell, I've done (& am doing) quite well with a non-MP browser.

What I can tell you is, given a trade-off between a MP browser or being able to use the extensions I use, well, I guess its obvious.

Ever since Chrome

Oh, the 'ol Chome argument.

Some extension developers rewrote extensions to be compatible with multiprocess, but not many.

And a good thing that.
How would you feel that you've spent all this time to do MP, only to be punched in the gut, when Mozilla says, oh, by the way, your new MP extension will shortly no longer be allowed to work - because we're jumping to webextensions!

Oh, & what's that you say? Your new MP extension - that you may want to now convert to webextensions - what, cannot be made as a webextension because there is no API to do the functions you need, well that's OK, because we won't allow those functions anyhow! WONTFIX. (Fun .)

If it was done years ago, Firefox might be in a better place now.

If it was done, with more forethought, in a better manner, FF might be in a better place now.

[therube]

If it suited the Mozilla agenda they'd say 99% of users use CTR.....
FYI: https://blog.mozilla.org/addons/2011/06 ... -on-users/

Interesting page that.
Funny how that now longer is the situation today .

The 'ol numbers game.


Likewise The New Firefox and Ridiculous Numbers of Tabs, points out the great speed increase in FF 55 startup for users that use lots of tabs. 1691 tabs!

And then on the other hand, FF chides users who use lots of tabs, "725 tabs, 62 windows" sorry, but that's just absurd. What happened to your testcase of 500 tabs - that wasn't enough??

And I'm sure their very own telemetry data shows that there are very few tabaholics.
So given that no one uses tabs, why bother.

(I am glad to see those improvements coming from Quantum Flow.)


(Disclosure: My name is therube, & I am a tabaholic.)

[JodyThornton]

If you want to run legacy extensions for a while, you can use Firefox 52 ESR until May / June 2018. Installing it may require building a new profile if you have run Firefox 55.

There are two issues here. Killing legacy extensions, and not allowing much UI customization in the new webextensions. Would you agree that Firefox needs to be a good multi-process browser? Ever since Chrome came out that was pretty clearly essential, and clock speeds have hit the wall and more cores is the way things are going. Some extension developers rewrote extensions to be compatible with multiprocess, but not many. If multi-process was fully turned on, full speed, no compatibility shims, that right there kills most legacy extensions. As a technology issue, it seems like that had to happen. If it was done years ago, Firefox might be in a better place now.

The second issue is future UI adjustment. Mozilla claims they will be willing to work with extension writers to extend the webextensions api and add some new features. However, features that affect the UI are usually not making the cut. Instead of being a technical issue, I believe this is a political one, I don't see why UI changes couldn't with security and performance be added to the API. IMO somebody decided they didn't want those type of extensions. But this is in a way good news - if it isn't technical and is just a choice, they can change their mind, which they may if they feel they are getting a bad response to the new browser. They'll just have to hope that some extension developers will come back and work with them, since all of these events must have been discouraging to extension writers.

Why am I getting a deja vu moment reading this? (I remember reading this very thing on Ghacks (???) ... good points though