Pale Moon forum

moonbat wrote: ↑

2020-07-17, 06:54

Lurker_01 wrote: ↑

2020-07-17, 06:15

unless this was done with the intent for somebody to actually step up, but i don't think this change will push people to do that.

Had Firefox not dumped XUL, these very extensions would still be only compatible with version 57 and require work from their developers to keep up with the latest changes. Can't expect everything to just run on cruise control without maintenance.

I think everybody expects that Firefox extensions would break over time, like i said it is unfortunate that that some firefox extensions are still working without any need of modification without the arbitrary check and were broken this way.
If this would have been a browser functionality change i think there would be less post about this or just posts to find alternatives.
edit: There is no ""install anyway"" that was in AMO or just enabling the extension after it was disabled. (yes in some cases you do indeed need to modify the extension version check for it to work) (in case if that were Firefox)

Lurker_01 wrote: ↑

2020-07-17, 07:04

and were broken this way.

Not really broken, you just get the warning, right?

letmeindude wrote: ↑

2020-07-17, 07:07

Lurker_01 wrote: ↑

2020-07-17, 07:04

and were broken this way.

Not really broken, you just get the warning, right?

It didn't break traditional extensions but every other kind of extension like dictionaries
edit: Through for traditional extensions some people find extra warning very annoying. (the extension indicator existed before to inform you that this is not a Pale Moon targeted extension)

Lurker_01 wrote: ↑

2020-07-17, 07:11

It didn't break traditional extensions but every other kind of extension like dictionaries
edit: Through for traditional extensions some people find extra warning very annoying.

Dictionaries I think were ported over from Firefox before they went Webextensions, so as of now only Tobin is maintaining those. And the extra warning is only on the addons page. Who goes to look at their addons page everyday? it's not as though the browser reminds you every time you start it or with an intrusive periodic alert like Windows update does.

moonbat wrote: ↑

2020-07-17, 07:17

Lurker_01 wrote: ↑

2020-07-17, 07:11

It didn't break traditional extensions but every other kind of extension like dictionaries
edit: Through for traditional extensions some people find extra warning very annoying.

And the extra warning is only on the addons page. Who goes to look at their addons page everyday? it's not as though the browser reminds you every time you start it or with an intrusive periodic alert like Windows update does.

I think this is a matter of preference at least for me, in my case i changed some tab/url bar looks and also some other changes that i didn't like because i could do it. i think if i were to really have a Linux install that i would use, it would take me a lot of time to customize(more than just gui preferences) or i will just give up.
There were some points mentioned why these warning were annoying between all the threads created and i wont list them here.
edit:The extension status was shown already so i think some people found the change to warnings too intrusive.
Anyway i am very thankful that there is no intrusive reminders implemented.

athenian200 wrote: ↑

2020-07-17, 02:26

I guess my perspective is, and always has been, that everyone is selfish. I'm selfish because I want this project to thrive for my own reasons, and if I want other people to help me make that happen, I need appeal to their selfishness too.

Something that might help people's perspective is something that was mentioned in passing in another post - that as the version number increases, it will start to conflict with version numbers of firefox releases since aurora.

It so happens whilst I was porting tab groups manager, I found some checks for firefox >3.5. So I added an issue for myself to get rid of all the firefox version checks. And scanning the code today I found a check for firefox version > 28. So next major release, that's going to trigger some code that isn't currently running.

I'm sure tab groups manager isn't the only extension that does that sort of thing - where things have been done for backward compatibility but never removed again.

thosrtanner wrote: ↑

2020-07-17, 12:49

And scanning the code today I found a check for firefox version > 28.

Since your extension targets both Pale Moon and Basilisk, if the extension really does something special for Australis then you could either bump the version to 52 (slight hack still) or better yet, just check for Basilisk's (aka Firefox's) GUID and do the Australis code if it sees Basilisk is running.

thosrtanner wrote: ↑

2020-07-17, 12:49

And scanning the code today I found a check for firefox version > 28. So next major release, that's going to trigger some code that isn't currently running.

How do you know it is some Australis interface code and not something else, because Pale Moon core is based on Firefox v52?

letmeindude wrote: ↑

2020-07-17, 13:42

How do you know it is some Australis interface code and not something else, because Pale Moon core is based on Firefox v52?

Because extensions have nothing to do with the core of the browser, but Australis has a different way to interact with the UI so checking for the browser version means checking for Australis.

vannilla wrote: ↑

2020-07-17, 14:40

checking for the browser version means checking for Australis.

Thank you, good to know.

Lootyhoof wrote: ↑

2020-07-17, 12:59

thosrtanner wrote: ↑

2020-07-17, 12:49

And scanning the code today I found a check for firefox version > 28.

Since your extension targets both Pale Moon and Basilisk, if the extension really does something special for Australis then you could either bump the version to 52 (slight hack still) or better yet, just check for Basilisk's (aka Firefox's) GUID and do the Australis code if it sees Basilisk is running.

I'm pretty sure the code is making an assumption which is just plain wrong, but I need to play with it a bit.

New Tobin Paradigm wrote: ↑

2020-02-13, 15:33

This is the same reason that Basilisk and Interlink retain their respective 52.9 version numbers even if it isn't widely displayed.

Why can't Pale Moon retain its version number 28? We can call it Pale Moon 29, but retain that magical 28.

Reasonable path forward. I was fond of the lowkey color symbols signifying target application prior to 28.11, but I can imagine people overlooking them and blatantly installing 5 year old Ff extensions and then being stumped when they misbehave with their 2 week old UXP browser. I'd say the shade of yellow conveys the message more clearly

Now what I'm wondering about is the next major version, presumably 29.x. Can one assume the dual-GUID system will be dropped at that milestone? If a legacy Firefox extension by grace of shared ancestry works apart from the targeted application, then the only work necessary for a potential forker in the short term is to change the GUID to target Pale Moon?

Of the 17 extensions I use, only 2 were acquired from Classic Add-ons Archive. Really stoked by the amount of Pale Moon-specific add-ons accumulated over the years. Kudos to all those involved.

docR11 wrote: ↑

2020-08-26, 20:39

then the only work necessary for a potential forker in the short term is to change the GUID to target Pale Moon?

Pretty much. You can also try contacting the original author to ask if they can fork it for Pale Moon, I've had some success doing that.

moonbat wrote: ↑

2020-08-28, 00:52

They follow a standard pattern now - Remove feature -> demote it to about:config -> demote it to ESR -> get rid of it altogether. And it's ridiculous the amount changes you have to make to a fresh install of a supposedly private browser to stop it snooping on you. The fact that there are so many 'hardened' prefs.js projects on Github speaks for itself.[/offtopic]

moonbat wrote: ↑

2020-08-28, 00:52

Pretty much. You can also try contacting the original author to ask if they can fork it for Pale Moon, I've had some success doing that.

Off-topic:

I've actually considered that for a small handful of lesser-essential extensions I still keep around. They've shown no signs of breakage over time with newer PM builds, and I don't foresee that happening with 29.x either. If I had to I could just edit the GUID for my own personal use, or share them on AMO if I get proper authorization.

Question about the unstable development builds of Pale Moon as of this post (v29.0.0a6).


I humored myself by downloading portable v29.0.0a6 and copying in my user profile from Pale Moon v28.12 0 to see if any of the traditional Firefox-targeting XUL extensions that I use would be disabled or the like, and it doesn't look like it does so despite v29 currently using a label in the Add-ons Manager saying "[extension name] is incompatible with Pale Moon 29.0.0a6".

So I'm kind of confused because the labeling suggests to me that the extension should straight-up either not work or get automatically disabled, and yet it still seemed to function and remain enabled. (I can understand that some extensions may very well bork out, but the two relatively simple ones I tested did not bork out).

Are unstable dev builds by chance not set to disable traditional Firefox-targeting XUL extensions, and that the final version stable v29 will in fact disable them in the way that traditional Firefox-targeted XUL themes are already disabled in current stable Pale Moon? Or is the wording of the label meant to be nothing more than a more blatant "you're on your own if this extension doesn't work" kind of thing?

They should be outright disabled in the latest unstable. Note that this doesn't necessarily mean they will be immediately whenever v29 is released, but we are working on making this change work as-intended for which the unstable is, of course, the designated testing ground. It doesn't seem to work entirely as-intended yet.

Aren't web technologies, such as JavaScript, supposed to be backwards compatible? To remind, the latest update that broke extensions consisted of removing non-standard functions:

Removed the non-standard object.prototype.watch()/unwatch() functions. Please note that this may affect some extensions; those will need to be updated to no longer use these non-standard functions

A lot of web standards used to be yes. But that entails a commitment to legacy and as you well know.. Legacy is NOT valid anymore in today's web un-standards as presented by the Google Cartel.

It simply isn't possible to conform to the web un-standards and old true standards anymore. Especially when something overlaps. This is totally by design. A design of chaos.

Maybe if JS had a runtime target to specify version of a standard but it doesn't and wouldn't work anyway going forward with these living un-standards.