Changing The Browser

Platform Development
#1

The browser update finally launched and it’s brought Sailfish up to a version of Gecko that’s “only” been out of support for over 2 years. What’s the justification for staying with Gecko when Jolla clearly (and unfortunately) can’t keep up with updates? Not that Qt is always on the latest version, but at least qtweb(kit,engine) will get incidental updates whenever a new Qt is pushed out, and that’s probably less active development work than having to create all of their patches and upgrades for their bespoke, unsupported Gecko embedding framework.

Not only does having an out of date browser hurt the ability to use web apps but it creates a massive security hole in the system unless Jolla is basically forking each version of Gecko and providing security patches themselves. It seems far safer and like less of a maintenance burden to just rewrite the Browser app against one of the Qt-provided web engines at this point or just rewrite the browser entirely if that’s too much work.

2 Likes
#2

If you missed it; there are significant license hurdles to doing a Qt uplift.
Also, if nobody uses Firefox, well… nobody uses Firefox, nobody will care about Firefox and the web becomes a Google-controlled monoculture.

12 Likes
#3

I actually did miss it. What’s the blocker on Qt? My understanding was that through the agreement with the KDE Foundation it should always be available; everything relevant to Jolla seems to be available under the LGPL.

#4

Current licensing is either full GPL or commercial.
Full GPL means that all Jolla apps need to be GPL too, which is a business decision to be made by Jolla.
Commercial means spending a few 100.000 Euro’s a year for access to Qt, while the financial situation for Jolla is tight, to say the least :slight_smile:

What I think is happening is that it gets postponed for a while to save money, untill it can no longer be postponed again.

Both the decision for Webkit and Gecko have pro’s and cons. Personally I think the mission of Mozilla is more in line with that of Jolla.

4 Likes
#5

There are probably better sources, but i’m clearly bad at looking… but at least it is recent:
https://irclogs.sailfishos.org/meetings/sailfishos-meeting/2020/sailfishos-meeting.2020-06-11-08.00.log.html

I must admit i’m curious as to why they didn’t choose to adapt the GeckoView bindings for Android to Sailfish OS, but i am confident that it was given a fair evaluation.

2 Likes
#6

And how Jolla could change that with their decisions in the first place?
Just because they would switch to a different engine (what everybody did in the open source world lol - no one uses Gecko anymore besides Firefox itself) how that would actually hurt the position of Firefox? The user’s aren’t even faced with the fact that the Sailfish browser is Firefox based.
Also Qt Web Engine is fully open source, so they can fork Blink at any time they feel that Google doing a lot of bad things, not to mention that the entire WebKit/Blink tree is based on kHTML which is still a thing for a plan C or something.

#7

They won’t, but that’s the same ridiculous argument as small western countries not doing anything for the climate because China and India exists. (If nobody does… nobody does.)
I still like and respect this decision very much all the same.

I don’t believe that was ever different - what are you on about?

But having a free ride doesn’t exactly make them maintain a position where they could.
Ever heard of “boiling the frog”?

My concern is not primarily about google being nasty, but how standards will go out the window as soon as there is just a single implementation. (And what will happen after that…)

3 Likes
#8

I understand your concerns I have two thoughts on this:

  • The main problem is Chrome itself not the engine. Firefox is better because of the excellent privacy tools they developed not because they use an aging engine. They could provide the very same thing with a different engine too.
  • In the open source world everything can be forked at any time and I’m pretty sure companies and communities could switch from Blink very easily if Google would make things unbearable. In an any-company vs the entire internet scenario always the specific company is the one who will loose that wants to change the internet unilaterally, no matter how big the company is.
#9

Yes, but also no.
The problem isn’t Chrome today it is Chrome in 5 years.
So i still maintain that the “problem” is in fact the engine, because of what a position it implicitly puts Chrome in when the web devolves into a monoculture.

The Internet was already pulled back from the brink once, i’m not sure it could be pulled off again.

I disagree, and i actually don’t care very much about them.
They only look great in comparison to the web landscape and competition, really it’s just common sense.
Same as i always smirk a bit at people touting Sailfish’s privacy focus… it used to be called “common decency” and “owning your device”.

To be maintained by who? A stalled fork hardly gets any traction or taken seriously in standardization.
Apple ($2.4 Trillion, with a T) accidentally tried when Blink forked… feel free to check up on how that went.

Again… boiling the frog. Had Chrome been what it is now when it was introduced, it would have been laughed off the Internet, and you seem to agree?

1 Like
#10

For what it’s worth this is incorrect and 100% Mozilla’s doing. Long ago it used to be possible to bind Gecko to toolkits and use it in 3rd party browsers, and back then quite a few browsers used it. Mozilla started neglecting and then completely removed the ability to use Gecko in non-XUL applications (until they removed XUL ofc) and the browsers dried up.

4 Likes
#11

In addition, if I remember correctly, to switch to the commercial Qt license, you need to pay for all the years you used it with the GPL license retroactively, which… might be a bit much after 10 years.