4 ms·

I've been trying out all the firefox forks that come out and end up back on master Firefox (on desktop, Mull wins on Android IMO) due to the inevitable limits o

I've been trying out all the firefox forks that come out and end up back on master Firefox (on desktop, Mull wins on Android IMO) due to the inevitable limits of developer time on the project. The value proposition is never quite worth it for the downsides. A web browser is not a trivial project to build and upstream changes are unrelenting. Just like I've seen in Linux for 20+ years, we see different tribes basically working on the same thing in slightly different ways, and I find my self thinking over and over "Why are all these people building different sandcastles when they could be working together?". Particularly with the Firefox forks where projects seem to struggle with the release cadence. It's a rhetorical question though - effective organization appears hard to scale especially when there is no money involved. How many cash rich companies have you worked at that have lost the plot? I would really like to see Firefox win again in the browser battles. I personally believe that its the UX where differentiation from the chromiums is opportune with many interesting and desirable new UI paradigms emerging in some the recent third-party browser upstarts.

ranger_danger 3y ago
> Why are all these people building different sandcastles when they could be working together? Why are chefs baking bread? There's buildings to construct.
crabbone 3y ago
Why forks? -- There are a lot of decisions made by the upstream that are... let's put it mildly, "opinionated" (but really, just idiotic). But you cannot argue with majority or whoever leads the project. So, many times you don't do it as a way to kill time. You do it because none of what you have works. If I had time and enough knowledge of any of the existing browsers codebases, I'd definitely try to create my own version. Here are things that are desperately needed and some even have been removed from browsers for no reason that I would like to have back. 1. Keyboard-driven interface. 2. Decent access to the browser's internal state for testing purposes. When it comes to (1), there used to be a common pattern where pressing Alt would underscore letters in labels necessary to press in order to select that element. This is long gone now (and, frankly, wasn't that great, but it was something...) To clarify: it's not gone because the feature was removed, it's gone because a lot of controls simply don't have any labels, which breaks the system. Today, most browser elements have no way of activating them with the keyboard. Just a big middle finger from browser vendors. When it comes to (2), the only thing we have is Selenium (and sometimes some extra stuff on top of that). But, Selenium has no access to anything that happens in the networking side of things (that's in the Web browser, which is all about networking... right?) So you cannot know if the page is loading anything, what it's loading, what's the status of things being loaded. It's ridiculous that people write tests with this handicap. With Selenium you also don't have access to code execution. So, you cannot see if the function was called, a variable was set etc. It's ridiculous how poor your access to the state of the program running in the browser is. You used to be able to side-load some JavaScript into the browser and get access to browser's chrome, but that was removed because it was a "security concern" (but really, power-user features being removed is just a plan for cutting down functionality that affects the least users, has nothing to do with security). Do I expect Firefox or any other browser to implement any of that any time soon or ever? -- Well, let's limit that to my lifetime, and the answer is a resounding "no". Hence forking.
- angra_mainyu 3y ago
  Have a look at NYXT
- replete 3y ago
  Although this wasn't my point – that forks are pointless –, I agree.
Hendrikto 3y ago
> Why are all these people building different sandcastles when they could be working together? Could they, though? What if upstream says ”thanks, but no thanks“ to your ideas? What if there are multiple, equally valid solutions to a problem, and you want to explore one that wasn‘t chosen? Progress isn‘t linear, and not everything can be compromised on. Just having more people work on a project, does not magically speed up development. You need leadership, a common goal, consensus, and focus. If all these projects merged their efforts, that would probably result in a lot of lengthy discussions and infighting, rather than improvements.
- replete 3y ago
  "It's a rhetorical question though - effective organization appears hard to scale especially when there is no money involved." I agree with you, growing beyond a single small team is another game entirely

ranger_danger 3y ago

> Why are all these people building different sandcastles when they could be working together? Why are chefs baking bread? There's buildings to construct.

crabbone 3y ago

Why forks? -- There are a lot of decisions made by the upstream that are... let's put it mildly, "opinionated" (but really, just idiotic). But you cannot argue with majority or whoever leads the project. So, many times you don't do it as a way to kill time. You do it because none of what you have works. If I had time and enough knowledge of any of the existing browsers codebases, I'd definitely try to create my own version. Here are things that are desperately needed and some even have been removed from browsers for no reason that I would like to have back. 1. Keyboard-driven interface. 2. Decent access to the browser's internal state for testing purposes. When it comes to (1), there used to be a common pattern where pressing Alt would underscore letters in labels necessary to press in order to select that element. This is long gone now (and, frankly, wasn't that great, but it was something...) To clarify: it's not gone because the feature was removed, it's gone because a lot of controls simply don't have any labels, which breaks the system. Today, most browser elements have no way of activating them with the keyboard. Just a big middle finger from browser vendors. When it comes to (2), the only thing we have is Selenium (and sometimes some extra stuff on top of that). But, Selenium has no access to anything that happens in the networking side of things (that's in the Web browser, which is all about networking... right?) So you cannot know if the page is loading anything, what it's loading, what's the status of things being loaded. It's ridiculous that people write tests with this handicap. With Selenium you also don't have access to code execution. So, you cannot see if the function was called, a variable was set etc. It's ridiculous how poor your access to the state of the program running in the browser is. You used to be able to side-load some JavaScript into the browser and get access to browser's chrome, but that was removed because it was a "security concern" (but really, power-user features being removed is just a plan for cutting down functionality that affects the least users, has nothing to do with security). Do I expect Firefox or any other browser to implement any of that any time soon or ever? -- Well, let's limit that to my lifetime, and the answer is a resounding "no". Hence forking.

angra_mainyu 3y ago
Have a look at NYXT
replete 3y ago
Although this wasn't my point – that forks are pointless –, I agree.

Hendrikto 3y ago

> Why are all these people building different sandcastles when they could be working together? Could they, though? What if upstream says ”thanks, but no thanks“ to your ideas? What if there are multiple, equally valid solutions to a problem, and you want to explore one that wasn‘t chosen? Progress isn‘t linear, and not everything can be compromised on. Just having more people work on a project, does not magically speed up development. You need leadership, a common goal, consensus, and focus. If all these projects merged their efforts, that would probably result in a lot of lengthy discussions and infighting, rather than improvements.

replete 3y ago
"It's a rhetorical question though - effective organization appears hard to scale especially when there is no money involved." I agree with you, growing beyond a single small team is another game entirely