Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Firefox Test Pilot

testpilot.firefox.com

Post reply on HN
Print view View original

91–100 of 152 posts

Re: Firefox Test Pilot

#91

Post by burfog »

chuckharmston wrote: post #10
kibwen wrote: post #4

Is the "Tab Center" feature a prototype for an official version of tree-style tabs, perhaps? I've seen that Servo/browser.html was experimenting with tabs-on-the-side, I'd love to see this graduate to more than just an experiment.

Potentially! The simple story is that the current tabbing model in was designed to save you from needing half a dozen browser window open. It just doesn't scale well past a dozen tabs or so, and we know that some users have dozens or even hundreds open at the same time. Tab Center is taking a fresh look at the problem with that in mind.

I do hundreds of tabs. The UI is fine, but performance is not at all OK.

The tab UI works because it is part of a larger 3-level system. There are tabs, windows, and virtual desktops. With 10 at each level, you can handle 1000 tabs... except for the performance.

It's important to avoid running things on pages that aren't in focus or even in view. It's important to avoid walking data structures that scatter nodes all over the address space, causing swap access and cache misses. Watch your RSS. Keep those extra tabs idle. Make sure the "Esc" key and the stop button actually work, stopping everything (all tabs, all video, all animation, all audio, etc.) until the user explicitly asks for something to run again.

Re: Firefox Test Pilot

#92

Post by kibwen »

Aldo_MX wrote: post #45
kibwen wrote: post #37

Earlier quoted context omitted.

Having a ten year-old bug isn't a mark of shame, if anything it's laudable when a project doesn't take the easy way out by silently closing ancient bugs like most projects do. Bugs should be fixed according to importance and priority, not simply according to age.

The problem is that "importance and priority" are pretty much subjective terms, for example, for many people Hello or Pocket were everything but a priority.

The solution is that it's an open source project. If you disagree with the priorities of the other developers, then you are empowered to take the initiative.

Re: Firefox Test Pilot

#93

Post by amjd »

greglindahl wrote: post #9

Very excited to see this launch! The Internet Archive has been working with browsers to check (opt-in) the Wayback Machine whenever endusers click on a dead link. We're in the queue for the next round of Test Pilot tests!

There's an existing add-on that (kind of) does this. It checks Google Cache and Internet Archive, among others.

https://addons.mozilla.org/en-US/firefox/addon/resurrect-pag...

Re: Firefox Test Pilot

#94

Post by billiam »

6a68 wrote: post #35
rhinoceraptor wrote: post #20

Earlier quoted context omitted.

Great! So now in like 2 years, hopefully I can simply use type="time" instead of implementing a half baked JS time picker or time validator. It's frustrating seeing them spend so much time on Firefox OS, Hello, etc. and not on making Firefox a really good, standards compliant browser.

I hear your frustration; people inside the company found some of those changes frustrating, too. Test Pilot is a step in a new direction that's more participatory and inclusive of the community, and injects a lot of data into the decision-making process. As far as web standards go, Gecko's the only open-source browser controlled by a nonprofit. The alternative is a webkit monoculture, a world controlled by Apple and…

Congratulations, guys! Really glad to see this finally shipped. It will ALWAYS be Ideatown to me;)

Re: Firefox Test Pilot

#95

Post by 746F7475 »

Just gave it a 10 seconds and decided this is not for me. Very first thing I enabled (Tabcenter) made Firefox completely unusable to a point where I couldn't even disable the feature or uninstall the plugin. Thankfully these pieces are named nicely in ~/.mozilla/firefox//extensions and thus can be removed manually.

Of course to even to read this advice after fucking up you need to know about firefox's -P startup option.

Re: Firefox Test Pilot

#96

Post by smacktoward »

Very nice! So far I'm liking this quite a bit, especially Tab Center.

On that note, one usability suggestion: having the tab sidebar run all the way to the top of the window means that the years (decades?) of muscle memory I've built up telling me to mouse to the top left of the window to hit the Back button now results in me hitting the New Tab button. It'd be better if the awesomebar area spanned the complete width of the window, and the tab center column started at the same height as the content area.

Re: Firefox Test Pilot

#98

Post by amake »

burfog wrote: post #90
rhelmer wrote: post #85

Earlier quoted context omitted.

Oh, I forgot to mention in my sibling comment - extensions are frequently pinpointed as the cause of (or a contributing factor to) performance problems. Additionally, many of them rely on older features in the Mozilla architecture, which makes it very hard to fix without breaking extensions that people rely on. While there has been and continues to be significant work in this area, the Web Extensions project has been…

No, don't blame extensions. I don't run any. Here is what a modern browser is expected to handle: Open and close tabs and windows, browsing the modern web without extensions. Keep it up for weeks, having 10 to 30 windows, and an average of 10 tabs (some at 40). You have 5 to 12 gigabytes of RAM to work with, and at least a 1920x1200 screen. (maybe a 4k screen) That's it. Expectations for older hardware: run for weeks…

Did you actually read and understand his comment?

> I don't run any.

You don't have to. The existence of extensions creates constraints on how they evolve the browser's internals.

Re: Firefox Test Pilot

#99

Post by mistermann »

burfog wrote: post #89
6a68 wrote: post #64

Earlier quoted context omitted.

Test Pilot isn't happening at the cost of stability and performance improvements. It's a complementary program that helps us ship better features through data-driven iteration. I hear your concerns, though. The platform and desktop teams are doing tons of great work on improving stability and performance--you might want to give Firefox another try sometime. If you do, maybe try out some Test Pilot experiments while y…

They "are doing tons of great work on improving stability and performance", but I've heard this before and somehow we got to where we are today. Do you alternate between improving these things and worsening them? Could you possibly not worsen them, or at least warn us to not upgrade when that happens? Regressions are really not OK. People are trying to use this browser. Well, mostly they were trying to use it. I stuc…

I have 12GB of RAM and probably ~120 tabs open at any given time. This used to be fine, but it seems Firefox is just getting worse over time, the main problem is the frequent 5 to 30 second pauses while it is doing....something.

Re: Firefox Test Pilot

#100

Post by mistermann »

6a68 wrote: post #64
mrmondo wrote: post #62

Honestly all I want from Firefox is better stability and performance - I don't care about almost anything else barring security. Firefox has become so unstable and slow (on OSX at least) that many people I know - myself included have switched to chromium. We don't want to use it but we have to because of all the crashes and massive slowdowns during long running sessions or when many tabs are open. This goes for stabl…

Test Pilot isn't happening at the cost of stability and performance improvements. It's a complementary program that helps us ship better features through data-driven iteration. I hear your concerns, though. The platform and desktop teams are doing tons of great work on improving stability and performance--you might want to give Firefox another try sometime. If you do, maybe try out some Test Pilot experiments while y…

Could you possibly comment on FF's seeming inability to unload tabs from memory? There are numerous add-ons that claim to unload tabs and free up that memory, but none of them work as advertised.

And, when FF crashes and reloads, restoring all the tabs, it seems like the tabs aren't loaded (the page is blank and does a reload when selected), yet if you look in task manager (Windows 7), all the memory seems to have already been consumed.

Should tabs be able to be unloaded from memory, in theory? If so, is it implemented this way? If not, is there a reason?

Post reply on HN