Slashdot Mirror
Firefox 32 Arrives With New HTTP Cache, Public Key Pinning Support
An anonymous reader writes: Mozilla today officially launched Firefox 32 for Windows, Mac, Linux, and Android. Additions include a new HTTP cache for improved performance, public key pinning support, and easy language switching on Android. The Android version is trickling out slowly on Google Play. Changelogs are here: desktop and mobile.
Firefox, bagel and lox
Breakfast of champions handy
And aftershave that makes men brave
When over Macho Grande
Burma Shave
Get thee glass eyes, and, like a scurvy politician, seem to see things thou dost not.--King Lear
This ran a day or two ago, yesterday, or day before.
Will the next version be Firefox64 ?
When will we go back to Firefox ONE ?
I have discovered a truly marvelous proof of killer sig, which this margin is too narrow to contain.
Just installed the latest Firefox and did a bit of random surfing. First impression: noticeably faster than before, probably even on par with Chrome.
Autoupdated this am.
Seems to work fine.
Memory use seems about the same. (I have 10 tabs open now... lots of "complex/rich" sites... 536 MB)
I don't read your sig. Why are you reading mine?
This stuff looks interesting. I can't wait until they fold it into Palemoon.
Because coolness of the technology aside, everything else about Firefox is increasingly pissing me off.
I have 3 tabs open one with flash and its only 196 megs (and some change), dunno what your fucking problem is
Might as well chime in:
Two tabs, 264mb
Required reading for internet skeptics
I prefer FF over other browsers. But it's mainly because it's what I'm used to and the tree style tabs. I'm sure you can get the same for Chrome. But I haven't gotten around to checking.
FF became pretty unstable a few versions ago, though I don't recall which. It seemed to be a memory leak or something. It got up to around 2.5 GB of RAM and then became unresponsive and would eventually crash. My system has 16 GB of RAM, and was never near 100%. The next release took a little longer to reach this point, and the one after that was even longer. I think the version before 32 only crashed once on me. And 32 is open right now as I post this using 2.1 GB of RAM. Granted, I have 13 FF windows open with 5 to 24 tabs open in each.
I post in EVERY Slashdot firefox article, whining for the same thing.
LESS focus on UI / features, MORE focus on stability / performance.
I've been using FireFox since the name it had before FireFox (I've forgotten it) - I think I used it since version 1 or god knows what.
For about the last 12 months, maybe 18, Firefox has become completely unstable for "extreme" browsers like me. I run anywhere from 30 to 150 tabs open at a time. I'd say a nice average would be around 60 tabs. When I'm researching something (often multiple things) I like to google what I'm reading, middle click open in the background the first 5 results. Then when on a forum, I'll middle click open 5 more results and so on. I like having those tabs queued up in the background for me to read.
You might think "well there's the cause of your stability problem!!" except this never used to happen. 18 months ago you could hit 200 tabs without FF crashing. Now, I'm scared to open more than 60. This is across multiple machines too.
I've even tried switching to WaterFox, no dice - I'm still able to crash FF regularly and I run very few addons either.
It's good to see the http cache changes, so they are working on performance but stability should be the #1 focus.
Oddly enough, I get exactly the same symptoms in Firefox for Android as I do Windows for fucks sake. If I hit enough tabs (about 8 on my Galaxy S3) - FF for Android shits the bed, presumably because it's out of ram and can't page well or something. Worst part is FF for Android doesn't remember my open tabs either. Miserable.
They've fiddled and fucked with the UI, replicating Chrome as much as they can (ugh!) for years, now can they stop? If I wanted ugly goddamn chrome I'd install it.
PLEASE fix the stability, PLEASE make it faster. I don't care how much ram it uses, I just want a modern experience with my browser.
Since these updates became more about upping the version number than adding anything really useful and substantive, they should seize this golden opportunity to call it Firefox 100000. Then as the updates roll on from here ...Firefox 100001, Firefox 100010, Firefox 100011, etc.
Grim Reefer, " I have 13 FF windows open with 5 to 24 tabs open in each." Yes, just try exactly this on Chrome and report back on memory usage.
Firefox mobile: Android 2.2 and ARMv6 processor chipset no longer supported
Youtube Videos still continue to play even when the tab is closed. The video stops rendering sometimes yet the audio continues. The browser itself freezes a lot. I get 'firefox is running but not responding' messages all the time. I don't see the fuss. It's bloated crap IMO
firefox would have crashed long before i was able to open up that many windows with that many tabs.
after a few hundred pageviews using no more than a few tabs at a time firefox gets close to 2 gigs memory used and stays there.. ui gets sluggish (even more than usual now since the omg-it-looks-like-chrome version), pages start stuttering on scrolling and taking longer to load up in the first place, and acknowledgement of mouseclicks can be delayed so much that what firefox 'clicks' on isn't even what *I* clicked the button on.... this on 8gb quad core win7x64.... and even when running only bare minimum addons like abp and noscript.
and yet i stick with firefox because a) its not google, b) it's not microsoft, and c) adblock and noscript are unequaled still in every other browser
The next time it gets that bad, open about:memory and see what stands out as eating the RAM. There have been issues with AntiVirus software, old crufty features like "ask me about every cookie", the YouTubeCenter addon gobbles up RAM (developer version doesn't), and video drivers with shared RAM being problematic. Adobe has also stopped caring about Flash on Firefox, so that's becoming a real turd in the punchbowl lately. But if you try to dig deeper and help Mozilla find out what the problem is, I'm sure they'll help find out what the cause is. They did for me, so if you're not a total asshat about it, they'll probably help you out too. They do care, they just need someone to help them find the actual problem.
Why would I ever need any of this? If anything ever happened, I would just bring up MyCleanPC and fix the problem.
I'll trust a respected companies solution over some guy-in-a-basement's free hosts program who is probably just out to scam me. At least MyCleanPC lets me hold someone responsible if I have a problem! What guarantee do I have that some "Anonymous Coward" will respond if I need support?
The "new" DownloadStatusBar addon which worked around the changes in (I think) FF29... well, it's broken again. Probably I can go Googling and find an update somewhere.
"Oh no... he found the
Yeah. The cookie jar was empty, so he had to take his nap with an empty tummy.
If you think I voted for Trump because of this post, you're wrong. I voted for Dr. Jill Stein of the Green Party. Again.
Two tabs. TWO.
http://imgur.com/8oaZW6D
Zero excuse for that bullshit. Slashdot and Fark are NOT loading up nearly half a gig of information.
Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
http://imgur.com/8oaZW6D
Only installed plugin is AdBlock Edge.
Two pages. Almost half a gig of RAM used.
Absolute rubbish.
Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
http://imgur.com/8oaZW6D
Two tabs, nearly double that, with my only extension being AdBlock Edge.
I find it funny everyone makes claims but aren't backing it up with screenshots.
Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
Just because something is using memory (it's there to be used, after all), it does not automatically follow that it is leaking memory, either in your sense (poor memory management) or the sense developers use it (broken memory management).
I've got mysqld running at home. The only DB on it is a few megabytes of data, and yet mysqld uses 1.5Gb. Why? Because it's available. It doesn't harm the system, as far as I can tell.
systemd is Roko's Basilisk.
Yup, from secuty point: something is wrong with the pinning CA, drop a fatal error message and you are ready.
But: No option to handle the connection as a unsecured connection. If you simply want to watch the intercepted information about funny cat video, the fact that there is a man in the middle attack is not interresting.
And: Something is alarml wrong error message. Nope, just that the connection might be intercepted, or not be from the original site. Nothing wrong with the PC of the user, but nothing he can do about it. Give him options!
A local DNS server can do more than all the above. A local DNS server can simply not resolve a specific subdomain, even an entire domain with a one liner. This prevents the browser from even trying to connect to an address.
One DNS server is sufficient for the entire network.
With high caching settings in a local DNS server, you can have something more reliable than hosts for all sites.
A local DNS server can get the advantages of hosts files and go beyond by using TCP resolution, which prevents known resolution exploits without the use of technologies like DNSSEC.
Hosts files can do 1, 2, 3, 4. After all, you can setup local zone files as you wish.
Of course, Windows was designed in mind of large hosts file, so DNS cache has to be disabled for this to work. Not a very good workaround. Setup a local DNS server and get the benefits of running the DNS server in a low-level user account that can't touch the rest of the system and not mess with Windows internals or break internals in the process.
Additionally, you don't need the terabytes of harddrive space needed to block entire domains by generating every single subdomain possibility that you would under hosts files if you wanted to block entire malicious domains.
Instead work with a user-mode process that is running with low-level user rights that cannot touch the rest of the system.
P.S. Slashdot's lameness filter really does not like me quoting APK posts.
Change is certain; progress is not obligatory.
Why does he have to be a 'Mozillian' because it worked for him? Couldn't he just be a user that doesn't experience the issue?
Khyber also claimed to me once that he ran an IPv6 only IRC network where there were lots of furry musicians on a domain that was unreigstered that I couldn't access because of some gibberish.
But it isn't, it shows reproducibility.
Except maybe he isn't emotionally invested in Firefox and people who have problems like this maybe just trolling or having some other issue that's causing this.
What if he doesn't care because he's not that emotionally invested in what browser he's using?
Good for them? Sounds like these people are getting what they wanted. There is no need to stop them.
Change is certain; progress is not obligatory.
Indeed, I increase the packet overhead for the purposes of security. I'd rather be more secure than not for such an easy change on all sites, as opposed to only the websites I have setup manually in DNS.
Using the scientific method, could you show an electric meter making a notable difference when running a DNS server verses hosts when doing look ups? I think YouTube would host a video of such a thing for free for you.
Change is certain; progress is not obligatory.
I run anywhere from 30 to 150 tabs open at a time. I'd say a nice average would be around 60 tabs.
I really cannot think of a reasonable workflow where that would make sense but I'm not trying to judge. As long as you are aware that you are doing something that almost nobody else does or even thinks is a good idea then go for it. Could be useful as a stress test I guess. I've been using Firefox (and Mozilla and Netscape before that) for a long long time and I've never seen behavior like what you describe but then I never thought it was a good idea to have 150 tabs open at once either.
My question would be, what have you done to isolate the problem? Are you sure it has anything to do with the number of browser tabs or is that merely coincidental? Are specific websites causing the problem? If something has changed is that something in Firefox or is it something in how the data is being served? You seem to be blaming Firefox but it isn't clear from your description that Firefox is the unquestionable source of the problem. It certainly should be on the list of possibilities but it isn't certain.
You're right. There's zero excuse for you blaming your problems on Firefox. Since you complained about other people not posting images here you go:
7 tabs open.
1 tab running a video
16 plugins installed
7 extensions running (10 installed).
350MB of RAM used. Go fix your browser instead of bitching about it on Slashdot.
http://s28.postimg.org/3zhxwhuzx/firefox.jpg
You can do it with one file in DNS, but you can't do it with hosts.
You can't prevent resolution of IPs entirely; the browser (or other application) still tries to connect regardless of the IP provided to it. You can't do TCP based resolution preventing all spoofing exploits from succeeding and you and you can't blacklist entire domains (that includes all subdomains) with a single line either.
I can do this in DNS just fine, for my entire network trivially.
I would rather keep things outisde of kernel space and inside user mode, where it can be contained. It also avoids issues where applications don't use resolv().
Breaking DNS caching on Windows doesn't seem like a good idea.
I don't need a separate system on my network (my Linux router is good enough). You can also set it up locally on your computer if you prefer; choose the deployment best for your network.
OpenDNS sucks where dynamic IPs are concerned, DNSSEC deployment is not widespread enough to prevent DNS spoofing such as the Kaminsky exploit you pointed out, so TCP resolution is still preferred.
P.S. WTF is with the lameness filter? It really doesn't like me quoting you.
Change is certain; progress is not obligatory.
IT DOESN'T FUCKING MATTER IF YOU AREN'T EXPERIENCING THESE PROBLEMS! Why do you freaks always insist on denying that very real problems exist just because you haven't personally experienced them?
Yes it does matter if others aren't experiencing these problems because you can't fix a problem you can't reproduce. The problem is with people who complain that "Firefox is broken" but then never provide adequate data to isolate what they are doing that is causing the problem. If User A says Firefox is slow when doing Behavior X and User B tries Behavior X as described and cannot replicate the problem then the source of the problem has to be something more than solely Behavior X or it might have nothing at all to do with Behavior X because User B is mistaken. It could be how the software interacts with other software. It could be a hardware issue. It could be a version mismatch. It could be any number of things, particularly when you are talking about some weird corner use case like opening 200 tabs at once. (why anyone would do that eludes me)
These aren't isolated problems that people are reporting, either. The reports are numerous, widespread, yet still quite consistent. Firefox is slow. Firefox suffers from memory leaks. Firefox has a shitty UI these days.
On a daily basis I run Firefox and Chrome side by side at work (yes I have a good reason) on Windows PCs and sometimes Macs. I run IE and Safari with some regularity too. I see no evidence that Firefox is meaningfully faster or slower than any of the other major browsers. There probably are some differences but they are so minor as to be meaningless to most people in most cases. I have seen memory leaks in the past but I see no current evidence of them with the current stable builds. While they may exist I'm certainly not seeing much evidence of them. Provide a reproducible use case though and I'll of course reconsider. As for the "shitty UI" that is more a matter of opinion than fact. I don't think Firefox's interface is meaningfully better or worse than Chrome, Safari or IE. I wish they would stop needlessly screwing with it but that's a separate issue.
ALL OF THOSE COMPLAINTS ARE LEGITIMATE, AND ALL ARE FACTUAL!
Right, because who would ever lie on the internet? It's such a bastion of integrity and fair play.
If Khyber spent even a fraction of the time he spends whining on Slashdot actually reporting the bugs to Mozilla and honestly helping them diagnose his problems, then they would likely have been solved by now. I say this as someone who doesn't even like Firefox, yet has had several of his "show stopper" bugs fixed because he had the crazy notion of helping Mozilla fix their "busted junk". It's easy to sit back and tell Mozilla to fix things, but if all you do is present anecdotal, whiny, non-actionable bug reports and vitriol on sites other than Mozilla's bug tracker, then don't be surprised if your problems never get solved.
Also, stay classy with your BS "10%" statistics and "ass slow" commentary. Chrome doesn't need such flimsy and childish arguments made for it. Lately more people I know have been jumping back to Firefox than the other way around, and the only reason the "10%" statistic holds water is because Chrome is now bundled as the default browser on newer Android devices. Compare only on desktops with a big enough subset of the web's userbase and Firefox hasn't really lost much ground at all. It has always been kneecapped by the fact that it's not bundled with any OSes or devices, much moreso these days since there are no mobiles or tablets bundled with it that have any uptake.
I don't care about usability of DNS. If I really did, I would make a GUI editor with rainbows.
Then please upload a video or some other material of the differences being measured. I really do not believe it makes a notable difference on modern systems.
I have a Linux router that acts as DHCP, PPPoE, DNS, NAT and passive tunneling for getting around websites that do geoip blocking. It's fine for me. Sure, if I split up each service on a separate machine, it would be wasteful for a small network, I agree. But who does that?
I don't see the problem? I doubled it, so what?
Change is certain; progress is not obligatory.
Zero extensions. If I had extension, I'd have listed them. Are you that incapable of rational thought?
Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
Grim Reefer, " I have 13 FF windows open with 5 to 24 tabs open in each." Yes, just try exactly this on Chrome and report back on memory usage.
I have no desire to install Chrome. Partly because I get sick of other programs trying to sneak it onto my computer. But I'm curious about what you are inferring. Will Chrome perform better than FF? Or worse? TIA.
If they are trying to keep up with Chrome, they better start adding those new HTML 5.1 features too.
Get free satoshi (Bitcoin) and Dogecoins
Yes.
Also, a DNS server can also be hardcoded to protect you. It can also be setup to have a longer caching period to provide resolutions while the DNS server is inaccessible.
That's a great theory, can you actually provide results that shows this has a notable affect?
I'm not interested in your 'moving goal posts' arguments. I also don't care about usability arguments at this point.
Arguments of efficiency are interesting, especially where DNSSEC is concerned. Servers that are up to date using DNSSEC, when receiving a request and they can't handle packet sizes beyond 512 (as is default on ISP networks for UDP), a response is given to use TCP to do the resolution. In other words, you're increasing the load and decreasing efficiency for people who don't run DNSSEC and decreasing load while increasing efficiency for people who do. This seems like a fair trade.
How much of a difference that load makes is another story however.
Change is certain; progress is not obligatory.
6 tabs open for the past 5 hours, 450MB.
Nice try, I didn't actually know about about:memory until you posted this, nor will I ever care enough about memory (I have 32GB in this machine) to ever type that in my browser again. Also I don't actually give a crap about Firefox because I use Chrome, but your anti-Firefox rant just needed correcting.
Interesting side note is that Chrome will show Firefox and IE memory usage in it's about:memory page. I decided to put that to the test. Both browsers have 8 identical tabs open, 7 slashdot tabs, and roundcube logged in, in the last tab.
Chrome 1035MB
Firefox 310MB
IE 11 261MB
I'm not entirely sure why Chrome shows other browsers. It would look like the marketing really does not work in its favour. Note though that Chrome has a lot of extensions installed, and IE is pretty much vanilla and appears to render the pages a lot worse than the other browsers.
I am saying your memory woes will be incredibly worse. Your use scenario of 13 FF windows multiplied with that many tabs is clearly well out of the bell curve.
Good to know. I hear a lot of people claiming that Chrome is faster and more efficient. I've often wondered about that. But like I said. It's pretty damn annoying how many programs try to install Chrome.
If it's such a trivial fact. Please show me the power differences seen. A video showing wattage while running a DNS server verses your hosts file and doing DNS lookups will be sufficient.
That's really only an issue if the old DNS server returns bad records now as opposed to not responding at all.
I can do this on my DNS server too.
Irrelevant.
Change is certain; progress is not obligatory.
I'm not aware of a notable difference, care to share your results? I would recommend a video recording of the system showing the wattage when using DNS server verses a large hosts file.
is this tiny amount of usage really an issue on modern computers? Could you show your evidence on the notable differences? Additionally, on Linux distributions, often the system resolver (ie: powerdns) is already a full DNS server, it's just configured to do auto configuration and and forward dns resolution and caching by default. In such a scenario, I am skeptical that the difference in system resources can even be detected.
So can hosts files. An IP that once pointed to the correct server may no longer be the same server.
I'd rather use dig, because then I can see all the IP addresses for anything using a round robin DNS.
There is nothing for me to prove, you're the one making the claims it uses more electricity, you prove it.
Change is certain; progress is not obligatory.
You really don't like backing your claims, do you? You've been making this claim about additional electrical usage for a while now and whenever pressed, you fail to provide it. I can only assume that you don't really know and you're just making this up as you go along.
If you read my post, I never admitted there was additional electrical usage. Sure, you can measure a DNS server is taking up resources, but is it impacting anything? I'm unconvinced there is a problem like you portray it to be.
I can too with DNS, I'm a Linux user.
I'm not. I'm telling you that some Linux distributions (particularly desktop ones) come with a DNS server (powerdns) configured out of the box. I was also pointing out in such a scenario, you're going to be hard pressed to even say there is even a recognizable difference in resource usage (after all, we can tell the difference between 200KiB use of memory verses 0, but 200KiB verses 200KiB ? not really)
I am not convinced there is a notable difference in electrical power supply though and I don't really see how DNS would even be notable for system performance either.
Seriously, what's the difference APK? Are we expected to use 400watts more for running a DNS server? Why are you constantly avoiding answering this? Why won't you show actual proof that it will increase the cost of electricity?
Change is certain; progress is not obligatory.
" your anti-Firefox rant just needed correcting."
Sorry, But unless you're some total n00b, Firefox has always had this problem. It's never been properly fixed since version 3.
Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
This is irrelevant to answering if this makes a difference to the electricity costs!
If the differences make little difference to the power consumption (something that requires little resources may not require the system to draw any further power at all than what is already available).
Now, since you're so adamant this will increase the cost of electricity, show us. Clearly if you know this to be true, you must have observed this and you're not just merely guessing.
How many more watts does it take to run a DNS server? What DNS servers did you run? What operating systems? What configurations? Come on APK, let's see the evidence to your claim!
Change is certain; progress is not obligatory.
It is not the answer and I provided reasons why.
Then prove it! How many more watts does it take to run a DNS server? It's a simple question, if you're correct, this can be measured and actually make some notable impact. It should be trivial for you to prove, but we both know why you aren't. After all, according to you, it impacts the electricity costs. So, it has to make enough of a difference to affect common units of usage.
You have failed to provide yet again, actual evidence where we see more power being consumed to just run a DNS server.
Irrelevant.
And you conceded hosts files can break windows services and I've pointed out that DNS can do all of this too without downing a windows service.
Change is certain; progress is not obligatory.
Well the total n00b thing is you ignoring the fixes that have actually happened. There was an entire campaign centered around reducing the memory footprint of Firefox and closing out all open memory leak bugs. That campaign started at Firefox 3.5 and the memory footprint has been improving very steadily since making it now quite light weight on par with a browser heavily integrated into the system and far better than the competition.
But don't let my facts (or those of any of the other people posting and modding against you) get in the way of your religious tirade.
So, yea. That screenshot I posted is of the latest version of Firefox. So something is NOT fixed.
So what actually got fixed, hrm? RAM usage was most fucking certainly not one of them.
Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
So, yea. That screenshot I posted is of the latest version of Firefox. So something is NOT fixed.
So what actually got fixed, hrm? RAM usage was most fucking certainly not one of them.
That screenshot that disagrees with all the others in your thread? I take it your problem is reproducible and that you have filled a bug report complete with all the info required to reliably reproduce your problem? If not then kindly screw yourself, if so then wait for the outcome of the bug report.
But coming on a forum and saying everything is broken and nothing has been fixed since Firefox 3 is absurdly wrong and childish, not to mention disagrees with what others are seeing.
Or maybe do you think RAM usage is subject to a single bug which can magically fix everything, in which case I have a pet rock which also cures cancer I can sell you.
http://i.imgur.com/DyizJh0.png
Still leaking like a fucking sieve even after upgrading (and again, the only thing installed is AdBlock Edge.)
Two images, two slashdot threads, and an HTML-based e-mail.
Over 650 fucking megs.
Firefox is fucking broken.
Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.
Still fucking rising, just sitting waiting on the pizza I phoned in about 25 minutes ago. Same tabs, not touched shit and instead went on Camfrog to relax.
http://i.imgur.com/OxQtmyj.png
Still waiting on Serviscope_minor to wake up to fucking reality and realize that Jessica Price isn't going to fuck him.