Firefox 52 – does it seem faster to anyone else?

Topic

New Reply

Got the Firefox 52.0 update last night just before shutting down. Started Firefox this morning, did a Restore Previous Session, to load a couple of weather pages I check every day– and the refreshes of those pages seemed much quicker than usual.

I see, from googling a few articles (ZDnet, Mozilla blog) that the Mozilla folks have added some major new function in this release, but now I’m wondering if they worked on improving rendering or page fetch efficiency or something, without publicizing that.

So my question is, is anybody else seeing this speedup?  Or should I look elsewhere than Firefox for an explanation?

For context, this is on my “daily driver” OS, Xubuntu 16.04.2 LTS, which also received a kernel update yesterday.  I’ll see if this machine’s Win7 partition shows a similar boost…

Reply | Quote

Replies

Can’t say I’ve noticed that much performance/ rendering wise but, thereagain, I already have FF tweaked and secured to the hilt AFAIK.

Perhaps the removal of Netscape Plugins API (NPAPI) has sped things up? Were you using plugins (not extensions) prior to the update? Are they still enabled or disabled?

‘Redesigned Responsive Design Mode to include device selection, network throttling, and more’ taken from here may have something to do with it also.

Windows - commercial by definition and now function...

1 user thanked author for this post.

AlanH

Reply | Quote

Hmm, good link– thanks!  Way down at the bottom of the release notes, huh?

Flash 24.0.0.221 is still there, having been granted special dispensation for now, and some “OpenH264 Video Codec provided by Cisco Systems, Inc” plugin “automatically installed by Mozilla” seems to have survived as well. (Not that I’ve ever paid any attention to it before.) I don’t recall any other plugins from beforetimes that have now vanished. I’ll see if there are any in the old Firefox on the Win7 side, before I update it.

Off to reboot into Windows…

1 user thanked author for this post.

Microfix

Reply | Quote

Maybe they fixed the issue with EMET 5.5? Do you use EMET?

http://www.dedoimedo.com/computers/firefox-emet-eaf-slowness-fix.html

 

Reply | Quote

EMET doesn’t appear to be present on the Windows side of this machine, and it wouldn’t have affected the Linux side. Good thought, though– thanks!

Reply | Quote

Well, it looks like it’s not going to be a fair comparison over here in Windows 7. Firefox 51 did indeed have a truckload of plugins (maybe 15 to 20), accumulated during eight years or so of this being a Windows-only machine, before it was “handed down” to me.  And upgrading to 52 jettisoned all but three: the two noted above, plus Google’s Widevine. So, that could easily account for the improved “snappiness” in rendering the weather.gov pages I tested just now on 51, and then 52.

I think I have a Linux box that hasn’t been updated to Firefox 52.0 yet.  I’ll see if it has plugin cruft. Kinda doubt it, though.

Reply | Quote

IMPORTANT NOTE: Please note that the “normal” version of Firefox 52 may not support Windows XP & Windows Vista [AND its server counterparts like Server 2003 & Server 2008 R0]. XP/Vista users MUST use the ESR (extended support release) version of Firefox 52 from here.

Quote from Firefox 52 system requirements page:
https://www.mozilla.org/en-US/firefox/52.0/system-requirements/

Windows XP/Vista/Server 2003 are no longer supported by regular Firefox releases. These users should migrate to ESR 52.

3 users thanked author for this post.

walker, Microfix, AlanH

Reply | Quote

I downloaded the latest ESR update yesterday, which was v. 45.8.

Reply | Quote

There is 52 ESR already. But it is in the optional category for this one and the next release, until it is proved to be reliable.
I use 52 ESR and with the right add-ons installed it introduced multiprocess for ESR, which was introduced earlier in the mainstream versions.
I see only one minor problem which has been reported few mainstream versions earlier in relation to screen “repainting” which sometimes require a manual refresh.
Recommended to install Add-on compatibility reported from Mozilla for identifying badly behaving and out of date add-ons in relation to multiprocess.

1 user thanked author for this post.

walker

Reply | Quote

Or even better, upgrade the Operating System.

Reply | Quote

@AlanH

For what it’s worth, I haven’t used plug-ins for years and only have 3 security extensions within firefox for linux and windows.

I think a lot of plug-ins just won’t work anymore due to dropping of NPAPI

Just sayin..

Windows - commercial by definition and now function...

2 users thanked author for this post.

AlanH, ch100

Reply | Quote

Microfix wrote:

I think a lot of plug-ins just won’t work anymore due to dropping of NPAPI

Yeah, a boatload of barnacles got scraped off the Win7 copy of Firefox on this machine today…

Reply | Quote

Good stuff AlanH, back on a plumbline in firefox.

Windows - commercial by definition and now function...

Reply | Quote

I see quite the opposite and seems to be a known problem dating few versions behind, maybe starting with 47, the screen repainting is delayed and as such perceived as slower. It is only a minor annoyance though.

Reply | Quote

ch100 wrote:

I see quite the opposite and seems to be a known problem dating few versions behind, maybe starting with 47, the screen repainting is delayed and as such perceived as slower. It is only a minor annoyance though.

ch100, have you tried adding this command to about:config to rectify?
1) In the configuration window, right click and choose New -> Integer
2) Name it nglayout.initialpaint.delay set to 0

This removes any paint delay

Windows - commercial by definition and now function...

1 user thanked author for this post.

ch100

Reply | Quote

Thank you, I didn’t know about this configuration.
After I upgraded to ESR 52, I reset my prefs.js file to start clean and did again few configurations through the GUI.
In the documentation for the configuration which you propose at
http://kb.mozillazine.org/Nglayout.initialpaint.delay

says:

“Lower values will make a page initially display more quickly, but will make the page take longer to finish rendering. Higher values will have the opposite effect.”

In fact I may have not explained it correctly, I have the delay in rendering the bottom of some pages, specifically pages on askwoody.com and not in the initial display of the page.
This behaviour is totally random.

Do you know if the default (builtin) value for nglayout.initialpaint.delay has changed in FF52?

Reply | Quote

There is a discussion here.
https://bugzilla.mozilla.org/show_bug.cgi?id=1283302
It may actually have become too short to 5ms or 0ms (like what you propose) instead of the traditional older value of 250ms.

Reply | Quote

@microfix

From reading at different discussions on the Mozilla Developer sites I found that the default value in the absence of the setting nglayout.initialpaint.delay used to be for many (15) years 250 ms due to the internal parameter PAINTLOCK_EVENT_DELAY set at that value.
The value has been revisited recently and the developers decided that starting with Firefox 52, the value has been modified to 5 ms for desktop Firefox while it was left to 250 ms for Android Firefox.

I opted for now to revert to the value 250 which appears to work better for my system. There may be variations in this behaviour from one system to another and/or type of Graphics Card and drivers installed and also the latency and speed of the internet connection.

There is still ongoing research at Mozilla as you would notice from the URL which I posted earlier https://bugzilla.mozilla.org/show_bug.cgi?id=1283302 and there are references to Mozilla 53 which has not yet been released.

Overall your suggestion was very useful and I will keep testing with various values for the next few days and weeks. If I will find anything significant I will post here. Until then, as I said, I opted for the old “safe” value which allows for a delay in the first display of new pages but also allows for background rendering so pages are not painted incomplete at first. This is largely a matter of preference and not a value which can be said as being correct or incorrect. Very much like progressive (“interlaced”) display of jpg images on sites like Google when searching for images. I don’t like that type of image rendering, but it seems to be currently preferred in the modern web page design.

1 user thanked author for this post.

walker

Reply | Quote

This may also be dependant on graphics card types, as there are a few drivers to consider from different manufacturers AMD, Intel etc., which are constantly changing, updating.

All my/our graphics cards are nvidia GeForce so I can’t verify whether my paint.delay tip works for others after looking closer into this. I only use nvidia reference WHQL drivers since before the aquisition of 3DFX.  Never had an issue with delay or tearing on GeForce cards and have been using this tweak form pre 2010 era.

Was just a suggestion and as you state, for sure. it will be addressed by the mozilla team. I’ll keep an eye on this also and revisit when developments arise.

Windows - commercial by definition and now function...

Reply | Quote

Can’t say I see much difference in rendering speed in Firefox 52, since I updated today.  It has always been pretty snappy for me, with my Windows 10 partition running on a SSD drive and a fast broadband connection.  Pretty much on par with Chrome and Edge here.

I was going to hold off on the update, until I saw this article today … jumped right on it!

Security vulnerabilities fixed in Firefox 52

https://www.mozilla.org/en-US/security/advisories/mfsa2017-05/

Windows 10 Pro 22H2

Reply | Quote

Something else you may want to check since Firefox 51, is that you are running Firefox with the new e10s (Electrolysis) multiprocess feature enabled.  It can become disabled due to incompatible extensions or accessibility features.

https://wiki.mozilla.org/E10s

“Electrolysis functionality hosts, renders, or executes web related content in background child processes which communicate with the “parent” Firefox browser via various ipdl protocols. The two major advantages of this model are security and performance. Security improvements are accomplished through security sandboxing, performance improvements are born out of the fact that multiple processes better leverage available client computing power.”

See about:support for status.  If all is well you should find:

“Multiprocess Windows     1/1 (Enabled by default)”

The first issue I encountered was with v51.  I had to disable my extensions one at I time until I sorted out those that were incompatible with e10s.

The second issue was with v52.  I kept getting 0/1 (Disabled by accessibility tools).  The only way I found to correct this was to use the “Refresh Firefox” feature.  It creates a new default profile and imports everything except browser settings and extensions.

Windows 10 Pro 22H2

1 user thanked author for this post.

AlanH

Reply | Quote

Nifty! Yes, it appears to be enabled here.

Reply | Quote

Just to finish up my portion of what has turned out to be a much more interesting topic than what I started with: I suspect the speed improvement was due mostly to the kernel update yesterday.

On the other Xubuntu machine I mentioned earlier, I checked the “snappiness” of Firefox 51, then let only the Firefox 52 update install. No perceptible difference. I then let each group of yesterday’s other updates through, retesting after each. Still pretty much the same. Then the kernel update, reboot, and retest– and finally, about a quarter to a third off the page rebuild time. (Measured by the highly scientific one mississippi two mississippi method. :D)

By the way, there was no plugin cruft on that machine’s Firefox 51, despite its having been running Xubuntu for about three and a half years. Just Flash, Widevine, and that Cisco video codec, all of which survived the upgrade to 52.

Thanks to all the participants here for the nifty Firefox info! Even if it was prompted by my looking in the wrong place for this morning’s gift horse.

Reply | Quote

AlanH wrote:

By the way, there was no plugin cruft on that machine’s Firefox 51, despite its having been running Xubuntu for about three and a half years. Just Flash, Widevine, and that Cisco video codec, all of which survived the upgrade to 52.

Widevine and Cisco OpenH264 codecs are default at installation and Adobe Flash Player seems to be the only plug-in allowed on FF v52.

Windows - commercial by definition and now function...

Reply | Quote

SAAS providers seem to be busy sending out notifications about Firefox 52, as the non-ESR v.52 won’t allow Flash, Silverlight, Acrobat and other plugins. Users requiring those options are being advised to change to using the ESR option, and if Java is required for x64, the 32-bit version should be used instead.
https://support.mozilla.org/t5/Problems-with-add-ons-plugins-or/Why-do-Java-Silverlight-Adobe-Acrobat-and-other-plugins-no/ta-p/31069

2 users thanked author for this post.

walker, ch100

Reply | Quote

Kinda No so for me.
Blue screened a W7 64 bit ultimate in it’s normal update it now. (had always worked)
Had to download a installer on another pc. It too Blue screened. 2nd install try worked.
Nothing found amiss on it. Was a wtf..
Was way faster   page secondary loading gone.
NoScript was disabled.
Leaving only Ghostery
Next? Some trusted sites still use flash.
No idea how to allow in 52 is me.
Roger

Reply | Quote

Update NoScript. It runs OK on my FF v52.

1 user thanked author for this post.

walker

Reply | Quote

@PKCano:   I am at a loss to explain the reason I am getting a prompt from Firefox to “update” to Version 52.0.2 esr.     This is the version which was updated on March 28th except it didn’t have the esr on it.

I’m uncomfortable in trying to update something which has the same number as the original one on March 28th, except for the added esr.   Could you please  advise me as to whether or not it is “safe” to install this update?   A user must be so careful of “everything” out there, that now even something such as this is creating havoc with users.   There is just not enough information on what this “update” is supposed to accomplish.

Thank you so very, very much for your assistance in the past, and hoping you will be able to advise once again.     Your expertise is amazing and outstanding in every respect, and admired by us all.     Thank you.  

 

Reply | Quote

esr stands for Extended Support Release. You are being offered it because you have Flash Player, Silverlight, Acrobat, or other add-ons that you are currently using. The esr version allows you to keep those add-ons.
The standard version (without “esr”) will remove them.
I recommend you go ahead and install the esr version so your Firefox won’t change.

1 user thanked author for this post.

walker

Reply | Quote

Firefox 52 ESR is still in beta.

To download the current release version, you are given ESR 45.9, which was released April 19th.

Reply | Quote

The link downloads esr Release version 52.1.0

Reply | Quote

No, by default you should be offered 52.1 ESR on that page, you’d need to switch to the ESR 45 ‘tab’ (Show: Firefox 52/Firefox 45) to access one of the (probably) more stable but older ‘current’ versions

If you are offered an update to an ESR version from Help > About Firefox, > Check for updates, you almost certainly are already using an ESR version from the same branch (which is my FF preference, much more stable, doesn’t need so many patches/updates – just works).

@Kirsty: you might be checking from within FF 45.8 ESR, so an update to another release from the 45x branch would be expected.

Reply | Quote

@PKCano:  Thank you so much for your advice!   I don’t quite understand Kirsty’s reply to your message, nor “Satrow’s” reply either.   I will perhaps wait a day or 2 to see if there are any other conflicting thoughts on this.

I rely upon your advice  as you are so very knowledgeable about “everything” relating to computers and the on-going changes.

Thank you once again for your guidance!   

Reply | Quote

My apologies @walker, I did not mean to confuse.

When I checked the latest Firefox ESR release notes, it only refers to this week’s release of v. 45.9 ESR.

After I installed that, I checked for updates, using Help>About Firefox>Check for updates, and was assured I had the latest version installed.

As @satrow mentioned, this may be because (as I need Silverlight for work) I only have a 45.* version of Firefox currently installed, the update check has not deemed me fit to require the v.52.*

I do hope that reduces, not increases, your confusion.

Reply | Quote