HTTP Switchboard for Chrome/Chromium:

I see, I never thought about that. Probably no need then. Thanks!

 

Thanks for the information.
EDIT: Do you mind if you change the filter you used in Adblock Plus when testing it together with Ghostery, Disconnect, and HTTPS. Fanboy's Ultimate list is the most comprehensive of all list (IMO). Only if you have spare time though or if you will test Kiss Privacy.

 

more goodness in the new 0.7.6.0 update!


Starting version 0.7.6.0, you can access the matrix for behind-the-scene requests by clicking the extension icon while on one of the HTTPSB pages, i.e. Settings, Statistics, Rule manager

it really pays to spend time reading the documentation because this ain't your gran'pa javascript blocker.

 

Ok, I didn't know about a better list, I might re-test using that list, but I will have to also re-test other extensions in order to ensure the tests all use the same web page, and to take into account latest versions.

About KISS, I won't bother, it appears to be a rather forgotten project (last update was two months ago, and before that seven months ago), so I rather not spend time on something which is not actively maintained and little used.

 

I have switched back to Firefox (ABP+ and Noscript) due to the below issues i am facing with Chrome -

- (big problem) I happen to see videos from sites like vidto.me,vidup.me and some other video sharing sites. Randomly, 3-4 times a week, my browser becomes unresponsive, video barely gets played and whole systems comes to the crawl.
Only solution to this problem is to restart my machine. If i see the CPU usage in task manager, the cpu graph and kernel time graph are at constant percent with no variations. I recently installed httpseverywhere. I am not sure if httpseverywhere or flash (and/or any relevant drivers) are culprit.

- (not so big problem) With longer periods of browser usage, i see some times pages takes longer time to load. (Browser is responsive only). I see the culprit here is "trying to establish connection..." part. Although this is greatly reduced to me with recent versions of chrome. But, still noticeable from time to time.

By the way i scanned with Hitman pro (with EWS switch), and MBAM. System is clean.

So, only thing in Firefox i am missing after using Chrome with HTTPSB is the easyway to control cookies and the referrer thingy

Anyone faced a similar issue wrt to Chrome browser hang when playing flash videos. It occurs so rarely, probably 3-4 times a week.
Thanks for reading the lengthy message And appreciate any help

 

Also, can you make sure you have unchecked the "Allow some non-intrusive advertising" in the ABP. I believe most of the users who install it, would definitely uncheck it.

 

seems to be working fine here although all i do is watch stuff on Youtube and the occasional hockey game on CBC Sports.

though these days, i don't know if i am being served Flash or HTML5 videos...

 

With click-to-play, Flash videos can easily be blocked. Unfortunately, that is not the case for HTML5. HTTP Switchboard seems to block them under other though, which is good. You can easily tell which is which by right-clicking the video.

 

Never had any freezes but I mostly only watch videos on YouTube and Dailymotion. But logically, if the problem is on Chrome/Flash then it should also happen in any websites which has flash videos. My blind guess is the security software you are/were using.

 

if video is being served with HTML5, then you wont see anything like "about version..." when you right click on it.
its like built into the browser. and part of the browser i think.

It is very frustrating. Not a straight forward thing to replicate the issue. Which only occurs in chrome. I am thinking it to be more of a problem with the flash version present in the chrome.
It might or might not occur in other websites. I started recently using chrome after HTTPSB only , to raymond.

I guess i will uninstall httpseverywhere and see how it goes.

Thanks, Harsha.

 

@harsha_mic,

are you using the libflashplayer.so, PPAPI (out-of-process), or both of these Flash plugins enabled in chrome?

 

hmm..it should be default. I have not changed anything w.r.t to this. I will check this once i go back to home in the evening.
Thanks for the suggestion wat0114.

Thanks, Harsha

 

Anytime. Hopefully it offers some clues.

 

Finally cleared the review process for Opera addons:

https://addons.opera.com/en-gb/extensions/details/http-switchboard

 

Here http://www.areweprivateyet.com/ it says:

"Fanboy's list has been merged into EasyList"

If true then I suppose the tests do not have to be redone?

 

I truly hate websites that won't load unless you allow cookies. I like HTTP-SB's interface much better than Noscript. I was a Noscript user for a very long time. HTTP-SB is locking down Chrome and the development has been in 5th gear, it's crazy. Gorhill is making many of my extensions unnecessary. I'm a strict blocking kinda guy and those cookies I want I make them read only and last till 2017 .

 

Fanboy's Ultimate List is actually a combination of many list.
Easylist
Easyprivacy
Enhanced Trackers list
Annoyances List

 

If you are telling me "Fanboy's list" is not the same as "Fanboy's Ultimate list", then I guess te test results would be different. I did follow EFF's instructions, which I expect most average users would do. To compare apples with apples, I must test as close as possible to out of the box settings, except here I decided that EFF's instructions would probably be followed by a majority of people interested in enhancing their privacy.

 

tnx a lot for your hard work ray!

 

I am confused here.
I am able to see exactly the same kind of domain/subdomain empty cells before this version too.
you can see one at tomshardware.com

 

The new version is not yet available, it's only the changelog.

 

This week-end I quickly wrote a developer tool extension to benchmark browser session. It is available at https://github.com/gorhill/sessbench#. It is rough around the corner and currently needs a restart if ever a benchmarked web page never finished loading, etc. Regardless, it still is usable to benchmark what I call "browser session", which is trying to mimic how a user would use a browser, aggregating results of all web pages benchmarked within a single session. So here are some results from a browser session script I assembled this morning, using links from the front page of Hacker News as a meaningful sample of links. I tested many blockers, and HTTPSB in two modes.

EDIT 2014-01-21: I revisited the core benchmark code of the dev tool extension in order to remove completely the overhead this extension was introducing in whatever it was benchmarking (it was significant). This changed the results enough for me to replace the results here with more reliable results.

EDIT 2014-01-21 again: Revisited to include 3rd-party hosts. Removed timing because too flaky, use bandwidth/net request count to compare meaningfully load speed.

Ordered by bandwidth:

HTTPSB out of the box
Bandwidth: 15,903,678 bytes
Network hits: 1,261
Cache hits: 21
Cookies (outbound): 0.0
Scripts: 0.0
3rd-party hosts: 66.6

Adblock+ with Easylist and Fanboy's Ultimate
Bandwidth: 21,962,549 bytes
Network hits: 1,714
Cache hits: 338
Cookies (outbound): 1801.6
Scripts: 252.2
3rd-party hosts: 132.0

Ghostery
Bandwidth: 22,888,486 bytes
Network hits: 1,810
Cache hits: 321
Cookies (outbound): 2062.6
Scripts: 254.4
3rd-party hosts: 150.4

HTTPSB allow-all/block-exceptionally
Bandwidth: 22,994,435 bytes
Network hits: 1,862
Cache hits: 296
Cookies (outbound): 2289.6
Scripts: 306.0
3rd-party hosts: 174.8

Disconnect
Bandwidth: 23,008,984 bytes
Network hits: 1,979
Cache hits: 461
Cookies (outbound): 3000.6
Scripts: 338.6
3rd-party hosts: 203.4

No blocker: Reference stats
Bandwidth: 27,026,350 bytes
Network hits: 2,716
Cache hits: 238
Cookies (outbound): 7185.8
Scripts: 686.8
3rd-party hosts: 603.4

The script used for Browser session benchmarker, using the top 15 most visited news web sites:

Code:

clear cache
clear cookies
repeat 5

http://news.yahoo.com/
http://www.huffingtonpost.com/
http://www.cnn.com/
http://news.google.com/
http://www.nytimes.com/
http://www.foxnews.com/
http://www.theguardian.com/
http://www.nbcnews.com/
http://www.dailymail.co.uk/
http://www.usatoday.com/
http://www.washingtonpost.com/
http://www.wsj.com/
http://www.abcnews.go.com/
http://news.bbc.co.uk/
http://www.latimes.com/

 

Had to revisit the results above, the benchmark extension was introducing a serious distortion in load time values. There is no longer any overhead introduced by the benchmark extension.

Also note that the benchmark extension plainly just parses and reports the content of the network tab, so it's really reporting reliable results, and other benefits include:

- Whatever overhead an extension introduce is reflected by the results.

- The results are really the outcome of whatever an extension does to net traffic, so that zero cookies and scripts are reported with HTTPSB out-of-the-box results confirm it does what it says it does.

 

tnx ray.

from the tests above, it looks like HTTP SB does the best job at blocking stuff while being also the fastest.

very well done m8!

 

The speed is tricky here, the timing depends a lot on network performance, which can change from one second to the next. Most reliable indicator for speed I would say are bandwidth and network requests.

I've fine tuned further the benchmarker to really try as much as possible to get meaningful numbers:

- Added "clear cookies" directive, this fixes the inconsistent number of cookies reported across benchmarks.

- Add third-party hosts count, a figure which matters regarding privacy concerns.

I will update the results above with the better fine-tuned benchmarks.

Currently implementing how cookies are counted: I want to count distinct cookies, and to separate 1st- and 3rd-party cookies. I will update when done.