uBlock Origin, Chromium 54 and Firefox 52.9 ESR

[Sampei.Nihira]

@PROBLEMCHYLD

 

Unfortunately it is not applicable to the UBO Legacy version:

 

https://github.com/gorhill/uBlock-for-firefox-legacy/issues/283

The list you entered is out of date.
You can delete Dandelion Sprout's Bad top-level domains.

This is the updated list:

! The Most Abused Top Level Domains
! https://www.spamhaus.org/statistics/tlds/
||rest^$all
||gq^$all
||work^$all
||tk^$all
||ml^$all
||viajes^$all
||casa^$all
||london^$all
||cf^$all
||fail^$all
||surf^$all
||exposed^$all
||fit^$all
||date^$all
||golf^$all
||email^$all
||top^$all
||cam^$all
||ga^$all

 

Edited March 20, 2021 by Sampei.Nihira

[Sampei.Nihira]

 

https://www.wilderssecurity.com/threads/ublock-a-lean-and-fast-blocker.365273/page-214#post-2997739

 

If among the members of MSFN there is someone who can help Just Off I invite him to come forward.

Edited March 28, 2021 by Sampei.Nihira

[Sampei.Nihira]

 

https://www.ghacks.net/2021/04/26/evaluate-your-content-blocker-with-ad-block-tester/

I allowed:
 

d3ward.github.io

so my normal browsing a bit more permissive (because otherwise the test doesn't work):

Not bad with my few lists in UBO.

Quote

48.127 network filters ＋ 32.019 cosmetic filters

 

Edited April 26, 2021 by Sampei.Nihira

[VistaLover]

@Sampei.Nihira :

In latest Serpent 52, I only use uBlock Origin 1.16.4.28 ; when I first ran the test, I got a score of 84%  ; 
enabling the built-in AdGuard Tracking Protection filter list, it goes up to 90%;
however, tracking down which hostnames were allowed in the tool's results, and creating global "block" rules in uBO:

* analytics.facebook.com * block
* analytics.pointdrive.linkedin.com * block
* widgets.pinterest.com * block
* analytics.pinterest.com * block
* trk.pinterest.com * block
* appmetrica.yandex.com * block
* yandexadexchange.net * block
* analytics.mobile.yandex.net * block
* extmaps-api.yandex.net * block
* adsdk.yandex.ru * block
* data.mistat.xiaomi.com * block
* data.mistat.intl.xiaomi.com * block
* data.mistat.india.xiaomi.com * block
* data.mistat.rus.xiaomi.com * block
* logservice.hicloud.com * block
* logservice1.hicloud.com * block
* logbak.hicloud.com * block
* insights.samsung.com * block
* analytics-api.samsunghealthcn.com * block
* supportmetrics.apple.com * block
* metrics.icloud.com * block
* metrics.mzstatic.com * block

... I can get 100% :

(... without enabling AdGuard Tracking Protection list, which is quite a big one... ) .

FWIW, another tester tool can be tried on: 

https://adblock-tester.com/

Happy blocking everyone...

 

[Sampei.Nihira]

@VistaLover

Probably the best test is the initial one, then the results tend to become unreliable:

https://www.wilderssecurity.com/threads/ublock-a-lean-and-fast-blocker.365273/page-219#post-3002981

 

Edited April 26, 2021 by Sampei.Nihira

[Sampei.Nihira]

Another method to get 100% is to use:

https://controld.com/

(Block Malware + Ads)

you can completely disable UBO.

It seems a bit too simple.

Edited April 27, 2021 by Sampei.Nihira

[XPerceniol]

Wow.. Wow.. Wow!

Thank you so much @Sampei.Nihira - what a great way to start the day here

First time I ran it: I got to 92%, so I added your list @VistaLover and reached 97%.

I added the following that were still showing 'not blocked':

freshmarketer.com
cdn.luckyorange.com
w1.luckyorange.com
browser.sentry-cdn.com
ads.facebook.com
analytics.oneplus.cn
click.oneplus.cn
samsung-com.112.2o7.net

And; BINGO!

Wishing you both a great week ahead and stay healthy :)

Edited April 27, 2021 by XPerceniol

[Dave-H]

I get a 100 score on https://adblock-tester.com/.
However https://d3ward.github.io/toolz/src/adblock.html doesn't work at all for me, even with d3ward.github.io whitelisted.
I just get an animation permanently and nothing else happens.
This on Firefox 52.9.1 ESR.

 

[VistaLover]

8 hours ago, Sampei.Nihira said:

Another method to get 100% is to use:

https://controld.com/

(Block Malware + Ads)
you can completely disable UBO.

... It's the latest offering from the Canadian makers of Windscribe VPN and related browser extensions (they currently only offer WEs for Chrome/Firefox/etc., but JustOff has also made an unofficial "legacy" fork that works in NM27/NM28/St52/St55 !

It's too bad the free ControlD DNS servers do not offer geo-block circumvention features, because that is the main reason I currently use Windscribe for (with "privacy" protection as an added bonus... ) .

Saluti

[VistaLover]

1 hour ago, Dave-H said:

However
https://d3ward.github.io/toolz/src/adblock.html
doesn't work at all for me, even with d3ward.github.io whitelisted.
I just get an animation permanently and nothing else happens.
This on Firefox 52.9.1 ESR.

Hi @Dave-H ; the d3ward test, hosted on GitHub, requires browser APIs that Firefox ESR 52.9.1 simply lacks, because it has been left to rot after its EOL in Sept 2018 (two and a half years already ) ; I won't even touch the security aspects of still using it, but the web itself has evolved greatly since then, usually for the worse, mostly according to what is being dictated by Google Chrome devs (which has been the topic of another recent post of mine...).

After Microsoft acquired GitHub, one of the first things they did was to fire the old GH devs, who were more "willing" to keep GH backwards-compatible with non-Chromium/older browser engines; the M$ team that stepped in soon abandoned support for older browsers and quickly introduced "new" web technologies (read Chrome-isms), since their new and shiny toy browser, ChrEdge/Win7+, uses now the Chromium engine...

One of these APIs is the AbortAPI (aka abortController) and it is now required for that adblock test to work successfully... I loaded the test URI in my portable copy of FxESR 52:

and the error is generated because of

    //Function to check an url and set red/green result
    async function check_url(url, div, parent, np) {
        // Lets set up our `AbortController`, and create a request options object
        // that includes the controller's `signal` to pass to `fetch`.
        const controller = new AbortController()
        const config = {
            ...{
                method: 'HEAD',
                mode: 'no-cors'
            },
            signal: controller.signal
        }

failing... OTOH, UXP-based browsers after October 2020 do have support for abortController API, hence the test loads and completes successfully there...

I know you are very much entitled to your choice of FxESR 52 as your "main" browser in your XP-partition, but perhaps you should seriously consider migrating to latest Serpent 52.9.0 (it isn't such a huge leap from FxESR 52, but it is a maintained application with considerably better response to "current" web than the now deprecated browser it was forked off... ) . Just my 2p, of course ...

Cheers

Edited April 27, 2021 by VistaLover

[XPerceniol]

@DaveH I should have said it doesn't work for me either on Firefox 52.9.1 ESR Unfortunately, I didn't do so well (63 points out of 100) with https://adblock-tester.com/

I just (now) ran it again and scored 76 out of 100 this time?

But 100% with Serpent 52.9.0

Edited April 27, 2021 by XPerceniol

[Sampei.Nihira]

Sometimes to check if the tests work it is necessary to reverse the problem.
I did it, I tried to get 0%.
It seems impossible to me, so the test is unreliable, as I wrote yesterday.

Another reason it became unreliable is because the lists maintainers added the missing trackers.

Edited April 27, 2021 by Sampei.Nihira

[Dave-H]

44 minutes ago, VistaLover said:

Hi @Dave-H ; the d3ward test, hosted on GitHub, requires browser APIs that Firefox ESR 52.9.1 simply lacks, because it has been left to rot after its EOL in Sept 2018 (two and a half years already ) ; I won't even touch the security aspects of still using it, but the web itself has evolved greatly since then, usually for the worse, mostly according to what is being dictated by Google Chrome devs (which has been the topic of another recent post of mine...).

After Microsoft acquired GitHub, one of the first things they did was to fire the old GH devs, who were more "willing" to keep GH backwards-compatible with non-Chromium/older browser engines; the M$ team that stepped in soon abandoned support for older browsers and quickly introduced "new" web technologies (read Chrome-isms), since their new and shiny toy browser, ChrEdge/Win7+, uses now the Chromium engine...

One of these APIs is the AbortAPI (aka abortController) and it is now required for that adblock test to work successfully... I loaded the test URI in my portable copy of FxESR 52:

and the error is generated because of

    //Function to check an url and set red/green result
    async function check_url(url, div, parent, np) {
        // Lets set up our `AbortController`, and create a request options object
        // that includes the controller's `signal` to pass to `fetch`.
        const controller = new AbortController()
        const config = {
            ...{
                method: 'HEAD',
                mode: 'no-cors'
            },
            signal: controller.signal
        }

failing... OTOH, UXP-based browsers after October 2020 do have support for abortController API, hence the test loads and completes successfully there...

I know you are very much entitled to your choice of FxESR 52 as your "main" browser in your XP-partition, but perhaps you should seriously consider migrating to latest Serpent 52.9.0 (it isn't such a huge leap from FxESR 52, but it is a maintained application with considerably better response to "current" web than the now deprecated browser it was forked off... ) . Just my 2p, of course ...

Cheers

Thanks, that explains the problem!

This is off-topic, but I'm actually constantly surprised at just how well Firefox 52.9 ESR does still work, as you say it's now years old!
It's had all the tweaks of course, which has improved its performance. It's still rather slow on javascript heavy sites like Facebook, but I'm impressed that it still works there at all!
Even videos still seem to be no problem on YouTube, Facebook, and Instagram, despite a few scares in recent years.

My only major reason not to switch to something else is really because I use Firefox Sync to synchronise my browsing data between Firefox on XP, Windows 10, and my mobile phone, and from what I can gather none of roytam1's browsers can now use the Mozilla Sync system, despite being originally Mozilla based.
Eventually I will have to bite the bullet of course, but at the moment it's actually quite rare to find a site which is unusable in Firefox 52 ESR, and all the sites I use regularly are fine with it, so until that changes (which it inevitably will of course) I will stick with it.
Cheers, Dave.

 

[VistaLover]

On 4/27/2021 at 8:05 PM, Dave-H said:

and from what I can gather none of roytam1's browsers can now use the Mozilla Sync system,
despite being originally Mozilla based.

 Only official Mozilla applications are entitled/authorised to use the Mozilla (Firefox) Sync infrastructure (and are thus allowed to carry the Mozilla-specific "authentication key" that enables connections with MozSync servers) ...

 In its initial life, Basilisk 52/UXP by MCP did have that auth-key, so its users could (ab)use Mozilla Sync servers (and sync between mobile Fx, too); but at a later point, Moonchild decided to "right the wrong" (basically, to keep Mozilla out of his back yard ) and newer Bk52 builds lost the ability to connect to MozSync...
 In any way, as Bk52 progressed further away from FxESR 52.9 (while Fx itself was moving quickly into its "Quantum" days), what could be synced safely between the two applications (via MozSync) was reduced to bookmarks, browsing history, browsing session, perhaps passwords; full profile sync (e.g. include installed extensions) was not possible, what was indeed probable was profile corruption if a user did not pay attention... 

As Basilisk 52 was eventually transformed into a "clone" (minus a few bits) of Pale Moon (but with the Australis interface), the two MCP apps now use Pale Moon Sync (MCP-maintained/owned) and can sync data between them; since mobile versions of the two do not exist, that leaves out syncing between a mobile device...  

UXP-based @roytam1's browsers, like NM28/Serpent52, also use the Pale Moon Sync infra (I suspect MCP aren't that enthused), but there exist at least two other "roytam1's browsers" that, AFAICT, are able to use Mozilla Sync: His Firefox 45esr fork, aimed at old hardware (CPU without SSE2) users, and his Serpent 55 fork !

Serpent 55 was initially based on MCP's Basilisk 55/Moebius project, later abandoned in favour of Basilisk52/UXP; despite the "55", it was forked off a Mozilla Firefox 53.0a1 code snapshot...
St55 currently is a "test" application, code from UXP and a few other projects is being applied to it; is updated by Roy at least once every month... It currently has better features compared to Fx 52esr, especially in the TLS/HTTPS section (has support for final TLS v1.3), has more Web Extension APIs compared to Serpent 52, but doesn't carry many of the recent WebAPIs St52 does; e.g. ,
https://d3ward.github.io/toolz/src/adblock.html
doesn't work in St55 (because of missing AbortAPI) ... And since Bk55 was "touched" () by MCP, St55 has been "optimised" to work as single-process application (I say this in case you are running FxESR 52 under XP with the "e10s hack"...).

So, you can at least "trial" Serpent 55 on your XP partition without (hopefully) losing your Mozilla Sync "workflow" .

In the longer run, you should definitely consider migrating to Serpent 52, sooner rather than later... A member in the official PM forums has recently posted that Basilisk 52 supports the third-party Sync solution called "EverSync", and the same stands true for St52:

https://www.everhelper.me/synchronizer.php
https://www.everhelper.me/everhelperplans.php
https://addons.mozilla.org/en-US/firefox/addon/fvd-synchronizer/

Their "For Firefox" extension is compatible with St52, you have to register an account with them (but the same is true for MozSync); they advertise that you can sync data between "any computer and any browser"; the same, supposedly, is true for mobile devices, once you install there their Android app - their free plan looks promising... BTW: I have no affiliation with them in the slightest, nor do I use any of their offerings (); just pointing out that Sync "solutions" do exist outside of MozSync...

My best regards

[Dave-H]

Wow, thank you so much for all that information!

Sounds like Serpent is the one to try then when I inevitably have to finally abandon Firefox 52 ESR.
Good to know that there are alternatives to the Mozilla Sync system, as I would really miss that facility if I had to lose it.
Keeping bookmarks and saved passwords in sync is my main concern.