[AdGuard for Windows] Beta v.6.0.224 RC

[vasily_bagirov]

Greetings!

Spoiler

There is a new beta version available out there. It serves mostly as a bug-fixer, with a couple of exceptions. We know that you guys mostly want to see more new cool features and less small technical changes, but they are necessary, especially since this version is a release candidate for R2. We promise, if this time there are no new bugs, we will get to 'real' stuff after the release The most notable changes in this beta are slightly new WFP network driver, which will hopefully fix some compatibility and stability issues, and adding of a $stealth modifier, which is explained in detail below.

Changelog:

Spoiler

[SIZE=3]UI changes[/SIZE]
[Changed] Adguard uses native-style tray menu now [url=https://github.com/AdguardTeam/AdguardForWindows/issues/1006]#1006[/url]

[SIZE=3]Ad Blocker[/SIZE]
[Improved] Added $stealth modifier that disables Stealth Mode for the specified website [url=https://github.com/AdguardTeam/AdguardForWindows/issues/339]#339[/url]

As we already mentioned in our [url=https://blog.adguard.com/en/news/improved-stealth-mode-module-for-adguard-for-windows.html]blog[/url], separate whitelist for Stealth Mode is on our roadmap. 
It is a much needed feature for anyone who actively uses Stealth Mode. We are not quite there yet, but we made a big step in the right direction by implementing $stealth modifier. 
It works similar to all other modifiers, just add it to any exception rule and it will disable Stealth Mode for that website specifically, even if the module itself is enabled. 
For example: [CODE]@@||example.org^$stealth

You can find more information about rules formatting on our website: adguard.com/filterrules.html

[Fixed] Adguard does not try to filter data: URLs any more #1007
[Fixed] $domain modifier now does work for blocking WebSocket connections #998
[Fixed] Adguard now doesn't ignore <base> tag href attribute while constructing absolute URL #978

Languages and Translations
[Added] Croatian language

Install/Update
[Fixed] An issue when update errors were shown as valid updates in tray notification #993
[Fixed] Adguard now detects Chrome in 'Filtered apps' on Windows 10 insider build #972

Networking
[Fixed] Adguard can now detect HTML if there is an xml declaration at the document start #1011
[Fixed] ESET compatibility issue when Chrome browser would randomly hang #899

Minor
#1014
[/CODE]

UPDATE: version 6.0.223

This version is a release candidate. It is a small update where we basically clean up some leftover bugs and defects from the last beta. Nothing too serious, just some things that need to be done before releasing a stable version.

You can find the full description on GitHub: https://github.com/AdguardTeam/AdguardForWindows/releases

And download it at this ink: https://github.com/AdguardTeam/AdguardForWindows/releases/download/6.0.223.1092/Setup.exe

Chengelog:

Spoiler

[SIZE=3]UI changes[/SIZE]
[Changed] Adguard window is now opened with a single click on tray menu [URL="https://github.com/AdguardTeam/AdguardForWindows/issues/1030"]#1030[/URL]

[SIZE=3]Ad Blocker[/SIZE]
[Changed] $replace rules should now have higher priority than other basic rules [URL="https://github.com/AdguardTeam/AdguardForWindows/issues/1023"]#1023[/URL]
[Fixed] $domain modifier now works for WebSocket connections [URL="https://github.com/AdguardTeam/AdguardForWindows/issues/998"]#998[/URL]

[SIZE=3]Networking[/SIZE]
[Added] Support for HTTP method REPORT [URL="https://github.com/AdguardTeam/AdguardForWindows/issues/1020"]#1020[/URL]
[Fixed] Issue with unexpected connection reset with TDI driver [URL="https://github.com/AdguardTeam/AdguardForWindows/issues/1025"]#1025
[/URL]
[SIZE=3]Minor[/SIZE]
[URL="https://github.com/AdguardTeam/AdguardForWindows/issues/1029"]#1029[/URL]

Update: a very small update consisting only of a couple of bugfixes:
[Fixed] Issue with unexpected connection reset with TDI driver #1025
[Fixed] Bug when parser builds wrong absolute URL #1021

Waiting for your comments!

 

[steve1209]

Updated smoothly, no issues here...

 

[Blaz]

Thanks, working fine.

It would be good if there would be an option/fix in order to inject Adguard into probably broken sites like http://www.mystreans.com/livetv/19.html

 

[mysteriously]

First look at the new build made me wonder if this is the first build where infinite web pages loading bug is fixed. If so, then it was ESET WFP driver incompatibility and had nothing to do with Sandboxie and HTTPS filtering.
Currently I don't have Adguard SSL cert installed for Firefox. SSL scanning is also disabled in ESET settings. I'll test it on the weekend.

 

[avatar]

It would be good if there would be an option/fix in order to inject Adguard into probably broken sites like http://www.mystreans.com/livetv/19.html

I am pretty sure we should not inject our code into such invalid html pages. These are rare cases which we can handle with $replace rules.

---------- Post added at 02:31 AM ---------- Previous post was at 02:29 AM ----------

First look at the new build made me wonder if this is the first build where infinite web pages loading bug is fixed. If so, then it was ESET WFP driver incompatibility and had nothing to do with Sandboxie and HTTPS filtering.

Wasn't it fixed in the previous beta version? We've made pretty important changes there that should have helped with it.

 

[Boo Berry]

Nice to see the $stealth modifier, this should help greatly.

Update went nice and smooth here too.

 

[avatar]

Nice to see the $stealth modifier, this should help greatly.

Yep, it couldn't wait anymore

What's important is that $stealth rules won't mess with other modules.
It affects only stealth mode module.

 

[Blaz]

@avatar
It just seems a lot of streaming sites are broken as I only see it with such sites. And Ad Muncher is injecting itself there. What exactly is broken with the site?

 

[avatar]

@avatar
It just seems a lot of streaming sites are broken as I only see it with such sites. And Ad Muncher is injecting itself there. What exactly is broken with the site?

Nothing with this particular site. But I am afraid that weakening html validation rules may lead to lots of unnecessary injections (to some ajax responses for instance).

 

[Blaz]

Add it as an experimental feature and we will see.

 

[pegas]

Updated smoothly and running fine, well done!

 

[avatar]

Add it as an experimental feature and we will see.

This particular issue may be solved by adding "title" to the list of tags we use to detect if html is valid.
So we'll add it and see how it goes.

 

[Blaz]

Thanks for considering it.

@@ $urlblock is only disabling all content delivered from that site? As when I tried it, it didn't unblock every URL which get initially requested from that site.

e. g. site google.com makes request to yahoo.com in order to load some pictures and the pictures had been still blocked.

 

[avatar]

@@ $urlblock is only disabling all content delivered from that site?

Yes.

As when I tried it, it didn't unblock every URL which get initially requested from that site.

e. g. site google.com makes request to yahoo.com in order to load some pictures and the pictures had been still blocked.

Hm, are you sure it was originated from google.com and not just some nested iframe?

 

[Blaz]

It was only an example.
It was the issue here:

http://forum.adguard.com/showthread.php?10273-*-Page-damage-throughout-http-www-dailytelegraph-com-au-newslocal-south-west

If $urlblock is only disabling requests from that mentioned site, there should be a modifier in order to (un)block all URL requests outgoing from the mentioned site in the url, e. g. dailytelegraph.com.au in a rule @@||dailytelegraph.com.au^$new-modifier

 

[avatar]

It was only an example.
It was the issue here:

http://forum.adguard.com/showthread.php?10273-*-Page-damage-throughout-http-www-dailytelegraph-com-au-newslocal-south-west

If $urlblock is only disabling requests from that mentioned site, there should be a modifier in order to (un)block all URL requests outgoing from the mentioned site in the url, e. g. dailytelegraph.com.au in a rule @@||dailytelegraph.com.au^$new-modifier

Something like that?

http*$domain=dailytelegraph.com.au

 

[Blaz]

Yes, something like you mentioned, so it would allow/block every URL request outgoing from the mentioned site.

 

[mysteriously]

Wasn't it fixed in the previous beta version? We've made pretty important changes there that should have helped with it.

I'm not sure. I'll test the most recent version more on the weekend.

 

[Phonemes]

With this latest beta, I often get "the connection was reset" messages in Firefox. I have seen these with earlier versions but not as frequently as at the moment. Reloading the page works in the majority of cases.

 

[vasily_bagirov]

With this latest beta, I often get "the connection was reset" messages in Firefox. I have seen these with earlier versions but not as frequently as at the moment. Reloading the page works in the majority of cases.

Does this happen when you try to open a website? Is https relevant for this issue?