Quote
anonymous
@siria
It would help if you update lines 68, 70 and 71.

@roytam1
Did a patch disable writing version.ini?
Please update patches and sources.

I don't think there is a change that disable writing version.ini, as there is no code that will write version.ini.

Quote
roytam1

Quote
hermes

Quote
roytam1
just took a look.

Flash plugin steals the focus and the engine under it can't get the element where the mouse cursor is.
Same problem exists when you tried to use F12 tool to get flash plugin object by clicking.

So for this situation, it seems to be no solution.

Not quite. There it was shown that if the plugin mouse gestures disabled everything works just perfect. And even if you turn it on, but not to do more even one mouse gesture all the same works. The problem arises if you make at least one gesture. Some elements (such as switch to full screen mode) will become almost inaccessible. As if they was are shut down a transparent layer (plug-in mouse gestures can do it?).

Flash has nothing to do with it, because without kplugin (mouse gestures) everything will work as it should.

This was a serious problem earlier in the flash version of youtube player. This is an old bug. It would be wonderful if can solve this problem. Anyway I suggest You to see an example. The real test much is more obvious.

with gestures enabled but not used, gestures is actually NOT getting in the windows messaging game YET.

At your discretion, as you think, really nothing can be done? Maybe there's a chance?
Just the contradiction lies in the mouse gestures module.

Quote
siria
Can we rename it please right-away, to KM77-Goanna for now?

Such large-scale changes of course are a new version. It is quite natural and honest.
Although can simply leave the name and change the version (like KM77, for example).
Or not 77. You can call it a 76+.



Edited 5 time(s). Last edit at 12/13/2017 02:10AM by hermes.

Quote
hermes

At your discretion, as you think, really nothing can be done? Maybe there's a chance?
Just the contradiction lies in the mouse gestures module.

So they have different window handle ID and parent ID.

Left is plugin window ID, right is engine window ID.

Which means windows messaging become complicated.

in such case, hijacking mouse events may not be reliable.

BTW in Win10 Build 1703 CBB, only 1st click on plugin is lost, next clicks are passed to plugin window correctly.

@roytam1
You can't say that there is no code that writes version.ini, because you uploaded it here.

Siria said in her last message: "version.ini in the Profile folder contains now a number". Candidates are changes of configuration, different libraries or any other change of sources.

Quote
anonymous
@roytam1
You can't say that there is no code that writes version.ini, because you uploaded it here.

Siria said in her last message: "version.ini in the Profile folder contains now a number". Candidates are changes of configuration, different libraries or any other change of sources.

Yes it is.

But for "no new number", does it necessary?

roytam1,

thanks for making all this possible. Somehow I knew a K-Meleon build based on Goanna could be done. Hope you can iron out all issues that may surface in the coming weeks/months, so that we can receive an official K-Meleon release once more. We need to patch the existing K-Meleon extensions as well where applicable, so that they continue to work with your build.

I wish you all the best. After all, you will face yet another challenge once the Pale Moon team moves to a newer Gecko base (Goanna 4.0?), as they have announced when they introduced the Basilisk project.

Frankly, I don't understand any of this odd discussion, going in circles, sometimes even slightly bickering (Dorian, help...!)

That previous goanna-builds had no version-ini-number was posted by anonymous, so I just assumed it to be true, but have not seen it myself. Have only since 1-2 days access to look at the program myself, and found it does contain (now?) a number. But a definitely wrong one.

I am just a user, without any dev programming skills, just love to play with user-macrolanguage. Everything beyond is greek to me.
But it's still very obvious that of course every new version needs a clearly identifying version number! That's only logical and common sense. Especially for such huge engine steps as this one.
And a special prob, as a macrowriter trying to create multiversion-macros for other users, which are using all sorts of various old KM-versions for various reasons, am struggling *extremely* and wasting enormous time just to swivel around all the various little alpha/beta/final bugs and old/new/missing features in the different builds. And more often than not failing to find acceptable workarounds for the buggier versions, despite all wasted time, grmpf. But just the simple FIRST task of figuring out the user's KM version "somehow", is already a nightmare since 7X, and now slowly turning impossible! May be that I can identify the first Goanna-builds, but every following build will have its own peculiarities again (see the discussions with Hermes), and needs a different version marker.

That there is such a thing as "version.ini" in the profile folder, I learned only very recently from anonymous too. Yet another possible version source. But yet again, for custom builds there's one big chaos. For what it's worth, just as example some past version-numbers:

# version.ini => convert hex to dec => add dots => usage known in...
# 17104933 => 1050025 => 1.5.0025 => KM 1.5.3 + 1.5.4 RC or final +?
# 17170436 => 1060004 => 1.6.0004 => KM ALL 1.6alphabeta + KM1.7alpha1/2
# dto: => KM CCFME0.097.0 same as 1.6 too! no readme-title
# 33554436 => 2000004 => 2.0.0004 KM 74 B / 24.2.0.5118 / exe 2014-01-13)
# dto: KM 74 B2 (g-20140204005730 / 24.3.0 / exe 2014-02-08)
# 33554438 => 2000006 => 2.0.0006 KM 74 B3 (g-20140301194541 / 24.3.0 / exe 2014-03-08)
# 33554439 => 2000007 => 2.0.0007 KM 1.8b24 adodupan (g-20140301194541 / 24.3.0)
# dto: KM 74 B4 (g-20140529202211 / 24.6.0esrpre / exe 2014-05-01)
# dto: KM 74 b4 (g-20140612234948 / 30.0 / exe 2014-06-13)
# 33554450 => 2000012 => 2.0.0012 => KM 74 final +KM74+1 V7 naruman (g-20140803202131 / 24.7.0)
# 33619975 => 2010007 => 2.1.0007 => KM 75.0 final +Pro +? (g-20150426163313 / 31.6.0)
# 33685507 => 2020003 => 2.2.0003 => KM 75.1 alphabetaRC?
# 33685509 => 2020005 => 2.2.0005 => KM 75.1 final +Pro +? (g-20150815021428 / 31.8.0)
# 33751044 => 2030004 => 2.3.0004 => KM 76b3u1, 76b3u2
# 33751045 => 2030005 => 2.3.0005 => KM 76RC1 up1
# 33751046 => 2030006 => 2.3.0006 => KM 76RC1 up2 and THIS GOANNA
# 33751047 => 2030007 => 2.3.0007 => KM 76RC2



Edited 1 time(s). Last edit at 12/13/2017 11:48AM by siria.

Quote
siria
Frankly, I don't understand any of this odd discussion, going in circles, sometimes even slightly bickering (Dorian, help...!)

That previous goanna-builds had no version-ini-number was posted by anonymous, so I just assumed it to be true, but have not seen it myself. Have only since 1-2 days access to look at the program myself, and found it does contain (now?) a number. But a definitely wrong one.

I am just a user, without any dev programming skills, just love to play with user-macrolanguage. Everything beyond is greek to me.
But it's still very obvious that of course every new version needs a clearly identifying version number! That's only logical and common sense. Especially for such huge engine steps as this one.
And a special prob, as a macrowriter trying to create multiversion-macros for other users, which are using all sorts of various old KM-versions for various reasons, am struggling *extremely* and wasting enormous time just to swivel around all the various little alpha/beta/final bugs and old/new/missing features in the different builds. And more often than not failing to find acceptable workarounds for the buggier versions, despite all wasted time, grmpf. But just the simple FIRST task of figuring out the user's KM version "somehow", is already a nightmare since 7X, and now slowly turning impossible! May be that I can identify the first Goanna-builds, but every following build will have its own peculiarities again (see the discussions with Hermes), and needs a different version marker.

That there is such a thing as "version.ini" in the profile folder, I learned only very recently from anonymous too. Yet another possible version source. But yet again, for custom builds there's one big chaos. For what it's worth, just as example some past version-numbers:

# version.ini => convert hex to dec => add dots => usage known in...
# 17104933 => 1050025 => 1.5.0025 => KM 1.5.3 + 1.5.4 RC or final +?
# 17170436 => 1060004 => 1.6.0004 => KM ALL 1.6alphabeta + KM1.7alpha1/2
# dto: => KM CCFME0.097.0 same as 1.6 too! no readme-title
# 33554436 => 2000004 => 2.0.0004 KM 74 B / 24.2.0.5118 / exe 2014-01-13)
# dto: KM 74 B2 (g-20140204005730 / 24.3.0 / exe 2014-02-08)
# 33554438 => 2000006 => 2.0.0006 KM 74 B3 (g-20140301194541 / 24.3.0 / exe 2014-03-08)
# 33554439 => 2000007 => 2.0.0007 KM 1.8b24 adodupan (g-20140301194541 / 24.3.0)
# dto: KM 74 B4 (g-20140529202211 / 24.6.0esrpre / exe 2014-05-01)
# dto: KM 74 b4 (g-20140612234948 / 30.0 / exe 2014-06-13)
# 33554450 => 2000012 => 2.0.0012 => KM 74 final +KM74+1 V7 naruman (g-20140803202131 / 24.7.0)
# 33619975 => 2010007 => 2.1.0007 => KM 75.0 final +Pro +? (g-20150426163313 / 31.6.0)
# 33685507 => 2020003 => 2.2.0003 => KM 75.1 alphabetaRC?
# 33685509 => 2020005 => 2.2.0005 => KM 75.1 final +Pro +? (g-20150815021428 / 31.8.0)
# 33751044 => 2030004 => 2.3.0004 => KM 76b3u1, 76b3u2
# 33751045 => 2030005 => 2.3.0005 => KM 76RC1 up1
# 33751046 => 2030006 => 2.3.0006 => KM 76RC1 up2 and THIS GOANNA
# 33751047 => 2030007 => 2.3.0007 => KM 76RC2

Oh thanks for the list of version numbers.

As I said before, since it is in pre-release stage, versioning is my least concern.

For why KMELEON_VERSION = 0x02030006, it is just because Dorian didn't push RC2 changes to hg repo.

Of course when it can get pass pre-release stage, versioning and packaging improvement can be done.



Edited 1 time(s). Last edit at 12/13/2017 12:43PM by roytam1.

Quote
hermes

Quote
roytam1
just took a look.

Flash plugin steals the focus and the engine under it can't get the element where the mouse cursor is.
Same problem exists when you tried to use F12 tool to get flash plugin object by clicking.

So for this situation, it seems to be no solution.

Not quite. There it was shown that if the plugin mouse gestures disabled everything works just perfect. And even if you turn it on, but not to do more even one mouse gesture all the same works. The problem arises if you make at least one gesture. Some elements (such as switch to full screen mode) will become almost inaccessible. As if they was are shut down a transparent layer (plug-in mouse gestures can do it?).

Flash has nothing to do with it, because without kplugin (mouse gestures) everything will work as it should.

This was a serious problem earlier in the flash version of youtube player. This is an old bug. It would be wonderful if can solve this problem. Anyway I suggest You to see an example. The real test much is more obvious.

One way in which this could be palliated would be adapting some of these Firefox extensions that do the same thing (assuming they work well with Flash players) and in the preferences add an option for the user to decide whether to use the K-Meleon module or the adapted extension.
It could be a temporary solution until the existing bugs in the mouse gestures are resolved.

Quote
Pale as the Moon
I wish you all the best. After all, you will face yet another challenge once the Pale Moon team moves to a newer Gecko base (Goanna 4.0?), as they have announced when they introduced the Basilisk project.

I suppose you are referring to Möbius/UXP, but until that happens it may be a long time since Basilisk is basically a "proof of concept" (or the Moonchild playground if you prefer), plus it is at an early stage of development.

Quote
Rekufad

Quote
Pale as the Moon
I wish you all the best. After all, you will face yet another challenge once the Pale Moon team moves to a newer Gecko base (Goanna 4.0?), as they have announced when they introduced the Basilisk project.

I suppose you are referring to Möbius/UXP, but until that happens it may be a long time since Basilisk is basically a "proof of concept" (or the Moonchild playground if you prefer), plus it is at an early stage of development.

And it is no-go for me (yet) because I have to acquire VS2015 Pro for MFC for KM.

Quote
hermes
@roytam1

I understand the problem described Chanson Yu. Yes, the problem exists. I have prepared for you a local test to reproduce this problem. The problem, as the Hydra, have a three heads.

The first problem was solved by Goanna engine. In normal KM 76RC2 it was impossible to drag files onto the window if you have enabled the KM module of the mouse gestures.
Here there is no problem.

The next problem is that if you ever use a mouse gesture (when switched on gestures, of course) there is a lock included via the <object> tag and other embedded content (like it is in another layer after that). It turns out that the object is hidden behind transparent glass. He realizes not all direct pressing and so on.

You can check it on the prepared sample: (~24 K
(direct link) https://getfile.dokpub.com/yandex/get/https://yadi.sk/d/YsEcY8MD3QVdZK

There's an empty player swf is inserted as iframe. All files is local. Player have (in the upper right corner) a "ring" to switch to full screen mode. Without the module of gestures, the player responds to each click and quickly change modes.

If you enable the module, but not use not a single gesture (using the right mouse button) all is still well. BUT! If you make at least one gesture, player will cease to perceive the pressing on the ring (for fullscreen mode). But it should be noted if you press VERY often, one of 10-15 clicks passes.

A small video to show the idea (only for visualization of the gestures here used old plugin, but essentially it does not change): (~215 K
(direct link) https://getfile.dokpub.com/yandex/get/https://yadi.sk/d/svHEm7_x3QVeLy

If you need support for FLASH / SWF here latest new version of FLASH 28 for KM: (~6.9 M
(direct link) https://getfile.dokpub.com/yandex/get/https://yadi.sk/d/OD9WVUZ53QVdZq
DLL only (but for this example can and old 10th version : )

And the last head of Hydra is that when you move sliders scroll can be lost following mouse gesture. Or more often the omission of a mouse gesture following the release of the slider.

Finally got some free time on observing windows messaging when gestures.dll involved.

So it now detects window-under-cursor and forwards button message back to plugin window. Please try this build:
http://o.rthost.win/gpc/files1.rt/km76g-gestures-nogdip.7z

Quote
ngu
Looks like it has nothing to do with gestures.dll. Even if you delete the gesture.dll file, KM76 still can not well handle AiQiYi flash. I think it is KM76 itself that causes this problem.

Thanks for reminding me,I always enable the gestures plugin,so never noticed the fact you found.

Quote
roytam1
Finally got some free time on observing windows messaging when gestures.dll involved.
So it now detects window-under-cursor and forwards button message back to plugin window. Please try this build:
http://o.rthost.win/gpc/files1.rt/km76g-gestures-nogdip.7z

Sorry for misleading you! As ngu said,KM76 still has the Control Bar issue of flash player even though you disabled the gestures plugin,so may something else causes that ...

________________________________________
K-Meleon zh-CN: http://pan.baidu.com/s/1o6NoHfG

@roytam1
When you use the Copy Link Text function in Context Menu,the contents K-Meleon fetched have a blank space before the target text,should it be there?

________________________________________
K-Meleon zh-CN: http://pan.baidu.com/s/1o6NoHfG

Quote
roytam1

For why KMELEON_VERSION = 0x02030006, it is just because Dorian didn't push RC2 changes to hg repo.

I don't have the change in my repo either, was probably lost.

Quote
roytam1
And it is no-go for me (yet) because I have to acquire VS2015 Pro for MFC for KM.

iirc you don't need pro for MFC


Here my bad adb+ ripoff

Quote
Dorian

Quote
roytam1
And it is no-go for me (yet) because I have to acquire VS2015 Pro for MFC for KM.

iirc you don't need pro for MFC


Here my bad adb+ ripoff

Great news! Thanks! This plugin is VERY GOOD, functionally more convenient than noscript.

Quote
Chanson Yu
@roytam1
When you use the Copy Link Text function in Context Menu,the contents K-Meleon fetched have a blank space before the target text,should it be there?

I agree. A trifle, but constantly have to edit. The absence of this unnecessary symbol would improve usability (as this feature is often used).

Quote
roytam1

Quote
hermes
@roytam1

I understand the problem described Chanson Yu. Yes, the problem exists. I have prepared for you a local test to reproduce this problem. The problem, as the Hydra, have a three heads.

The first problem was solved by Goanna engine. In normal KM 76RC2 it was impossible to drag files onto the window if you have enabled the KM module of the mouse gestures.
Here there is no problem.

The next problem is that if you ever use a mouse gesture (when switched on gestures, of course) there is a lock included via the <object> tag and other embedded content (like it is in another layer after that). It turns out that the object is hidden behind transparent glass. He realizes not all direct pressing and so on.

You can check it on the prepared sample: (~24 K
(direct link) https://getfile.dokpub.com/yandex/get/https://yadi.sk/d/YsEcY8MD3QVdZK

There's an empty player swf is inserted as iframe. All files is local. Player have (in the upper right corner) a "ring" to switch to full screen mode. Without the module of gestures, the player responds to each click and quickly change modes.

If you enable the module, but not use not a single gesture (using the right mouse button) all is still well. BUT! If you make at least one gesture, player will cease to perceive the pressing on the ring (for fullscreen mode). But it should be noted if you press VERY often, one of 10-15 clicks passes.

A small video to show the idea (only for visualization of the gestures here used old plugin, but essentially it does not change): (~215 K
(direct link) https://getfile.dokpub.com/yandex/get/https://yadi.sk/d/svHEm7_x3QVeLy

If you need support for FLASH / SWF here latest new version of FLASH 28 for KM: (~6.9 M
(direct link) https://getfile.dokpub.com/yandex/get/https://yadi.sk/d/OD9WVUZ53QVdZq
DLL only (but for this example can and old 10th version : )

And the last head of Hydra is that when you move sliders scroll can be lost following mouse gesture. Or more often the omission of a mouse gesture following the release of the slider.

Finally got some free time on observing windows messaging when gestures.dll involved.

So it now detects window-under-cursor and forwards button message back to plugin window. Please try this build:
http://o.rthost.win/gpc/files1.rt/km76g-gestures-nogdip.7z

Effect of improvement is clearly visible. The responsiveness of the embedded object is improved in 3-5 times, but unfortunately, after at least one mouse gesture still be problems with responsiveness. Is it possible to raise the sensitivity of the window under the cursor?

If PostMessage function used, can try replace to more fast (but blocking window before execution) SendMessage function?



Edited 2 time(s). Last edit at 12/14/2017 09:49PM by hermes.

Quote
Dorian

Quote
roytam1

For why KMELEON_VERSION = 0x02030006, it is just because Dorian didn't push RC2 changes to hg repo.

I don't have the change in my repo either, was probably lost.

Quote
roytam1
And it is no-go for me (yet) because I have to acquire VS2015 Pro for MFC for KM.

iirc you don't need pro for MFC


Here my bad adb+ ripoff

Thanks!

a rebuilt adblock.dll seems working fine!
http://o.rthost.win/gpc/files1.rt/km76g-adblock.7z

Quote
hermes

Quote
roytam1

Finally got some free time on observing windows messaging when gestures.dll involved.

So it now detects window-under-cursor and forwards button message back to plugin window. Please try this build:
http://o.rthost.win/gpc/files1.rt/km76g-gestures-nogdip.7z

Effect of improvement is clearly visible. The responsiveness of the embedded object is improved in 3-5 times, but unfortunately, after at least one mouse gesture still be problems with responsiveness. Is it possible to raise the sensitivity of the window under the cursor?

If PostMessage function used, can try replace to more fast (but blocking window before execution) SendMessage function?

Using SendMessage in a WindowProc() is no-go.



Edited 1 time(s). Last edit at 12/15/2017 01:33AM by roytam1.

Just having fun again with macroplaying, updated useragent macro with site-switcher, and wondering:
do you happen to know if it's mandatory for that native gecko/goanna function that the pref ...override.websitehere.com is "fully" the same as hostname(URL)? Meaning, would it also work on site "abc.example.net" if users have just stored "example.net" or "www.example.net"...?

Quote
siria
Just having fun again with macroplaying, updated useragent macro with site-switcher, and wondering:
do you happen to know if it's mandatory for that native gecko/goanna function that the pref ...override.websitehere.com is "fully" the same as hostname(URL)? Meaning, would it also work on site "abc.example.net" if users have just stored "example.net" or "www.example.net"...?

if you set "general.useragent.override.example.com" then it affects to *.example.com

Quote
Dorian

Quote
roytam1

For why KMELEON_VERSION = 0x02030006, it is just because Dorian didn't push RC2 changes to hg repo.

I don't have the change in my repo either, was probably lost.

Quote
roytam1
And it is no-go for me (yet) because I have to acquire VS2015 Pro for MFC for KM.

iirc you don't need pro for MFC


Here my bad adb+ ripoff

from https://msdn.microsoft.com/en-us/library/hh967573.aspx :
"You can create MFC or ATL programs with Visual Studio Community Edition or higher. The Express editions do not support MFC or ATL."

OK then it will be fine for me.

Quote
roytam1

Quote
siria
Just having fun again with macroplaying, updated useragent macro with site-switcher, and wondering:
do you happen to know if it's mandatory for that native gecko/goanna function that the pref ...override.websitehere.com is "fully" the same as hostname(URL)? Meaning, would it also work on site "abc.example.net" if users have just stored "example.net" or "www.example.net"...?

if you set "general.useragent.override.example.com" then it affects to *.example.com

Great! Thanks!

Quote
roytam1
a rebuilt adblock.dll seems working fine!
http://o.rthost.win/gpc/files1.rt/km76g-adblock.7z

Dorian, roytam1 thank you very much! The plugin works again!

Just wondering, is that pref still needed for addons? (not contained)
kmeleon.install_firefox_extension

@roytam1

Victory! When testing latest gestures.dll, checking a bug described by me earlier,
the bug is completely eliminated if:

1. You need to use the latest version kplugin mouse gestures (by roytam1)
http://o.rthost.win/gpc/files1.rt/km76g-gestures-nogdip.7z

2. Need to set dom.ipc.plugins.enabled as TRUE.

In these circumstances, no error was observed more in all tests.

I propose to include the corrected plug-in and this specified setting in default configuration
(together with a rebuilt built-in AdBlock plugin).

The source code of ffmpeg can also be added (maybe, as separate folder) to the KM source
code on github? To whole kit was in one place... it would be wonderful.



Edited 3 time(s). Last edit at 12/15/2017 01:35PM by hermes.

First regular/weekly release with known pre-release bugs fixed.

★ K-Meleon 76 - Goanna 3.4.1 (Build 20171215)

Changelog:
Out-of-tree change:
- update Goanna3 to git rev 92ca83173

1998c9c5..e2867829:
- regen jsbridge files (5ca2ae6)
- Update for Goanna (2ff1f9c)
- Update ClearCache to cache2 API (6d2b3c2)
- privacy: Update for cache2 API (746d1d4)
- adapt mozilla.patch for Goanna 3 (85cf101)
- Goanna xulrunner-sdk patch (a418371)
- Goanna old resource:/// behavior patch (optional) (7414c8f)
- Gestures: remove GDI+ drawing (cfc9d23)
- add winEmbed to build (c992abb)
- Fix: CanCopy() always returns true, use !IsSelectionCollapsed() instead. (d772af1)
- winEmbed: Add Enter key "Go" function in ComboBox. (3a0d2e0)
- update default prefs (fa9b565)
- update app default prefs (75992d2)
- don't use MOZILLA_BUILDID (bd6ba75)
- console2: truncate long source line (e08e4ee)
- Trying to fix About Cache Info (80304c1)
- ditch super-long filename and let user enter new name instead (a32244b)
- maybe just truncate at 240th char instead? (8b1dd6c)
- gestures: tab to 4 spaces (ebec8bb)
- gestures: prevent detecting gestures over plugin window (a797fe6)
- adblock: add adblock to build (6508914)
- adblock: fix link lib path for my environment (69839d9)
- kmHelper: update to 76RC2 (9d81050)
- kmHelper: add UserAgentOverrides support (dc86c5c)
- adblock: exclude unused part from compilation, add missing return paths (3c48801)
- bump KMELEON_VERSION (b366a79)
- improve Copy Link Text (e286782)



Edited 1 time(s). Last edit at 12/15/2017 04:39PM by roytam1.

WOW!!! Great job, comrade!!! Respect!

Quote
roytam1
First regular/weekly release with known pre-release bugs fixed.

Luxuriously! Only one comment: need to set dom.ipc.plugins.enabled as TRUE (by default).

[_check_ wish true and GUI for wrapping long lines] .

Speaking of default settings, just discovered this in about:config, because I needed it:
plain_text.wrap_long_lines = false
This is for the view of *.txt lines or similar.
Wouldn't "true" be better for most people?
But probably matter of opinions ;-)

My current new prob is that wrapping a plain-text file by injectCSS doesn't seem to work anymore, as it did in older builds (not sure until which versions exactly).



Edited 3 time(s). Last edit at 01/15/2018 12:26AM by siria.

Quote
roytam1
First regular/weekly release with known pre-release bugs fixed.

★ K-Meleon 76 - Goanna 3.4.1 (Build 20171215)

1. Video doesn't play : https://www.sketchup.com/learn/videos/826
2. HIGH RAM usage and it hangs

Quote
KMfan2
1. Video doesn't play : https://www.sketchup.com/learn/videos/826
2. HIGH RAM usage and it hangs

Add LAV filters from the first post of the thread.