[_check_ loader and duplicate FF-compat-macros] .

...and in last KM76 versions the loader.exe was gone, thought that's intentional, now it's back (old source?)

And the KM75-beta bug with 2 UA macros (both just for compat-mode) is back again.
It would be great too if both could be replaced with the macro that John and I recently posted here:
http://kmeleonbrowser.org/forum/read.php?9,143209
It's basically an updated version of the old one that was included until KM1.6, except that it already includes the FF-compat-toggle (of course BOOL, should be changed for Goanna), and it includes a separate little js-file for the UA list (so far only 1 entry as example, more would be better), and the macro offers to open this list by menu for editing.
Perhaps this macro needs a little bit more polishing if included out-of-box, like checking translations or so, but would see it as a big step forward even as is.



Edited 1 time(s). Last edit at 01/14/2018 01:07PM by siria.

@roytam1
Did you use MozillaBuild 1.8 or 1.10 to build this SDK?

Quote
anonymous
@roytam1
Did you use MozillaBuild 1.8 or 1.10 to build this SDK?

I use 2.1, following Goanna's builkd guide:
https://forum.palemoon.org/viewtopic.php?f=19&t=13556

@roytam1
Thanks. They never had time to update this:
developer.palemoon.org/Developer_Guide:Build_Instructions/Pale_Moon/Windows

Trying to fix console2 goes slow when adding super long lines, but why editing xml binding does not make effect?

EDIT: You have to delete browser\omni.ja and startupCache folder in Profile in order to make effect. 7z archive updated.

From now on, console2 will truncate long source lines to first 100 chars.
Which makes loading DuckDuckGo index page with console2 opened way faster.

By the way, GRE-included chrome://global/content/console.xul is always working well.

EDIT2: I'm thinking separated files in browser or all-in-one in browser\omni.ja which is better.



Edited 2 time(s). Last edit at 12/09/2017 09:33AM by roytam1.

Separated are easier to extend and omni.ja is supposed to be loaded faster (really cannot say whether this is so or not).

There have been cases when users had to edit omni.ja's themselves.
Just noticed: in the gecko-build I can unzip them, in goanna-build not.

Quote
siria
Just noticed: in the gecko-build I can unzip them, in goanna-build not.

It seems to be just the case of old zip. I can transparently open both JA-types by Total Commander (or free Double Commander) with 7zip plugin installed.

Quote
siria
Waah... I see you have already included my trouble-info macro - that's GREAT, except that it's not ready yet by far! Uh oh, must hurry up...

And while you're at macros ) Dorian has added a very long very hard missed function: $OnSwitchTab
Prob is, he only added it to the macros.dll, but forgot the var-creation and function-creation in main.kmm!
Otherwise it works, we've recently tested it. Could you add those 2 snippets in main.kmm yourself, that would be nice? Just search for "ActivateWindow" and copy its lines (in 2 places)

Best You do the kmm - maybe. You know kmm code.

Quote
siria
There have been cases when users had to edit omni.ja's themselves.
Just noticed: in the gecko-build I can unzip them, in goanna-build not.

Update 7zip version maybe.

@siria

Quote
siria
There have been cases when users had to edit omni.ja's themselves.
Just noticed: in the gecko-build I can unzip them, in goanna-build not.

By the fact that it's not really an clean archive, the so-called garbage in the header, similar to SFX. Anyway 7zip 17+ can extract it without any problems. In fact, modification onmi.ja this it basis for firebug integrating.

@roytam1

about:addons hangs at the opening (possible modification of omni.ja as reason).

Quote

From now on, console2 will truncate long source lines to first 100 chars.

Controversial modification, (just as example) I often use both debug consoles (in same time, in parallel). Mostly original. Many extensions write there useful information, then its reading will not be possible or fragmentary.
If it prevents someone it's easy to turn off completely:

user_pref("consoleservice.enabled", false);

For example, I use it to check where coming JSON requests ( https://en.wikipedia.org/wiki/JSON ).

(example #1 = 511 symbols)

_ttps://disk.yandex.ru/monitoring.txt?event=load&app=disk&tm.navigationStart=0&tm.unloadEventStart=0&tm.unloadEventEnd=0&tm.redirectStart=0&tm.redirectEnd=0&tm.fetchStart=0&tm.domainLookupStart=0&tm.domainLookupEnd=0&tm.connectStart=0&tm.connectEnd=0&tm.requestStart=0&tm.responseStart=0&tm.responseEnd=329&tm.domLoading=0&tm.domInteractive=0&tm.domContentLoadedEventStart=0&tm.domContentLoadedEventEnd=0&tm.domComplete=0&tm.loadEventStart=0&tm.loadEventEnd=0&nav.type=1&nav.redirectCount=0
  by  /monitoring/i

(example #2 = 166 symbols)

_ttps://push.yandex.ru/v2/subscribe/websocket?client=ufo&uid=12574499&session=125744991512842997027&service=disk-json%2Cdisk-notifier%3Aweb
  by  /push\.yandex\.ru/i

(example #3 = 216 symbols)

Blocked  _ttps://disk.yandex.ru/monitoring.txt?blockName=listing&event=block_render&app=disk&allTime=573&requestsTime=149&requestsCount=8&selfTime=424&updatesCount=7&factTime=691&factSelfTime=542
  by  /monitoring/i

And they are very long, then how to read them? I don't think it is a good idea. That is a pity if this will be in a final.

But here's something exactly need to truncate: incredibly long page titles when you save them.

For example, try to save this page: http://www.polystyle.com/index.jsp
(if not available you can open from google cache)
http://webcache.googleusercontent.com/search?q=cache:z0f1JU-RwXcJ:http://www.polystyle.com/%2Bhttp://www.polystyle.com/index.jsp&newwindow=1&dcr=0&gbv=1&hl=ru&ct=clnk

Browser won't even will not show dialogue to save because the header a lot more than 250 characters.

<title>PHP Obfuscator, Perl Obfuscator, Java formatter, C# Obfuscator, Javascript formatter, PHP formatter, Visual Basic formatter, C++ formatter, ASP formatter, Actionscript formatter, C# formatter, CSS formatter, HTML formatter, Javascript formatter, JSP formatter, pascal formatter, Delphi formatter, Perl formatter, Python formatter, CMAC formatter | Polystyle - Multi-Language Source Code Formatter</title>

Ie 397 symbols! Because even the save dialog is not only not working (with error), but doesn't appear at all.

I had written a script to fix it.

if(getpref(BOOL,"kmeleon.download.useSaveDir"))setpref(BOOL,"kmeleon.download.useSaveDir", false);
if(getpref(STRING,"browser.download.lastDir")=="C:\\TEMP"){}else{setpref(STRING,"browser.download.lastDir", "C:\\TEMP")}
if(getpref(STRING,"kmeleon.download.saveDir")=="C:\\TEMP\\"){}else{setpref(STRING,"kmeleon.download.saveDir", "C:\\TEMP\\")}

_CUT_TiTLE_{injectJS("if(document.title.slice(0,7)=='[JS] - ')document.title=document.title.slice(7,document.title.length);document.title=document.title.slice(0,48)")}

_SAVE_1_{
if(getpref(INT,"kmeleon.general.saveType")==2){}else{setpref(INT,"kmeleon.general.saveType",2)}
macros(_CUT_TiTLE_)
id(ID_FILE_SAVE_AS)}

_SAVE_2_{
if(getpref(INT,"kmeleon.general.saveType")==3){}else{setpref(INT,"kmeleon.general.saveType",3)}
macros(_CUT_TiTLE_)
id(ID_FILE_SAVE_AS)
setpref(INT,"kmeleon.general.saveType",2)}

_DUAL_SAVE_{
macros(_SAVE_1_)
macros(_SAVE_2_)}

_RESET_TiTLE_{injectJS("document.title='_'")}

_DUAL_SAVE_KEY_{setaccel("ALT S","macros(_DUAL_SAVE_)")
setaccel("ALT VK_F1","macros(_RESET_TiTLE_)")}

$OnInit=$OnInit."_DUAL_SAVE_KEY_;"

_SAVE_FiX_{if(getpref(INT,"kmeleon.general.saveType")==2){}else{setpref(INT,"kmeleon.general.saveType",2)}}

$OnLoad=$OnLoad."_SAVE_FiX_;"
$OnCloseTab=$OnCloseTab."_SAVE_FiX_;"

The idea was to be truncated to 48 characters in the title of the page when saving. If you do not, these pages (even saved in a roundabout way) with such a long name be not undeletable, not renameable and not be defragmented files. Then need to remove with programs like as unlocker.

Therefore written a script above about couple of years ago. Of course it is better to have the value of the variable where you define the length of the header of the page. And truncate to it when saving.

document.title=document.title.slice(0,48)

Javascript function document.title.slice copes with this (as an example). Of course it needs to be activated not always, but only when calling the save dialog.

A simple test set (containing the header in ~250 characters) KM will refuse to save the document:
(direct link) https://getfile.dokpub.com/yandex/get/https://yadi.sk/d/LkUJnB0x3QTwub



Edited 5 time(s). Last edit at 12/09/2017 09:30PM by hermes.

Quote
roytam1


By the way, GRE-included chrome://global/content/console.xul is always working well.

EDIT2: I'm thinking separated files in browser or all-in-one in browser\omni.ja which is better.

Thx.


chrome://global/content/console.xul cannot debug errors in kmm.

omni.ja can be packed when ready for release. Before that people may need to edit.

Without packing could help to make a New Moon and then only add KM specifics to copy there later?

Packing to ja is done fore speeding loading up AFAIK. That is essential when Chrome and Firefox people run for the speediest browser - not for XP systems anymore.

@roytam1

Forgot to mention that now not put and do not work any extension xpi.

Also not working is the clear button to clean the console (the original console). Access to her without editing macro: need remove the macro console2.kmm from the macro folder and restart the browser.

Quote

I'm thinking separated files in browser or all-in-one in browser\omni.ja which is better.

Removal [BROWSER_FOLDER]\BROWSER\omni.ja causes problems with some extensions that worked in builds collected 08.12.2017 (some versions of AdBlock, at least, for example).

In 08.12.2017 builds, both consoles are fully functional (including cleaning feature of the original console). Maybe just a little to roll back the changes?



Edited 8 time(s). Last edit at 12/09/2017 07:40PM by hermes.

@roytam1

I would like to ask a question, if You don't mind. Looking through the keys Your compilation (and the original build of LAV filters) showed the presence of HEVC / H265 decoder.

configure line:
./configure --disable-everything --disable-protocols --disable-demuxers --disable-muxers --disable-filters --disable-programs --disable-doc --disable-parsers --enable-parser=vp8,vp9,aac,aac_latm,h264,hevc,mpegaudio,flac --enable-decoder=vp8,vp9,aac,aac_latm,h264,hevc,mp3,flac --disable-static --enable-shared --disable-debug --disable-sdl --disable-libxcb --disable-securetransport --disable-nvenc --disable-iconv --disable-swresample --disable-swscale --disable-avdevice --disable-avfilter --disable-avformat --disable-d3d11va --disable-dxva2 --disable-vaapi --disable-vda --disable-vdpau --disable-videotoolbox --enable-asm --enable-yasm --toolchain=msvc --disable-sse --disable-sse2 --disable-sse3 --disable-ssse3 --disable-sse4 --disable-sse42 --disable-avx --disable-xop --disable-fma3 --disable-fma4 --disable-avx2 --disable-aesni --build-suffix=-lav

In connection with question: ie supposed to be supported H265 in MP4 container? Right?
I tried to decode a test example: (~8 K
(direct link) https://getfile.dokpub.com/yandex/get/https://yadi.sk/d/yQ9rV35n3QUBLT

But got the message about the wrong mime type.
Maybe it's superfluous, but... is it possible to turn it on?

All other tests pass. A complete set of tests (the same tests YouTube, only locally): (~46 K
(direct link) https://getfile.dokpub.com/yandex/get/https://yadi.sk/d/V1STQvqB3QUBfQ

Made for ease of testing (including on very weak machines or specific emulators).

Quote
hermes
@roytam1

Forgot to mention that now not put and do not work any extension xpi.

Also not working is the clear button to clean the console (the original console). Access to her without editing macro: need remove the macro console2.kmm from the macro folder and restart the browser.

Quote

I'm thinking separated files in browser or all-in-one in browser\omni.ja which is better.

Removal [BROWSER_FOLDER]\BROWSER\omni.ja causes problems with some extensions that worked in builds collected 08.12.2017 (some versions of AdBlock, at least, for example).

In 08.12.2017 builds, both consoles are fully functional (including cleaning feature of the original console). Maybe just a little to roll back the changes?

yeah I'm going back to omni.ja way.
so GRE console and about:addons are working again.

7z archive is updated. please delete browser folder before update.



Edited 1 time(s). Last edit at 12/10/2017 12:16AM by roytam1.

Quote
hermes
@roytam1

I would like to ask a question, if You don't mind. Looking through the keys Your compilation (and the original build of LAV filters) showed the presence of HEVC / H265 decoder.

configure line:
./configure --disable-everything --disable-protocols --disable-demuxers --disable-muxers --disable-filters --disable-programs --disable-doc --disable-parsers --enable-parser=vp8,vp9,aac,aac_latm,h264,hevc,mpegaudio,flac --enable-decoder=vp8,vp9,aac,aac_latm,h264,hevc,mp3,flac --disable-static --enable-shared --disable-debug --disable-sdl --disable-libxcb --disable-securetransport --disable-nvenc --disable-iconv --disable-swresample --disable-swscale --disable-avdevice --disable-avfilter --disable-avformat --disable-d3d11va --disable-dxva2 --disable-vaapi --disable-vda --disable-vdpau --disable-videotoolbox --enable-asm --enable-yasm --toolchain=msvc --disable-sse --disable-sse2 --disable-sse3 --disable-ssse3 --disable-sse4 --disable-sse42 --disable-avx --disable-xop --disable-fma3 --disable-fma4 --disable-avx2 --disable-aesni --build-suffix=-lav

In connection with question: ie supposed to be supported H265 in MP4 container? Right?
I tried to decode a test example: (~8 K
(direct link) https://getfile.dokpub.com/yandex/get/https://yadi.sk/d/yQ9rV35n3QUBLT

But got the message about the wrong mime type.
Maybe it's superfluous, but... is it possible to turn it on?

All other tests pass. A complete set of tests (the same tests YouTube, only locally): (~46 K
(direct link) https://getfile.dokpub.com/yandex/get/https://yadi.sk/d/V1STQvqB3QUBfQ

Made for ease of testing (including on very weak machines or specific emulators).

I suppose mozilla/moonchild people will add HEVC soon, but no ETA.

Quote
roytam1

yeah I'm going back to omni.ja way.
so GRE console and about:addons are working again.

7z archive is updated. please delete browser folder before update.

About addons was working for me all the time. I had to toggle just one pref.

Thought I was alone with the problem because I use an old profile.

p.s. I have an extension/addon that suppresses "addon current" checking.
So after toggle all was ok.

When I unpack the 7z normally it deletes all old rubbish in browser.
Why not now.

Greetings from Hanover



Edited 1 time(s). Last edit at 12/10/2017 12:47AM by guenter.

Quote
hermes
@siria

Quote
siria
There have been cases when users had to edit omni.ja's themselves.
Just noticed: in the gecko-build I can unzip them, in goanna-build not.

By the fact that it's not really an clean archive, the so-called garbage in the header, similar to SFX. Anyway 7zip 17+ can extract it without any problems. In fact, modification onmi.ja this it basis for firebug integrating.

@roytam1

about:addons hangs at the opening (possible modification of omni.ja as reason).

Quote

From now on, console2 will truncate long source lines to first 100 chars.

Controversial modification, (just as example) I often use both debug consoles (in same time, in parallel). Mostly original. Many extensions write there useful information, then its reading will not be possible or fragmentary.
If it prevents someone it's easy to turn off completely:

user_pref("consoleservice.enabled", false);

For example, I use it to check where coming JSON requests ( https://en.wikipedia.org/wiki/JSON ).

(example #1 = 511 symbols)

_ttps://disk.yandex.ru/monitoring.txt?event=load&app=disk&tm.navigationStart=0&tm.unloadEventStart=0&tm.unloadEventEnd=0&tm.redirectStart=0&tm.redirectEnd=0&tm.fetchStart=0&tm.domainLookupStart=0&tm.domainLookupEnd=0&tm.connectStart=0&tm.connectEnd=0&tm.requestStart=0&tm.responseStart=0&tm.responseEnd=329&tm.domLoading=0&tm.domInteractive=0&tm.domContentLoadedEventStart=0&tm.domContentLoadedEventEnd=0&tm.domComplete=0&tm.loadEventStart=0&tm.loadEventEnd=0&nav.type=1&nav.redirectCount=0
  by  /monitoring/i

(example #2 = 166 symbols)

_ttps://push.yandex.ru/v2/subscribe/websocket?client=ufo&uid=12574499&session=125744991512842997027&service=disk-json%2Cdisk-notifier%3Aweb
  by  /push\.yandex\.ru/i

(example #3 = 216 symbols)

Blocked  _ttps://disk.yandex.ru/monitoring.txt?blockName=listing&event=block_render&app=disk&allTime=573&requestsTime=149&requestsCount=8&selfTime=424&updatesCount=7&factTime=691&factSelfTime=542
  by  /monitoring/i

And they are very long, then how to read them? I don't think it is a good idea. That is a pity if this will be in a final.

But here's something exactly need to truncate: incredibly long page titles when you save them.

For example, try to save this page: http://www.polystyle.com/index.jsp
(if not available you can open from google cache)
http://webcache.googleusercontent.com/search?q=cache:z0f1JU-RwXcJ:http://www.polystyle.com/%2Bhttp://www.polystyle.com/index.jsp&newwindow=1&dcr=0&gbv=1&hl=ru&ct=clnk

Browser won't even will not show dialogue to save because the header a lot more than 250 characters.

<title>PHP Obfuscator, Perl Obfuscator, Java formatter, C# Obfuscator, Javascript formatter, PHP formatter, Visual Basic formatter, C++ formatter, ASP formatter, Actionscript formatter, C# formatter, CSS formatter, HTML formatter, Javascript formatter, JSP formatter, pascal formatter, Delphi formatter, Perl formatter, Python formatter, CMAC formatter | Polystyle - Multi-Language Source Code Formatter</title>

Ie 397 symbols! Because even the save dialog is not only not working (with error), but doesn't appear at all.

I had written a script to fix it.

if(getpref(BOOL,"kmeleon.download.useSaveDir"))setpref(BOOL,"kmeleon.download.useSaveDir", false);
if(getpref(STRING,"browser.download.lastDir")=="C:\\TEMP"){}else{setpref(STRING,"browser.download.lastDir", "C:\\TEMP")}
if(getpref(STRING,"kmeleon.download.saveDir")=="C:\\TEMP\\"){}else{setpref(STRING,"kmeleon.download.saveDir", "C:\\TEMP\\")}

_CUT_TiTLE_{injectJS("if(document.title.slice(0,7)=='[JS] - ')document.title=document.title.slice(7,document.title.length);document.title=document.title.slice(0,48)")}

_SAVE_1_{
if(getpref(INT,"kmeleon.general.saveType")==2){}else{setpref(INT,"kmeleon.general.saveType",2)}
macros(_CUT_TiTLE_)
id(ID_FILE_SAVE_AS)}

_SAVE_2_{
if(getpref(INT,"kmeleon.general.saveType")==3){}else{setpref(INT,"kmeleon.general.saveType",3)}
macros(_CUT_TiTLE_)
id(ID_FILE_SAVE_AS)
setpref(INT,"kmeleon.general.saveType",2)}

_DUAL_SAVE_{
macros(_SAVE_1_)
macros(_SAVE_2_)}

_RESET_TiTLE_{injectJS("document.title='_'")}

_DUAL_SAVE_KEY_{setaccel("ALT S","macros(_DUAL_SAVE_)")
setaccel("ALT VK_F1","macros(_RESET_TiTLE_)")}

$OnInit=$OnInit."_DUAL_SAVE_KEY_;"

_SAVE_FiX_{if(getpref(INT,"kmeleon.general.saveType")==2){}else{setpref(INT,"kmeleon.general.saveType",2)}}

$OnLoad=$OnLoad."_SAVE_FiX_;"
$OnCloseTab=$OnCloseTab."_SAVE_FiX_;"

The idea was to be truncated to 48 characters in the title of the page when saving. If you do not, these pages (even saved in a roundabout way) with such a long name be not undeletable, not renameable and not be defragmented files. Then need to remove with programs like as unlocker.

Therefore written a script above about couple of years ago. Of course it is better to have the value of the variable where you define the length of the header of the page. And truncate to it when saving.

document.title=document.title.slice(0,48)

Javascript function document.title.slice copes with this (as an example). Of course it needs to be activated not always, but only when calling the save dialog.

A simple test set (containing the header in ~250 characters) KM will refuse to save the document:
(direct link) https://getfile.dokpub.com/yandex/get/https://yadi.sk/d/LkUJnB0x3QTwub

It seems not that bad that not truncating error message as it will wrap to next line, and it is not bad as source line(counting to 3k column)

for saving super-long page title, I'd take moz/MCP way: let filename empty and user have to enter it.

also trying to fix view cache info menu item.

k-meleon.exe in 7z is updated.

@roytam1
Consoles working, thanks!

Quote

about:addons are working again

In last update - not : ( Installation of the XPi is also ignored.

Quote

for saving super-long page title, I'd take moz/MCP way: let filename empty and user have to enter it.

May by still truncate? At least up to 100 (64? 48 or other of a certain number) symbols? Simple, YES zeroing is a solution but... then you lose all the information about the title. And let's be honest, entering meaningful and detailed name for file (or many files) most will not.

Quote

It seems not that bad that not truncating error message as it will wrap to next line, and it is not bad as source line(counting to 3k column)

Not managed to test, for do this me need to connect the extension for analysis.


@guenter

Quote

About addons was working for me all the time. I had to toggle just one pref.

Thought I was alone with the problem because I use an old profile.

p.s. I have an extension/addon that suppresses "addon current" checking.
So after toggle all was ok.

Tests (at least i make that) are always carried out first in a clean virtual machine, to exclude any influence of any third-party terms and conditions. You also need to consider: the files may change often.



Edited 2 time(s). Last edit at 12/10/2017 02:12AM by hermes.

Quote
hermes
@roytam1
Consoles working, thanks!

Quote

about:addons are working again

In last update - not : ( Installation of the XPi is also ignored.

did you delete startCache in profile as well?
I tried installing ABP and it works.

Quote
hermes

Quote

for saving super-long page title, I'd take moz/MCP way: let filename empty and user have to enter it.

May by still truncate? At least up to 100 (64? 48 or other of a certain number) symbols? Simple, YES zeroing is a solution but... then you lose all the information about the title. And let's be honest, entering meaningful and detailed name for file (or many files) most will not.

KM already tried to truncate filename to 260 chars, but still not working.

	TCHAR* szFileName = new TCHAR[MAX_PATH+1];
	_tcsncpy(szFileName, W2CT(fileName.get()), MAX_PATH);
	szFileName[MAX_PATH] = 0;
	MakeFilename(szFileName);

Edited 1 time(s). Last edit at 12/10/2017 02:12AM by roytam1.

@roytam1

Quote

KM already tried to truncate filename to 260 chars, but still not working.

In the posted example (as direct link) used 250 characters and it was enough for mistakes. I think over 128 is overkill.



Edited 1 time(s). Last edit at 12/10/2017 02:28AM by hermes.

@roytam1

Specially recorded test in a clean virtual machine (~147 K:
(direct link) https://getfile.dokpub.com/yandex/get/https://yadi.sk/d/_i3-apQL3QUG5u

about:addons are not working : (



Edited 1 time(s). Last edit at 12/10/2017 02:29AM by hermes.

@roytam1

Conducted experiment: the (real) limit KM is 248 symbols.

But I think over 128 is overkill.



Edited 3 time(s). Last edit at 12/10/2017 02:51AM by hermes.

Quote
hermes
@roytam1

Specially recorded test in a clean virtual machine (~147 K:
(direct link) https://getfile.dokpub.com/yandex/get/https://yadi.sk/d/_i3-apQL3QUG5u

about:addons are not working : (

OK I forget to update GRE omni.ja, 7z archive updated.

Quote
hermes
@roytam1

Quote

KM already tried to truncate filename to 260 chars, but still not working.

In the posted example (as direct link) used 250 characters and it was enough for mistakes. I think over 128 is overkill.

updated and truncate in 240 chars instead, less code changes involved is good.

Quote
roytam1

Quote
hermes
@roytam1

Specially recorded test in a clean virtual machine (~147 K:
(direct link) https://getfile.dokpub.com/yandex/get/https://yadi.sk/d/_i3-apQL3QUG5u

about:addons are not working : (

OK I forget to update GRE omni.ja, 7z archive updated.

Quote
hermes
@roytam1

Quote

KM already tried to truncate filename to 260 chars, but still not working.

In the posted example (as direct link) used 250 characters and it was enough for mistakes. I think over 128 is overkill.

updated and truncate in 240 chars instead, less code changes involved is good.

XPi are working, the page is can be saved, the title is adjusted, the console is fully functional)
It's all good now) Thanks! Good and very useful work!



Edited 2 time(s). Last edit at 12/10/2017 07:53AM by hermes.

Quote
hermes

Quote
roytam1

Quote
hermes
@roytam1

Specially recorded test in a clean virtual machine (~147 K:
(direct link) https://getfile.dokpub.com/yandex/get/https://yadi.sk/d/_i3-apQL3QUG5u

about:addons are not working : (

OK I forget to update GRE omni.ja, 7z archive updated.

Quote
hermes
@roytam1

Quote

KM already tried to truncate filename to 260 chars, but still not working.

In the posted example (as direct link) used 250 characters and it was enough for mistakes. I think over 128 is overkill.

updated and truncate in 240 chars instead, less code changes involved is good.

XPi are working, the page is can be saved, the title is adjusted, the console is fully functional)
It's all good now) Thanks! Good and very useful work!

So please check if there is any other oddities (comparing to 76RC2) so I can fix it (other than adblock.dll since it has no source code).
If no other problem reported, hopefully I can make first regular release(like my New Moon binaries) with weekly engine update.

@roytam1

Quote

So please check if there is any other oddities (comparing to 76RC2) so I can fix it (other than adblock.dll since it has no source code).
If no other problem reported, hopefully I can make first regular release(like my New Moon binaries) with weekly engine update.

For a full and meticulously testing will take some time.

The readme.html's are still claiming completely wrong Versions, especially for ES+DE+RU even in the title.
And all languages are linking to much outdated release notes, either KM75 or KM74.
http://kmeleon.sourceforge.net/wiki/ReleaseNotes74

For release notes, can you please create a new wiki page ReleaseNotesGoanna or similar? Or anyone else here can do that too.
It could be almost empty for now, at least contain a link to this forum thread, and (more or less slowly) be filled later, some day...

Alternative suggestion, would like best:
unofficial builds often have their own readme-page, completely different from the usual one.
Simply delete all readme-translations, and replace the english readme.html with an own page, just a short Goanna description and some links? Keeping the obsolete KM74/KM75 pages without the slightest change would be very much misleading IMO.

and this too if possible?

Quote
siria
...and in last KM76 versions the loader.exe was gone, thought that's intentional, now it's back (old source?)

And the KM75-beta bug with 2 UA macros (both just for compat-mode, 1 is enough) is back again.
It would be great too if both could be replaced with the macro that John and I recently posted here:
http://kmeleonbrowser.org/forum/read.php?9,143209
It's basically an updated version of the old one that was included until KM1.6, except that it already includes the FF-compat-toggle (of course BOOL, should be changed for Goanna), and it includes a separate little js-file for the UA list (so far only 1 entry as example, more would be better), and the macro offers to open this list by menu for editing.
Perhaps this macro needs a little bit more polishing if included out-of-box, like checking translations or so, but would see it as a big step forward even as is.

Edited 3 time(s). Last edit at 12/10/2017 11:07AM by siria.

Quote
siria
The readme.html's are still claiming completely wrong Versions, especially for ES+DE+RU even in the title.
And all languages are linking to much outdated release notes, either KM75 or KM74.
http://kmeleon.sourceforge.net/wiki/ReleaseNotes74

For release notes, can you please create a new wiki page ReleaseNotesGoanna or similar? Or anyone else here can do that too.
It could be almost empty for now, at least contain a link to this forum thread, and (more or less slowly) be filled later, some day...

Alternative suggestion, would like best:
unofficial builds often have their own readme-page, completely different from the usual one.
Simply delete all readme-translations, and replace the english readme.html with an own page, just a short Goanna description and some links? Keeping the obsolete KM74/KM75 pages without the slightest change would be very much misleading IMO.

and this too if possible?

Quote
siria
...and in last KM76 versions the loader.exe was gone, thought that's intentional, now it's back (old source?)

And the KM75-beta bug with 2 UA macros (both just for compat-mode, 1 is enough) is back again.
It would be great too if both could be replaced with the macro that John and I recently posted here:
http://kmeleonbrowser.org/forum/read.php?9,143209
It's basically an updated version of the old one that was included until KM1.6, except that it already includes the FF-compat-toggle (of course BOOL, should be changed for Goanna), and it includes a separate little js-file for the UA list (so far only 1 entry as example, more would be better), and the macro offers to open this list by menu for editing.
Perhaps this macro needs a little bit more polishing if included out-of-box, like checking translations or so, but would see it as a big step forward even as is.

links and points in browser\omni.ja\chrome\kmeleon\content\kmeleon\about.xhtml is updated.

I'm not good at writing documentation, can someone do me a favor?

for UA issue, should I just replace it with "UserAgent for K-meleon 7x.x 2017-11-27s.zip"?



Edited 1 time(s). Last edit at 12/10/2017 12:26PM by roytam1.

Quote
roytam1
for UA issue, should I just replace it with "UserAgent for K-meleon 7x.x 2017-11-27s.zip"?

USERAGENTS:
In my opinion yes (anyone else?) That is, just the kmm+js files are enough.
Both KM75 compat-macros can be deleted then (1-uagent.kmm is just older version of 2-useragentmode.kmm)
The function of 2-useragentmode.kmm is included in our newer macro too, so can be deleted too.

There is just one prob: the different compat-mode toggles
(but not worse compared to the prob in current build, so for now it is still a progress)

In gecko (in your kmeleon.js too) the compat pref is:
BOOL: general.useragent.compatMode.firefox
In goanna (in your "kmprefs-additional.js"):
INT: general.useragent.compatMode

So it's 2 different pref names, right?
Does the gecko-pref has any effect in goanna? And vice-versa?
Another tiny prob, the INT pref is unknown for non-goanna-experts, if you would add a comment line in the js what 0/1/2 are doing exactly that would help too. .

I would like to modify the new macro to include both compat-modes somehow, for universal use in all KM-versions. Just a bit complicated to decide how and what exactly, currently rather unsure.... or perhaps keeping the compat-stuff in separate macros again.... but that makes it hard to show them in the same menu, especially in the default button menu, which is created in toolbars.cfg...

And still one of the main probs for goanna-macros:
just exists no very simple engine-check, if gecko or goanna!
Best would be a simple pref, created in kmeleon.js.
And/or platform.ini declaring "=goanna" in a [K-Meleon] section.

In KM1.X things were so easy, they contained a pref for vendor+vendorSub, and engine was always gecko.
No idea why Mozilla wanted those prefs deprecated, but if written into js-file they are still usable anyway.
general.useragent.vendor "KM-Goanna" perhaps? (have no clue though what is 'appropriate' or not)
Or revive one of those even older obsolete prefs?
general.useragent.appName
general.useragent.appVersion
general.useragent.productComment
general.useragent.vendorComment
Or how about a completely new pref in kmeleon.js?
Hehe... Supposing Dorian wouldn't mind???
kmeleon.engine=gecko or goanna
kmeleon.engineSub=3.4.1

Prob is you're flying at rocket speed and I'm far too slow trailing behind :-/