KernelEx Apps Compatibility List (New)

[schwups]

My icons are ok. Just testet.

 

Sumatra 3.1.1

Kexbases: v11

Kexbasen: v12

KernelEx: v12

no Kext

Edited March 5, 2016 by schwups

[MiKl]

Tested Sumatra 3.1.1 and icons are fine on my system too and I could also print. IIRC it was reported that printing is broken in latest

versions. But unfortunately this version freezes when I click on the menu button.

[MiKl]

IrfanView 4.42 and Plug-Ins seem to be fine. But unfortunately I can not open pdf's.

Error message states:  'Decoding error ... PDF.DLL could not be loaded'. I tested several comp-modes on exe and dll.

[loblo]

PDF.DLL now requires fontsub.dll. Find it and put it somewhere in the irfanview or system path and it'll work.

[MiKl]

I tried a few versions of fontsub.dll but with this dll in place Irfanview crashes completely !!

Maybe one of the things that is not working on 98SE ... but thanks for checking.

Edited March 11, 2016 by MiKl

[loblo]

Works fine for me with the first I found which is version 5.1.2600.0 (xpclient.010817-1148).

[DougB]

FontSub.dll has been working for me as well, v5.1.2600.5888 from XP SP3 (or maybe USP4).  I put it in the Windows\System\ folder since it's a Ms file, but it would work in the I_View32.exe folder too (not the Plugins\ folder).  (IrfanView v4.42)  

- Doug B.

[MiKl]

Hmmmm, the strange world of Windows. On one of my setups I could easily install the latest j*v* but only there - on the other machines I have received this annoying error 1723.

However, in case Jumper is interested here is the error report.

 

I_VIEW32 führte eine ungültige Anweisung in
Modul PDF.DLL bei 018f:01970d8c aus.
Register:
EAX=01fd9208 CS=018f EIP=01970d8c EFLGS=00010206
EBX=0000001f SS=0197 ESP=007ebbb8 EBP=007ebbd0
ECX=007ebb83 DS=0197 ESI=01fd8d79 FS=5def
EDX=00000002 ES=0197 EDI=000001a2 GS=0000
Bytes bei CS:EIP:
66 0f 11 05 68 8a fd 01 0f 28 05 90 26 c7 01 66
Stapelwerte:
0196e8fa 01c713d8 01c2546b 00000000 00000001 82399200 007ebbe0 01c253cc 01c7135c 01c719e4 007ebc14 01c1e1eb 00000000 ca11e37c 00000000 00000001

[jumper]

I have no idea what you mean by "j*v*" or "error 1723".

Enable API hook on pdf.dll on both systems and compare the console logs.

[dencorso]

May I guess "j*v*" should mean "java"?

[loblo]

On 29 February 2016 at 9:46 AM, loblo said:

On the positive side, and before any of those changes, some apps that did crash in kexbases.dll and that I could only run from within Dependency Walker now run fine on their own with the latest version of this dll.

 

Amongst the handful of those I think it is worth mentioning the recently released version 6 of Microsoft Visual True Type which appears to be compiled with Visual Studio 2015 (most of the code in that is from the late 1990's however). It did require some extra hexing hodgepodge  however as some definitions  I did need to add to kstub822.ini broke other software (Networx). It also requires Vista's UIAutomationCore.dll which I got from a dll dump website.

 

I have not tried so far to see if disabling system dlls would help with the Gimp plugins and git crashes as I am still using original kexbasen.dll. Kstub822.ini cleanup won't help for sure as the crashes occur without stubs dlls at all.

 

Edit: A screenshot of VTT running (flawlessly) under Windows ME.

 

 

I am sorry to say this but this doesn't really belong here as this is possible only with a KernelEx development build which breaks a few other things working with the stable KernelEx version... Altough I told LoneCrusader to do whatever he thought was best with this post that for some unfathomable reason upsets jumper so much in the KernelEx updates thread that he holds of releasing new builds because of it , I've got to say that IMHO it belongs exactly where it originally was and not here. 

[LoneCrusader]

1 hour ago, loblo said:

I am sorry to say this but this doesn't really belong here as this is possible only with a KernelEx development build which breaks a few other things working with the stable KernelEx version... Altough I told LoneCrusader to do whatever he thought was best with this post that for some unfathomable reason upsets jumper so much in the KernelEx updates thread that he holds of releasing new builds because of it , I've got to say that IMHO it belongs exactly where it originally was and not here. 

Other discussion of running with jumper's add-on/development builds already existed here as well. There is no perfect solution, and no reason to start an entire new thread devoted to what works or doesn't work with the development version; KEX has enough threads as it is. I agree that it could also fit very well in the thread where it was, but in the end it's jumper's project and jumper's thread. We can only hope that he will eventually reach a "stable" version that incorporates new successes without breaking older ones.

[MiKl]

@Jumper, I have already mentioned 'error 1723' a few posts up in this thread (Feb., 11th) !! But you don't have to look into this, please concentrate on KernelEx but it would be nice if I could get some feedback in the Auxiliary thread from you.

[DougB]

Latest KoolPlaya (DirectShow media player) v1.3.1.9ß (2014-Feb-02) is again working using updated KernelEx DLLs r11+r12 (no Kext/KStubs or ROS):

http://www.koolplaya.de/index.php

Download actual latest version:

http://www.koolplaya.de/download/beta/

Requires the kpsamuplugin.dll file from the parent directory:

http://www.koolplaya.de/download/

Note: Will play non-encrypted DVDs, but not encrypted ones.

- Doug B.

[halohalo]

English version of K15TK (AMD Kaveri APU tweaker) runs on KernelEx 4.5.2 with Kstub822. But I have to edit Kstub822.ini.

[Kernel32.dll]
ActivateActCtx=f2e
CreateActCtxW=t1e
DeactivateActCtx=f2e
ReleaseActCtx=f1

GetFileSizeEx=f2e
HeapSetInformation=f4e

The utility is the only one that can change the clock speed by limiting P-State range on Windows 98 SE. I also tried AmdMsrTweaker 1.1 on KernelEx 4.5.2, but it needs DOS prompt and Windows 98 SE always freezes and shows a black screen when I access the DOS prompt window.

CPU-Z 1.6.8, CrystalCPUID, HWiNFO32 5.2.2 can't show correct clock speed. But HWiNFO32 5.2.2 shows multiplier, voltage identification, and benchmark scores that can prove K15TK works.