Chozo4

I know it's a bit late but with RP7, disable KernelEX extensions for 'RunDLL32.exe' either the compatibility tab or through registry: REGEDIT4 [HKEY_LOCAL_MACHINE\Software\KernelEx\AppSettings\Flags] "*\\RunDLL32.exe"=dword:00000001 If for some reason that doesn't work, revert RunDLL32.exe to the stock version using the System File Checker. Only thing lost by doing so is access to the RP7 system properties info and use of the 'version spoofing' in the RP7-specific compatibility options tab.

Hm... interesting. Although the first one I'm not too sure on to be honest. The program based off it the author also wrote is 'mtate (madotate) which, while giving transparency to windows when 'tilted' to the side did not update the programs GDI display when dynamic content was involved (mIRC for instance) and so needed to be brought out of such a mode to update the current 'view'. Such a demonstration of the 'transparency' in that way leaves room for concern if that's how it's handled normally throgh the libraries. If implimented in some way on a 'window manager' based level as content won't update as a result if the side-effect is the same as in the use of madotate. ------ The second one I just tried using the given demonstration. As with the other one, it updates based on actions upon it by removing the window off the desktop, capturing the desktop, then putting back and then applying the effects. This is evident by the small 'blink' everytime you move it. this is also evident by you moving it over a window or other application then moving the window or application out from underneath. As a result you have the image still plastered on the 'transparent' item. This application also has the same effect of not updating dynamically with a constantly updating app underneath.

Small issue to note with the Network status monitor... Object: Network Status Monitor Issue: Error message every connection update Cause: Occurs when there are no connections to pull up from the TcpTables

I know it's slightly older of a topic but didn't see the point in making a whole new one as I'd thought this somewhat relevant. I've tried the methods listed here but couldn't seem to copy certain files in the japanese language character set. Changing the Active Code Page (ACP) to 1251 did not seem to rectify the issue, nor did changing the ACP to 932 (Japanese NLS). However, after a lil fiddling around - I'd noticed the key 'OEMCP' and matched the ACP to that (CP_432.nls). Needless to say, I was then able to copy over these files using the japanese subset even though most of the characters were listed by explorer as other random characters. The only side effect I'd noticed offhand were those items named using the ALT+0160 character (for 'label-less' icons) had their name changed to an 'a' or were simply accessable. Changing the names using the old character set rectified the issue. Then again, alt+0160 shouldn't have been used anyway as some disk scanners do not like that due to lack of dos accessibility. Better solution for this would be to make real label-less icons using the CLSID and Desktop Namespace sections of the registry.

Ah, just installed this one and it now looks polished and is very stable besides. There are no rendering issues I can see under 32-bit mode and they all display perfect under an NVidia Card. Quite a fine piece of work Revolutions Pack is and has become. I look forward to any future development with this and will continue to monitor for any changes. Huge thumbs up in your direction, Tihiy.

That's as it should be. RP didn't support 32-bit icons in .ico, but when you had icons with 32-bit format only, it accepted them. Now, it won't work since 32-bit icon loader is separated, and it can't load from .ico. You'll have to wait. And yeah, rendering 32-bit icons with 16-bit shell icons is a Windows 98 shell bug really, since it can't switch icon list depth at runtime properly.AH! That explains it then, thanks for letting me know on that. They had worked perfect in in the last release of RP9 so had thought it was a broken feature in RP9. Good to know things changed alot in that regard. I don't mind the wait and am in no rush. Tango icon set bug, really. What bittorrent do you have? http://www.bittornado.com/ I use the Experimental version of the client.

Edit: I'd grayed out the issues that have been resolved and are no longer relevant to the current release of Revolutions Pack. They were not removed as I'd thought them best as a bug-fix history for references sakes. Alrighty, that's definately fixed with the scanline issue. However, the alpha blending issues with icons still seem to exist (but don't think it was the target for this release so I'm okay with that ^^). If it's any help, vid card on my system is an NVidia GeForce 6800Ultra 256mb. If it helps any - here is a small subset of the icons I use which are exibiting alpha blending issues in 32-bit mode. The alpha blending seems to be fine for these for example under 16-bit but not 32-bit. [ url removed ] http://hosting02.imagecross.com/image-host...onSet_32bit.jpg In 16-bit, there is a green background behind the tango icon in the RUN-Dialog. Additionally, the bittorrent icon shows nothing other than a plain box behind it. It is one of those icons that use a small icon for the centre 'document' picture instead of a full sized image. Additionally, since testing from yesterday - I'd like to say that RP9 has shown to be fully stable again as of v9.0.9. I've not yet had a single crash or lockup involving '<unknown>' or 'krnl386.exe' since then (so far ). RP9 is showing great promise in it's development and I'm proud to help stress test it along the way. I'd initially had my doubts with RP9 as I'd been an RP7 user all along (I'm stubborn as all heck) but I'm glad to have switched over. Keep up the wonderful work you're doing.

Edit: I'd grayed out the issues that have been resolved and are no longer relevant to the current release of Revolutions Pack. They were not removed as I'd thought them best as a bug-fix history for references sakes. Alrighty, working fine so far save for a couple more things to report immediately noticed on first installation of RP9.0.9. Issues ---(See Image)Icons now don't seem to be getting rendered right, oddly enough. Some get transparency, others don't, and others have a couple lines. Switching between 32-bit and 16-bit through RPConfig seem to switch the issues around. Repairing icon cache using TweakUI and Switching between depth modes in both Display Properties and RPConfig don't seem to rectify it. http://hosting02.imagecross.com/image-host...4IconIssues.jpg ---When restoring back from full screen, the restore window animation doesn't take effect (restore goes back to size and titlebar flickers alot during the supposed animation). That and the corners in skins such as 'Black Mesa' show the pink transparency colour upon restore. Most noticable with applications such as Internet Explorer, pressing F11 to switch modes twice. edit in: Side note, the lack of skinning happens on applications after changing between fullscreen+resolution change and window. Corrected --Ethereal no longer has issues with deleting pen objects. --Secondary monitor display in the Display Properties no longer show issues and have been fixed upon resolution slider change. I will note anything else that arises during testing. I hope I'm not being a royal pain with all these issues cropping up. >.>

Edit: I'd grayed out the issues that have been resolved and are no longer relevant to the current release of Revolutions Pack. They were not removed as I'd thought them best as a bug-fix history for references sakes. Which version exactly, and how can I make those console windows appear? I have 1gb memory. 256mb dedicated to a large tempfile / download ramdrive - swapfile placed in said ramdrive to consume any space not taken up by said temp files if / when the swap is ever called on. I've the conservative swap option on so it uses primary ram first. Alas, I've yet to breach that much of the physical ram for the swapfile to ever be used yet. For Ethereal, it's v0.10.0 exactly. The errors appear in a first command window immediately upon startup of the program and each instance of the same 'delete object failed' occurs everytime any buttons or menus are used and any redraw over the application (moving another window over it). The second occurs on starting a packet capture - resulting in two windows simultaniously producing GDK errors but of the same nature. Download link for it is at the following: ftp://ftp.neolink.ru/pub/ethereal/ethereal-setup-0.10.0.exe This version was the version recommended for use by the ethereal page itself in it's documentation in the event v0.99 did not work under windows 98 (Which in fact, didn't at all).

Edit: I'd grayed out the issues that have been resolved and are no longer relevant to the current release of Revolutions Pack. They were not removed as I'd thought them best as a bug-fix history for references sakes. Is it possible to get Dr. Watson diagnosis? I tried but I cannot catch that one. For one reason or another Dr. Watson even bombshells over it in an endless loop - completely locking up the system after about 20 or so instances in a matter of seconds (each time trying to log its' own crash in tandem). The only logs I can get were involving one of the Game Clients (Mabinogi) I run on a daily basis involving two dll's that seem to crash in response to '<unknown>' due to RP9 (which is modifying user.exe?). WatsonLogs.rar

Hmm... decided to give winamp 5.x another kick around - see how it works out with KEx4. Alas, it works fine other than an age old issue since even KEx3. After either playing for a few songs or entering the options - winamp shoots up to a full 90~100% CPU usage from normal 1~5%. Issue Report: O/S: Windows 98SE v4.10.2222 KEx Ver: KernelEx Core v4.0.12 App: NullSoft Winamp v5.551 Descript: CPU usage ramps up from 1~5% to 90~100% during normal use or after options access. Reproduce: Run winamp, play an mp3 or so and open the options menu, close options, then take note of the CPU use. Work Pre-KEx?: No, requires KernelEx to function. Notes: After install, remove Gen_ml.dll and Gen_Jumpex.dll from the plugins directory before running as these crash winamp (On load and on configuration saving respectively). During the 90%~100% cpu issue, the cpu use seems to go down to normal if you have a sub-menu open from the options screen (such as a plugin config menu) temporarily. However, winamp returns to 90%~100% after closing the options submenu's. It seems to be but a single (the main?) thread causing the winamp proccess to completely max out the CPU according to WinTop. Additionally, pausing or stopping playback does not lower the max CPU use of winamp's main thread.

Edit: I'd grayed out the issues that have been resolved and are no longer relevant to the current release of Revolutions Pack. They were not removed as I'd thought them best as a bug-fix history for references sakes. Issue #4 {Resolved as of RP9.0.9} Been getting back to my general routine with everyday things using RP9. However, more things seemingly are cropping up. After a while I'll start to get oddball General Protection Faults and Invalid Page Faults in random applications. Be they RSRCMTR.exe or other programs such as winamp, trillian, etc. I can ignore them and restart the apps but after a while it will happen again followed by a complete crash to a blank, empty, white 16bit system error message immediately after a half missing Illegal Operation notice. These seemingly have 2 Things in common though. The Modules they fault in are most often in KRNL386.EXE and lesser often in <unknown>. Disabling neither GDI or USER resource salvation prevent it so it may not be linked to that. I do not experience them without RP9 otherwise. I'd love to provide steps for reproduction but alas, this issue seems almost completely random and cannot be reproduced easily. For the sake of it, here is one of the errors I'd caught. RSRCMTR caused a general protection fault in module KRNL386.EXE at 0001:0000b0d3. Registers: EAX=001a409c CS=0167 EIP=0000b0d3 EFLGS=00000212 EBX=821a7000 SS=2c37 ESP=00008b22 EBP=00008b2a ECX=0002000c DS=066f ESI=00020000 FS=0000 EDX=7e07bf60 ES=2c37 EDI=00008b38 GS=2d47 Bytes at CS:EIP: 67 66 8b 02 66 25 fc ff ff 0f 66 26 01 45 0c 66 Stack dump: 0035018f 066f0000 662f8b56 066f04ff 2c378b38 001816cf 00000000 00000020 0be40020 409c001b 3e3e001a 353a0000 1c10f90f 35bb8b66 00001767 39bf0000 Issue #5 /Cosmetic/ {Resolved as of RP9.0.9} During testing, I'd noticed that disabling User Resource Salvation also disabled desktop icon label transparency regardless of it's setting. Unsure if this was by design or a bug. However, it's not disruptive at all as a workaround is using transparency.exe to cover the labels if the need to ever disable URS arises. Issue #6 /Minor Annoyance/ {Resolved as of RP9.0.9} Using 'Ethereal 0.10', installation of RP9 causes it to produce a constant stream of the same error over and over in multiple Console Windows. (ethereal.exe:4286670927): Gdk-WARNING **: gdkgc-win32.c:release_pen:885: Delete Object failed (ethereal.exe:4286670927): Gdk-WARNING **: gdkgc-win32.c:release_pen:885: Delete Object failed (ethereal.exe:4286670927): Gdk-WARNING **: gdkgc-win32.c:release_pen:885: Delete Object failed (ethereal.exe:4286670927): Gdk-WARNING **: gdkgc-win32.c:release_pen:885: Delete Object failed (ethereal.exe:4286670927): Gdk-WARNING **: gdkgc-win32.c:release_pen:885: Delete Object failed (ethereal.exe:4286670927): Gdk-WARNING **: gdkgc-win32.c:release_pen:885: Delete Object failed (ethereal.exe:4286670927): Gdk-WARNING **: gdkgc-win32.c:release_pen:885: Delete Object failed (ethereal.exe:4286670927): Gdk-WARNING **: gdkgc-win32.c:release_pen:885: Delete Object failed Not quite for me, maybe because i'm using GDI from Me. Anyway, this bug is a mere cache malfunction which i corrected; it shouldn't ever show in real apps (i hope), so wait a bit for fix. I'm busy with fixing DIBENG misfortunes leading for icons-with-non-standard-palette corruption, and they're tough ones. This is 98 shell specific behavior; as with overstretch bugs, i can't really fix it. Alrighty, no problem at all. Just bringing up any issues I come across that I've not seen already mentioned. Glad to know those were known of already at least. -------- Edit in: I have a small feature reimplementation request that I miss. Is there any possibility to re-implement the ability to disable skinning on a Per-App basis like back in RP7 or is it still there but hidden now?

Are you running KEx4 RC2? Also, do note that only the firefox / netscape version will work. The ActiveX version will not (flashplayer_10_ax.exe or something to that effect). WHat I find also worked (might be in this case) was changing the OS version in the other reg path too. The one on the second post indicates: [HKEY_LOCAL_MACHINE\Software\Microsoft\Windows NT\CurrentVersion] Instead, using regedit, navigate to: [HKEY_LOCAL_MACHINE\Software\Microsoft\Windows\CurrentVersion] .. and change the key 'VersionNumber' to something like 5.10.2222, making sure to keep note what your version said before changes so as to set it back to the correct one after install.

Edit: I'd grayed out the issues that have been resolved and are no longer relevant to the current release of Revolutions Pack. They were not removed as I'd thought them best as a bug-fix history for references sakes. Just tried RP9 today on 2 fresh installs. It's pretty well done - although I've found a rather crippling GDI related issue on my end. Installed components: -DirectX9 -Unofficial ServicePack v3.0beta3 (Other install was v2.1a) -Nvidia 6800Ultra, Driver v82.16 -IE6sp1 (Other install was IE5.5) -Required DIBENG.dll and other files needed as requested and indicated for RP9 Issue #1 {Resolved as of RP9.0.9} Symptoms: -Complete curruption of the GDI layer. Icons garble, Hardward acceloration seems dead (Rendering slows to a crawl), menus, windows, etc cause garbage left behind (similar to an explorar grash due to no resources but the resources are all over 80%) Reproduction (on my end): - (*) open an explorer window to 'My Computer' (or any folder) - Open display Properties and select the Settings tab - Move the resolution slider back and forth *=If you disable GDI Resource Salvation, you can skip this step altogether. At this point if you have two shown monitors in the settings display, the second one will look garbled but then the first one will follow suit. Once the first one garbles the GDI is completely corrupted rendering wise and artifacting/garbage ensues. I would have provided a screenshot of the end result but any and all screenshots taken resulted in a pure-black screenshot image. If you do not open 'my computer' or such similar windows and just use the display properties resolution slider only, you will see the second display corrupt but no GDI corruption occurs. http://hosting02.imagecross.com/image-host...condmoncurr.jpg Issue #2 {Unrelated with RP9 - is a shell restriction} Symptoms: - Taskbar size is not restored when changing from 'Watercolor EB_Superbar' theme to a non 'LargeBar' type of theme. Reproduction: Select 'Watercolor EB_Superbar' then apply. Then select a non-largebar theme such as 'Watercolor EB' or even 'Royale_Small'. Continuing to select a superbar theme afterward during change continues to make the taskbar larger and larger. SuperSUPER-bar: http://hosting02.imagecross.com/image-host...persuperbar.jpg Issue #3 {Resolved as of RP9.0.9} Symptoms: -Access Violation (RPCONFIG.EXE) - Minor, does not interrupt functionality. [Access violation at address 00440180 in module 'RPCONFIG.EXE'. Read of address FFFFFFFF.] Reproduction: Select the [X] to remove a wallpaper image in the list without a wallpaper selected. If there is any more information you need which I can provide you - let me know and I'll provide what I can.

Yes, I know I'm quoting myself here.. heh. I'd solved this issue after checking the differences in CORE.INI between RC1 and RC2. Core.ini from RC2 has the GetVersion strings removed under the Configuration names. For Trillian Astra to function as it did in RC1, Re-adding 'KERNEL32.GetVersionExW=std' under [DCFG1.names.98] resolves the issue.

Have you tried using the 'Nightly Tester Tools' for firefox 3 to see if those two plugins really, in fact, will not work? The 'Nightly Tester Tools' overrides the compatibility version block and allows you to force older plugins to activate. At worse if it doesn't work afterward you can just reinstall firefox 2 (or a backup in the event you did that beforehand with your ff2 install).

Installed fine, everything seems alright and the changes stated (firefox, etc) are fine. However, Trillian Astra is broken again whereas paths are incorrectly registered, calls user profiles in its' own folder (should be %windir%/application data/trillian), sounds no longer work, and crash logs are no longer saveable (broken path). I greatly appreciate the effort you have put forth into this Xeno. Both you and Tihiy are doing a wonderful job for this community with KernelEx amongst others.

So far so good with KernelEx v4.0-RC1. Haven't tested everything overall but I've noticed that a few programs were spitting out missing file notices after KernelEx update, Namely DWMAPI.DLL and MSWEBAPI.DLL. Downloading those files and putting them in their proper places resolved my issues with that. To add a program better functioning: --Trillian ASTRA +Now correctly loads/saves certain configuration files and user data in their proper places (%windir%/Application data/Trillian) rather than in it's own install folder. +Notification sounds are finally working. +Crash dumps are now saved properly instead of the jumbled path '?<?rogram Files\trillian' Issues encountered: File: WINHELP32.EXE (v4.10.0.1998 - Stock Win98se) Error: Looping of 'WINHLP32 caused an invalid page fault in module <unknown> at 0000:80000001'. Resolution: Check in 'Disable KernelEx extentions for this program' under compatibility tab. File: RUNDLL32.EXE (v5.1.0.2600 - Revolutions Pack 7) Error: 'RUNDLL32 caused an invalid page fault in module <unknown> at 0000:80000001'. Resolution: Restore stock RunDLL32.exe from installation CABS/CD using System File Checker. Notes: This will disable the functionality of changing windows version in the first compatibility tab (RP7). Use the KEX compatibility tab (second one) instead for that purpose. Have you tried checking in 'Disable KernelEx extentions for this program' in the compatibility tab and seeing if that resolves it? Note that this reply underwent heavy edits during some intial testing

Cannie, I've been wondering the same if it's the card thats damaged but to function normally in all 3d and 2d applications aside from 16-bit console use makes me wonder it may not be the card at all. I have an mx4000 and fx5700 but both of them are too great of a performance drop difference and prevent me from playing any of the games I do on a near daily basis. MX4000-128mb: Very overclockable, runs smoothly on lower end stuff, no pixel or vertex shader capabilities FX5700-64mb:Hardly overclockable, budget line version, runs choppy and poor, basic vertex+pixel shader capabilities I might as well buy a whle new mainboard and see if that has anything to do with it in the near future. If it doesn't, money wasted, otherwise a new vidcard to follow. Thanks for the help though.

Are you able to copy+paste the error at all, either through clicking details button or even the 'view details' link on the 'sorry for inconvinience' type error box? Such information would be very desirable to help fix the issue in kernelEx, additionally posting the error in the KernelEx thread as well could be helpful. Also, are you using Revolutions pack along with it? If so, few programs with KEX (the versions that started including further unicode support) and RP7.0 are slightly more buggy and require either compatibility mode skinning or skinning disabled (in the programs properties) off for it to function properly. Being that I don't use Messenger, but rather Trillian[Astra].. I'm not of much help here otherwise.

I don't think it's for WinXP only considering it's been supported on win98 since the earlier 6x.xx verson of the drivers. Wouldn't have bought it otherwise, although in this case - used. *shrugs* If only it was as easy as using newer drivers but the issue happens irregardless of driver version. The widewin9x ones (v82.69) are also a lil buggy as they seem to cause errors in nvopengl.dll under certain SUNJava Applications that use OGL through the browser on my end (been using 82.16 because of the such since). Thanks for the input though, anything is appreciated.

Well, I'm officially about had it really and am incredibly tired from and over it. I apologize if this post seems incredibly mixed up). Tried for the last couple months trying to figure this out without having to fork over cash to find out. The issue I'd already gone over in the following thread in Guru3D , incredibly helpful bunch but still couldn't solve it. I'd gone there first assuming it was purely a graphic card issue (and may still be, I'm not sure anymore ). http://forums.guru3d.com/showthread.php?p=2931325 Everything is pretty much discussed there but for the sake of not having to type it all out again/paste the whole thread worth, I'd linked to it. In a nutshell: New vidcard, GeForce 6800Ultra 256mb, EPoX 9nda3-j, using the command prompt causes -- System wide lockup for a split second during load -- Occasionally, will randomly cause the screen to 'blank out' as if re-forcing a new resolution -- May further cause the screen to completely blank out for 60+ seconds, during this time all actions such as sound studder to split second sound and stop every 15-20 secs, over and over till display resumes then everything is back to normal. -- If done during while hardware acceleration is active (3d game such as unreal tournament), the display may completely corrupt all together. There is absolutely no issues with anything else, ONLY the command prompt and any applications linked to it which is really miffing me. I can't even do my web development (or even serve my pages) anymore on this system due to the PHP interpreter using the command prompt in the background (well, WINOLDAP at least). Killing explorer and running the prompt causes the display to not resume at all after the random blankout happens. Seems explorer is re-initializing the gui as a safeguard?. Running a dos screen session for extended periods (a fullscreen game such as STRIFE) and closing may cause the display to 'blankout' for split seconds every 45-60 minutes on its own but without any audible system lockups while playing audio. Apimonitor shows after every instance, getclientrect() is called, etc. It blanks out upon GUI refreshes it seems (stops while using the task list using ctrl+alt+delete but occurs during scroll then resumes after closing it), etc. I've tried: -- Reinstalling the OS (Win98se) 20+ times, all fresh settings (due to the next 3 methods tried) -- Tried it with MinDows(gfx drivers and not) and Windows ME (full updates and not) -- Different settings, different updates -- no drivers but gfx card -- no nvidia gfx drivers at all (only time the issues vanish is without gfx drivers) -- Butchering and changing Bios settings -- Flashing to and trying all Mainboard BIOS versions available -- Flashing and trying multiple BIOS for the 6800u (BFG, Leadtek, etc) -- Flashing bios after changing and normalizing voltages -- Underclock and standard clock on both mem,cpu, vidcard -- Ran Memtest+ overnight, 50+ passes, no errors -- Ran a GFX Video memory stress tester, 312+ passes, no errors. -- Trying most nvidia gfx driver versions, starting from 7x.xx to latest -- Trying different driver versions for mainboard chipset (Nforce3) -- Changing GDI.exe/gdi32.dll to the one from windows 95 and updated winme -- Different memory managers (himemx.dll, xmgr.sys, etc) -- Changing the maxpage settings in system.ini for vcache -- Tried USP2.1/3.0a,b/ME->98/RP7/etc -- Taking out one of my 512mb memory sticks to physically limit ram to 512mb. -- Bought a brand new PSU off newegg for $48.00 ......Issue still consistantly occurs. Seemingly, i can stop it for a while after fiddling with the resolutions/depth back and forth on the second screen (extended desktop) despite no extra monitor hooked up. After several aggressive resolution changes on that non-existant monitor and trying the prompt, the issue of blanking out stops (but not the lockups during load) untill you push it with too many console windows at once (during testing such as 20+) cause a BSOD in VCOND.VXD. Also returns during another display res change or refresh. Perhaps this which may indicate the card / drivers having issues with the dual-DVI monitor support on it? All in all, I'm frustrated, feeling defeated, the whole nine yards. I really don't want to shell out for another GFX card and/or Mainboard just to find out what the cause is. It's enough that I bought a new PSU which didn't fix the issue (despite it being the suggestion both here in the hardware forum and on the Guru3D forum). As I think back, there was the occasional issue when using even my GeForce mx4000 128mb. The command console would randomly cause a fault in address 0000:009b on its own (or something incredibly similar, all zero save for the last two bytes). So the question remains: mainboard/gfx card incompatibility or windows software issue? Any suggestions to fix, or even any experiences regarding console instability? Past discussion is in the link at the top for references sakes. Also, has anybody had ANY issues using an AMD 939 socket mainboard regarding such issues as the system stall during winoldap/prompt load? If you have used a socket 939 AMD mainboard and not ever had such issues, what mainboard was it and how robust is the bios as far as clocking features are concerned?

Have you tried Sysinternals 'FileMon' and does it fit the needs you seek? http://technet.microsoft.com/en-us/sysinte...s/bb896642.aspx I've found it invaluable to discover what file operations are going on. Be it file open, read, seek, write, close, etc to help discover which applications need to have their settings ajusted to 'tone down' on file accesses from the disk (generic and slow thumbdrive in my case). Perhaps the Freeware 'System Information Utility' by Juan M. Aguirregabiria? http://asakura.t35.com/?pg=file (Look for sinfo9x.rar)* *Note: I compressed and supplied a trimmed out version of it (help files removed, etc) which I use because the authors page is gone(?) and the only site that hosts a copy of it provides the nt/2k/xp version of it only. So as far as I know there is no other way to currently obtain the win9x version of it. .. or finally, Iarsn System Software - TaskInfo: http://iarsn.com/taskinfo.html

I've always stuck with using Juno for email services. Coupled with using an external mail manager such as Mozilla Firebird or any other client of your liking.. and it's pretty good. Been using the same service for nearly 12 years without any complaints whatsoever.

As long as the CPU you intend to use is backward compatible with 32-bit applications (meaning is has x86-64 extentions), there should be no problems. From personal experience, it runs fine for example on an AMD Athlon 64 4000+ (SanDiego-Core).