rloew

Thanks for the clear explanation. You say "it's possible not even a Pentium III may be enough (to play FP > FP 9.0.47.0 -- I added what's in parentheses). In *all* my comments about FP 9 and Pentium lll, I had *accepted* your and halohalo's assertions that Pentium lll *was* necessary and sufficient. And , *now*, you are backing off from this -- a little -- (now that's not very nice -- LOL!). First I discovered that Adobe lied to me, and I now discover that you might have "stretched the truth" a little. LOL ! Best not to play mind games with the feeble-minded -- right ? LOL ! You are being rather unfair. Neither Halohalo nor Dencorso said that a Pentium III was sufficient, only that it was necessary. You assumed, and I hoped, that it was sufficient. It is not very urgent, but an answer would be helpful. If the Pentium III doesn't work, then you are probably out of luck. If it does, then further research may be of value (see below). Although mostly about the Crusoe Transmeta, there were some references to Pentium IIs. It mentions the possibility that some SSE instructions may execute incorrectly without causing an "Illegal Instruction" fault. My Patcher cannot fix these, but it might be possible to find and Patch them manually or with a Program.

The file is not "educated", only the cached copy of it in RAM and/or Swap. It will be a "lame brain" as it was never modified and the cache will have been cleared. Since the copy is made from cache whenever possible, it would contain some or all of the Patches made. There are no guarantees as parts of the file may be in swap or no longer cached if space was needed for something else. I'm not sure this is going to help if the Patches are not a complete solution.

Hello, rloew, I just saw your Post # 61. As for the Pentium lll issue, I thought I'd summarize all the discussion about it -- in *this* thread -- to give you an overall view of the thoughts mentioned here. Halohalo, in Post # 2 said, "Since 9.0.115.0, Flash Player 9 requires SSE." As I pointed out in my Post # 3, "The first CPU to support SSE (is) the Pentium lll ...." So, halohalo's statement is equivalent to stating that "Since 9.0.115.0, Flash Player 9 requires Pentium lll (at a minimum)." Then halohalo (in Post # 4) and dencorso (in Post # 5) gave links to "prove" that Pentium ll is the *whole* and *complete* reason that FP 9 > 9.0.47.0 will not work. It was kind of like: "That's it. Case closed." I went through all of the threads listed by Halohalo and Dencorso. Only one post implied a complete fix and he was running XP and using FP 10. The others only said it no longer crashes, which we have achieved. Not necessarily simpler, but ones that do not use the problematic sections of code. Some posters in the listed threads mentioned Image size being a factor. The closer the better. I'm not sure of the range of speeds available, but there is no reason to assume it would not work with Windows 98. I already knew your system details. I have no information on minimum requirements, so there was no point in addressing your specific numbers.

The Patcher is totally dormant until an "Illegal Instruction" occurs so it would have no effect whatsoever normally. Copying the Patched files would only help if the Videos did not play properly the first time or maybe the second time but worked after that. Do these videos play with a Pentium III?

True enough. It doesn't always work. So, I might say it's complicated... But it's *not*, really. Your problem, as with most other cases where simply adding a MaxPhysPage and Xeno86's VCache is not enough, is probably due to your video cards. In other cases, there are also some LAN cards that use virtual addresses in the 3rd GiB, as well as some more modern motherboards that won't allow the simple tweak method to work. But for your average PIII/P4 (or Athlon) board with onboard video and LAN, it'll work in most cases. So the hardware that really forces one to use the RAM Limitation patch is reasonably recent, and was quite rare four years ago. And that's my main point. Moreover, even with forgiving hardware, more than 3 GiB RAM (also rare some years ago) leads to unstable configurations, with rare exceptions, unless one uses RLoew's patch. If you look at the 4 GiB section of the list you'll see 4 machines, two of which don't use RLoew's patch... these are some of those exceptions I've just mentioned. Don't get me wrong: myself, I'm a satisfied customer of RLoew's, and I think his is *the best* solution. But that does not make it the only one. And I think we're fortunate to have options. *And* I'm quite tired of people insisting in saying running with > 1 GiB is not *at all* possible. To dispel that urban myth is the main reason why I started the list, to begin with. The two issues I have seen so far, that require my Patch to resolve, are oversized Registries and Ethernet Drivers that hog the first 16MB of RAM. I added the /M Option to my Patch to deal with these issues. Also RAMDisks can be a problem especially when they exceed 512MB. I developed a Custom set of RAMDisks that do not have this limitation, and one can use memory above 4GiB.

The erratic behavior sounds like timing problems. Your processor may be too slow when combined with the newer Software. Performance would tend to improve when a Video is rerun as the Patches should have been already made and the Patcher is no longer invoked. But you had successful plays followed by unsuccessful ones. The Patches may be aggravating the problem as the original instructions probably were chosen to speed up performance on newer Processors. Since you are not getting any "Illegal Instruction" Errors, the Patcher cannot do more.

More like almost 7 years. My High Capacity Disk Patch was released in 12/03. My new Windows 98 Disk Patches will Boot from Hard Drives larger than 2TiB. XP cannot do this. Microsoft has already abandoned XP, so who is going to upgrade it?

An Invalid Page Fault occurs when an Instruction tries to access memory that is not present and the operating system does not expect it to access that memory. This is not related to "Illegal Instructions" that my Patcher was designed for. The problems you are seeing don't appear to be related to my Patcher or the identified issues between Pentium II and III. Has anyone tested these versions of Flash and browsers on a Pentium III or later? My Handle is RLOEW not RLEOW.

Actually I found that Windows 98 contains code that only comes into play above 2GB of RAM. At a later stage, they decided to shrink some internal tables to save space. This is what limits the amount of RAM. There was a bug, but it didn't come into play until 1.4GB. I have Patched Windows 98 so it can handle up to 4GB and written a SDK to support more.

You mean that after Win9x/ME caches a file, due to executing it, it'll never go back to the file (until reboot), so that if one uses flash intensively enough, so that all missing instructions get patched, and then copies the file, the resulting copy will be made from the cached copy, not from the actual stored file. It makes perfect sense, but I hadn't realized it'd work like this until you mentioned it. However if the machine is used intensively enough, with many other programs, without ever going back to flash, the cached file may end up flushed to make room for others, may it not? So the surest way to obtain a patched file would be to use flash intensively enough, and then copy the flash files right away, just to be on the safe side, right? It will go back to the File if the cached copy is discarded to make room for other stuff. Depending on usage and File Cache size, some or all of the Patches may or may not be there when the files are copied. It may even be necessary to copy the files while the Flash Video is running. It may also be necessary to repeat the process multiple times, and probably with multiple Videos to get all of the Patches saved. Of course there may be other problems that will not be fixed. The first problem larryb123456 found may not even be related to Pentium II issues.

Is this problem repeatable? Try playing this Video without my Patcher running. What happens at 2:37, assuming it gets that far?

Don't be too quick to uninstall the newer Flash if it doesn't work. I may be able to add more Patches quickly. If you still get "Illegal Instruction" Errors, open the details, write down the 16 Byte Code Trace at the bottom and post them.

Many of the things you are worrying about would have been listed in my Warnings if they could occur. Since this is a test, not a release version, I don't expect you to keep the Patcher running for any extended time so Programs such as Photoshop should not even be run. There may be more A: Although the files themselves are not modified, copies made from them may incorporate some or all of the Patches made if the File is or has been Executed. After the tests are complete, you may be able to take advantage of this to make the changes permanent so you will no longer need the Patcher running. B: Correct C: Correct D: No. KernelEx is a package of modified API's between Applications and the Kernel. It is not even Kernel code. My Patcher is incorporated into the Kernel itself and has only handles the 3 known Pentium III Instructions in the Flash code. E: Only 3 types of "invalid instructions". No other kinds of problems. F: Correct. G: Only Executables can be affected, never Bitmaps. H: If it did, my Patcher would Patch Photoshop itself, not the Bitmap File. You would also have to copy the affected Photoshop Executable or DLL to see the change.

#1 Yes. #2a Yes. #2b Yes as long as both are now identical. Otherwise copy SYSTEM.INI to another name. #3. Yes. Real DOS Mode can be run by holding down a SHIFT key during boot. Then just replace the SYSTEM.INI file with the backup. #4a. Yes. #4b. Yes. Only the Flash Code in RAM is affected. #4c. E-Mail. The file is tiny. #4d. Less than 1MB. #5a. It is a text file. You open it with a text editor such as EDIT or NOTEPAD. Then insert the line, save and exit. #5b. Yes. I would add the following line to the [386Enh] Section: DEVICE=P3CPU.VXD #5c. Either will disable it. You will need to replace the SYSTEM.INI, not just rename it. If you only change the VXD file, you will get warnings during reboot. #5d. Yes. #5e. Yes. #6a. Yes. #6b. A Hard Crash will lead to a Bad Shutdown. #7a. Windows 98 caches Program Files in RAM. The Code is supposed to be "Read Only" so Windows assumes that it represents the File Data. My Patcher runs in Kernel Mode so it can change the Code. If another Program reads the affected File, it will see the modified data and not the actual data on Hard Disk. #7b. Copying or Moving the File or Files that are currently triggering the "Illegal Instruction" errors. #7c. Yes. Don't do it while the Patcher is running, especially if you have opened a browser since the last reboot. #7d. It is running after the Computer is Booted with the File and SYSTEM.INI Line present. It is no longer running when disabled (see #5c) and the Computer is rebooted. #7e. Yes. At a later date it may be advantageous to take advantage of the File Cache to make the Patches permanent but this is for later. As to your earlier question as to why some Videos work with the newer Versions. The websites don't determine what instructions are executed on your Computer, but different Videos use different features of the Flash Software. Some portions of the Flash software use the problem instructions while others don't. So some websites might work but others like YouTube use the Flash features that cause the errors.

I downloaded it in April. The copy I got was corrupt so I didn't do anything with it. I downloaded it again earlier tonight, got an even shorter file, that was unuseable. I just downloaded it a third time and got a good copy. I will look at it. I tried the SVGA Drivers both in Normal Mode and in Safe Mode. Windows worked but DOS Boxes do not. I could run a Full-Screen DOS but got a Black Screen on exit.

My Patcher does all the work. All you have to do is drop one file into your SYSTEM folder, add one line to your SYSTEM.INI file and reboot. Then you run the video. By the time you get back from the hills, the video should say "THE END". In the C:\Windows\System folder -- correct ? What is the name of this drop-in file (so I can check out the extension, etc.) ? I have two System.ini files: in C:\Windows and in C:\Backup. By adding one line to the System.ini file, you just mean "rename" it -- right ? We'll just rename the .ini in C:\Windows and leave the one in C:\Backup alone -- right ? If you sense I might be a little overly cautious here, it's because this is my only computer and if it gets messed up, I will not be able to get another one. I know that the System.ini has to do with startup, etc. and specifies what programs are loaded at startup -- that is, if I understand Google correctly. I certainly don't understand the totality of the file or -- if the truth be known -- anything at all about it. Have you modified System.ini just so that it affects the browsers and Flash Player only -- (or I guess just for Flash Player since it's to fix a bug in it) ? I have many important programs that I can't afford to lose (like Photoshop, Corel Draw, Adobe Illustrator, and others, etc). All of my installation CD's and manuals, etc. were destroyed (in a sad, sad story that I won't go into here) -- so if the programs are uninstalled or corrupted, or if they just disappear, I'll be totally out of luck. I also need to be reassured that if the patch doesn't work -- (or if my other programs don't work as before) -- that Windows will open as usual so I can navigate thru my folders to undo the patch. To summarize, can you explain -- that is, if you don't mind -- in somewhat simplified terms -- how the patch works, and how it interacts with my Windows 98 system and other installed programs, like Photoshop, for example ? By the time I get back from the hills, I sure hope it's the video saying "THE END", and not my computer saying "THE END". LOL ! When I feel just a little bit more comfortable knowing the details of the patch and how it works, I will certainly volunteer. I don't like the word "test", because it causes me to wonder about the ramifications if the "test" fails. I don't think it will be much longer before I get the explanations and reassurances I need for me to test it. Dencorso, did you kind of double-check the code (if that's the right word) of the patch ? But right now, the patch is kind of like a "pig in a poke". In the newspaper not too long ago, I read a story about a guy who installed a "pig in a poke" into his computer. In no time at all, the pig grew to be ten feet tall and ate his computer. Then it ate his house and his car. Finally, for good measure, it ate his wife. Thanks for your efforts. BTW, all this is very interesting to me. larryb123456 It is not a Patch. As Dencorso surmised, it is a VXD that intercepts "Illegal Instructions" when they occur and Patches the code in RAM. It does not modify any files. The added line in SYSTEM.INI simply invokes the VXD at startup. There is no effect on any Program that is not already causing "Illegal Instruction" errors. You can always backup SYSTEM.INI by creating a differently named copy before editing it. The Patcher can be disabled by restoring the original SYSTEM.INI file. The VXD File can then be removed and the system rebooted. If the Patcher fails, your Flash software will crash as it did before. The Patcher can affect the File Cache, so do not copy or move the Flash files while the Patcher is running. Disable the Patcher and reboot first.

The 1TiB Problem occurs only with certain alignments of data so there is a 1 in 4 chance that a given Partition will have the problem. The problem appears when trying to read a Directory where the Directory itself is above the 1TiB limit.

By being a computer novice, certain words send up red flags, shut my brain down, and make me want to run for the hills. Among these words are patch, registry, hexeditor, KernelEx, etc. My Patcher does all the work. All you have to do is drop one file into your SYSTEM folder, add one line to your SYSTEM.INI file and reboot. Then you run the video. By the time you get back from the hills, the video should say "THE END".

I have completed testing on a new set of Patches that will support this Hard Drive when it becomes available as an Internal Drive. Using an Emulation Overlay, I was able to Partition, access and Boot From Sectors that appeared to be above 2TiB.

I have written the Dynamic Patcher. It Patches test code, but I don't have a suitable Flash setup to fully test it. There are no Klingon girls in my area. No Romulans either.

I downloaded it in April. The copy I got was corrupt so I didn't do anything with it. I downloaded it again earlier tonight, got an even shorter file, that was unuseable. I just downloaded it a third time and got a good copy. I will look at it.

I haven't seen a motherboard yet that forces DOS Compatability mode for IDE Hard Drives, at least PATA ones. I have a Patch for SATA.

Another option, that I might be able to write for Windows 9x, is an Interrupt hook that can intercept these instructions and replace them dynamically. I wrote one to run GOMPLAYER on an AMD-K6. It ran but was too slow as it could not replace the CMOV Instructions just interpret them. I eventually added a logger and used that to do patches.

Although I added the /A Option to my RAM Limitation Patch to support 512 MiB Video Cards based on my experience with an ATI Radeon Card, it neither helps nor hinders operation of NVidia Cards with 512MiB. The issue I observed with the ATI Driver, does not occur with the NVidia Drivers discussed on this forum. My Patch may help with ATI cards but it won't help with the NVidia cards.

Publicly available is not the same as Public Domain. Windows 98 itself was once Publicly available but never was Public Domain. These Patches are available to general or generic users. Only cheap users do not have access. I don't have a lot of PCI-E Motherboards to test, but the 7100 GS worked fine. A 7200 works but has the shutdown problem.