Mk 16 torpedo tactics?

[TorpX]

Originally Posted by WH4K

But since I "reinstalled" SH4 and applied the 4GB patch earlier today, I think the Tokyo Bay crashes are cured, so perhaps I'll give it another go.

What is the 4GB patch and what exactly does it do?

[WH4K]

Originally Posted by TorpX

What is the 4GB patch and what exactly does it do?

There really needs to be a sticky explaining the hows and whys of the "4GB patch." I think it applies to SH3 as well as SH4, although I haven't tried it with SH3 yet. I'll take a stab at explaining.

The "4GB patch" is a nifty utility written by a Daniel Pistelli, who offers it for download at his website, NTCore (http://www.ntcore.com/4gb_patch.php).

The patch allows 32-bit applications (which both SH3 and SH4 are) to use up to 4GB RAM when run on a 64-bit OS, such as Windows 7 or Vista 64-bit.

Formerly, 32-bit apps could only use 2GB max, regardless of how much was actually available.

It's a bit unclear what the 4GB patch does for me, because I only have 4GB total. After applying the patch, I did not see SH4's RAM use climb higher than about 1.6 GB. So I'm not sure what the deal is. The patch may not be responsible for the apparent increase I've seen in SH4's stability.

Many of the fan mods cause SH4 to use much more RAM than it was originally intended to require. So anything we can do to let it have more RAM may be of benefit. I'm just not sure how this applies in my case, since I only have 4GB RAM.

[paulhager]

If you go over to the SH4 forum, I've posted accounts about a number of long range shots. Most recently for Patrol 8 of CAPT Glenn Ford, using the 3D TDC radar mod, I actually got 2 hits at the amazing range of 13,200 yards on a ship. Since the range of the Mark 16 is 13,700, that's got to be about the max.

Later in the same patrol I sank a Yamato at a little over 4,000 yards with 4 Mark 16's. Unfortunately, a 3,300 yard, 6 torpedo salvo at the sister ship completely missed. I'm still scratching my head over that.

[WernherVonTrapp]

Originally Posted by WH4K

There really needs to be a sticky explaining the hows and whys of the "4GB patch." I think it applies to SH3 as well as SH4, although I haven't tried it with SH3 yet. I'll take a stab at explaining.

The "4GB patch" is a nifty utility written by a Daniel Pistelli, who offers it for download at his website, NTCore (http://www.ntcore.com/4gb_patch.php).

The patch allows 32-bit applications (which both SH3 and SH4 are) to use up to 4GB RAM when run on a 64-bit OS, such as Windows 7 or Vista 64-bit.

Formerly, 32-bit apps could only use 2GB max, regardless of how much was actually available.

It's a bit unclear what the 4GB patch does for me, because I only have 4GB total. After applying the patch, I did not see SH4's RAM use climb higher than about 1.6 GB. So I'm not sure what the deal is. The patch may not be responsible for the apparent increase I've seen in SH4's stability.

Many of the fan mods cause SH4 to use much more RAM than it was originally intended to require. So anything we can do to let it have more RAM may be of benefit. I'm just not sure how this applies in my case, since I only have 4GB RAM.

I've been doing some research on this 4GB patch. It seems there have been other, similar apps in the past, including an adjustment in Windows and mentioned on the Microsoft website. It would seem that the consensus of users claim no increase in performace. There are some who claim it helps but they seem to be more a product of the Placebo Effect. In fact, from what I've dug up, anything that purports to allow 4GB of memory in a 32 bit app (regardless of 32 or 64 bit systems) is questionable. There's also some info that these patches involve hacking the Windows Kernel and may leave your system open to exploits. I've also delved into some forums and read what the users of these type of patches write. Though a lot of them claim to have experienced no problems with these patches, they also claim either no noticeable improvement or haven't even bothered to check, benchmark, etc. In essence, they're using these patches blindly simply because someone created them, and/or, says they work.
The whole point to this is that one should thoroughly research any patches to be installed in their PC and hopefully, no one installs it simply on the strength of an unknown who posted the patch.

[TorpX]

Thanks for the warning, Wernher. It doesn't sound like something I want.

[WH4K]

The "warning" above sounds Chicken Little-ish.

I can't speak for other "4GB patches," but there is no evidence that Daniel Pistelli's (aka NTCore) 4GB patch does anything to the Windows 7 kernel. I don't understand exactly what his patch does, but I could say the same thing about the Windows 7 kernel itself, or SH4.exe.

I'm not sure that Windows 7 would allow kernel alterations without kicking up a big fuss, if at all. I have had no evidence of nefarious goings-on, such as MSE alerts, bluescreens, or unexplained network traffic. So the simpler explanation is that the NTCore patch is just what it says it is, not some very clever Trojan.

I had not yet run across claims of "improved performance" with the 4GB patch (NTCore or anyone else's). That does sound nuts. Why would anyone think the program would magically "run faster" because of a mysterious software trick? Placebo effect, indeed.

The last time I recall hearing "runs faster" claims, they regarded "memory manager" software, common in the 1990's MS-DOS era. There were several alleged "run faster" programs or utilities, all of which were powerless to increase your CPU's clock speed and did not magically give you more RAM.

[WernherVonTrapp]

Don't take it personal WH4K. My post wasn't directed toward you. It was a matter of fact statement based on 1 single hour of web research. It's not at all Chicken-Littleish. It's common sense. I mentioned that during my research, I found many forum members (on many forums) that claimed they experienced no problems with the patch. On the other hand, they claim that they haven't noticed any improvements or enhancements to their Windows system either. You have just reinforced this with your own claim.
From what I've read, and I'm not an expert mind you, you cannot allow a 32 bit program to use 4GB of memory (on a 32 or 64 bit system), without altering the Windows Kernel. It's not possible, from what I dug up. Any time you mess with the Windows Kernel, you potentially open up your system to an exploit or hack. Hacks occur every day without the PC users ever knowing that it took place. That's one of the reasons why it's called hacking.
I don't see why, if you do some research yourself, you wouldn't find the same things I did on-line. It wasn't an attack on you WH4K and I'm sorry if I came across that way since I sincerely had no intention of it appearing that way.
Here, I'm not an expert and I don't do programming but, on the same website where you got your 4GB patch from, there is some explanantion involving 32 & 64 bit apps and RAM utilization. Now, from what I can decipher, he mentions "call tos" or other references that sound like he's using or fooling the Windows Kernel. That, in essence, is an alteration of the original Kernel. Maybe I'm misunderstanding this but check it out for yourself:
http://www.ntcore.com/files/vista_x64.htm

Mind you, I had checked other websites also, that contain content about getting 32 bit apps to use 4GB of memory.

[WH4K]

Originally Posted by WernherVonTrapp

...On the other hand, they claim that they haven't noticed any improvements or enhancements to their Windows system either. You have just reinforced this with your own claim.

Not so fast. I have had an improvement: After the NTCore patch, SH4 doesn't crash nearly as much as it once did. Hardly ever, in fact.

Originally Posted by WernherVonTrapp

From what I've read, and I'm not an expert mind you, you cannot allow a 32 bit program to use 4GB of memory (on a 32 or 64 bit system), without altering the Windows Kernel. It's not possible, from what I dug up.

That's not what the article you linked says.

Originally Posted by WernherVonTrapp

Any time you mess with the Windows Kernel, you potentially open up your system to an exploit or hack. Hacks occur every day without the PC users ever knowing that it took place. That's one of the reasons why it's called hacking.

Any time you step outside, you potentially open yourself up to having a grand piano land on your head.

That doesn't mean it's reasonable to worry about it, especially when the thing (changes to the Windows kernel) isn't even happening.

This line of thought is a dead end.

Originally Posted by WernherVonTrapp

... Now, from what I can decipher, he mentions "call tos" or other references that sound like he's using or fooling the Windows Kernel. That, in essence, is an alteration of the original Kernel. Maybe I'm misunderstanding this but check it out for yourself:
http://www.ntcore.com/files/vista_x64.htm

There is not a single use of the term "call to" in that article.

Using features of the system as intended is not "altering" the kernel. You apparently skipped the section on Patch Guard. Go back and read it, and perhaps you'll understand.

Another part you must have skipped, under the "Windows on Windows" heading:

...32bit applications have a maximal 2GB space (4GB if explicitly required) and the rest of the space is handled by the system. This doesn't change much of course, since on x86 user mode applications had 2GB of virtual memory space out of 4GB (the other 2GB were reserved for kernel mode). On x64 these two other GB can now be accessed by 32bit applications. In order to achieve this, the IMAGE_FILE_LARGE_ADDRESS_AWARE flag has to be set in the File Header's Characteristics field. You can do this programmatically or manually with a normal PE editor like the CFF Explorer, just like this: (example truncated)

So we see it is entirely possible to let a 32-bit app access more than 2 GB RAM without "hacking" or in any way altering the Windows kernel. Even if it did try, Patch Guard would probably stop it.

I have seen no evidence that the NTCore 4GB patch does anything that could reasonably be regarded as harmful. The plural of anecdote is not "data."

Rather, the NTCore patch simply keeps SH4 from crashing where it once crashed constantly, regardless of whether SH4 uses more than 2 GB RAM. That is good enough for me.

No insult perceived or intended - I just think you are worrying needlessly, and perpetuating conclusions not supported by the facts.

[Jan Kyster]

Originally Posted by WernherVonTrapp

...There are some who claim it helps but they seem to be more a product of the Placebo Effect...

Erhm, but no. And it's not about performance issues either. Or NTkernel...

Have a look in this thread, post 14 and 15: http://www.subsim.com/radioroom/showthread.php?t=177503

SH4 can max. address 1.65 GB. Trying for more = CTD (as loading OM Museum).
A patched SH4.exe loads OM Museum fine and use 2.21GB...

A noticable find is that the memory overhead in SH4 is actually very small, 80-160MB, which you don't need that much traffic to use.
And I've feedbacks from many users that confirms their SH4 doesn't crash anymore after applying the patch.


The LAA-patching simply allows a 32-bit program to be aware of memory above the 2GB barrier. The patching itself is only a bit being set to "1" in SH4.exe

"This is an application that assists in making applications large address aware. When a 32-bit application is Large Address Aware, it can access up to 4 GiB on x64 operating systems and up to 3 GiB on x86." Source: http://www.techpowerup.com/forums/sh...d.php?t=112556

[WernherVonTrapp]

Well, then, maybe I am missing something here, if trying to get Windows to utilize it's full 4GB of memory (instead of just having it display 4GB) is not a performance issue. If the patch actually does stop SHIV from crashing, isn't that an improvement in performance to some degree? I understood what the programmer meant when he spoke of "Large Address Awareness". As far as the Kernel is concerned, that word pops up quite a bit on that web page, and I didn't only rely on the info found on his page. I was checking other similar apps and Microsoft's own website.
My reference to the Placebo Effect was an inference drawn upon the many users (in my searching) who claimed that they had the patch installed but didn't see any effect, improvements or enhancements to their gaming. In fact, many never checked their systems or had them benchmarked to compare the before/after effect on their PC. To me, that's tantamount to a placebo effect.
If one wants to install the patch blindly, that's their prerogative. I refer to the closing in my original reply, "The whole point to this is that one should thoroughly research any patches to be installed in their PC and hopefully, no one installs it simply on the strength of an unknown who posted the patch."
There was no malice intended in wanting users to beware of a patch that involved, from my searching, some serious questions.

Post Script: I knew I should've posted link references but there were so many and I was too lazy, and I thought that would've been too presumptuous.