[DrPizza]

What does it do that NT's own FAT lazy-write caching doesn't do?
What does it do that NT's own NTFS journelled lazy-write (meta-data is written immediately, data is lazy-written) caching doesn't do?

[DrPizza]

Quote:

Originally posted by AlecStaar
Try it and see!

No, because I don't know what it does, and I'm not in the habit of installing software whose purpose and mechanism is unknown.

Quote:

You'll see increased speeds on disks... just try it!

See, I'm not convinced that I will see increased speeds, because I don't know what it does. For all I know it might do nothing more than enable the unlimited file cache registry key (which I already have enabled).

Quote:

* If I told you? You could write your own!

Surely it can't be so trivially simple that a quick description of what it does that NT doesn't would be sufficient for me to write my own version?

Quote:

(The evidence is just you trying it and running benchmarks on it like WinTune97, SiSoftSandra, DrHardware, etc., standard tests alot of folks use is one way for you to see what it does!)

The only thing I'm interested is real-world performance.

[DrPizza]

Quote:

Originally posted by AlecStaar
Ok then, fine... do not try it! It's your loss maybe, you know? You're being 'close-minded' & all I am asking you is, try it and see!

No, I am asking for more information about the product. Before I install it, I want to know what it does, and I want to know why it's better than the built-in caching done by the OS.

Quote:

When I use it here, as I am now? I can see/feel my system going faster, as well as having benchmarks show a HUGE gain in diskspeeds!

Again, benchmarks are of no interested whatsoever. One can show pretty much anything with a benchmark.

Quote:

It's not that... NOT that simple!
It's not using the SessionManager, Memory Management, LargeSystemCache=1 setting registry hack if that's what you mean.
It's a driver Peterb/DrPizza. Tuneable & parameterizeable & creates its OWN memory cache that compliments the native one.

This already sets the alarm bells off. If it uses its own memory cache, that necessarily means depriving the system cache (because the system cache is free to grow as large as it likes; if this other program is using a large block of memory for its own cache, the system cache can't then use it).

Quote:

Again, do try it and see... I would wager you purchase it as I did & then developed part of its engines!

Again, outline what it does.

Quote:

DrPizza: Programs that place as finalists for 2 years in a row at Microsoft Tech-Ed? Where guys like yourself & myself are the ones looking stuff over, along with our bosses??

Speak for yourself. My boss wouldn't know one end of a computer from the other. I have no interest in award ceremonies (plenty of software I wouldn't care to use receives awards; they're a poor indication of quality).

Quote:

Are not getting that high up (and, in SQL Server categories with DBA's looking at them) for no reason! It also won BackOffice Mag's B.O.S.S. award, & Windows 2000 magazine (then Windows NT mag) in April 1997 gave it a clean bill of health & good review in 1997 April issue page 63! Read it yourself, the article is on their site... that was only the initial versions too, it has gotten better (Heck, I wrote parts that made it better, & helped it to do as well as it did at Tech-Ed).

It being good in 1997 doesn't mean it'll assist Windows 2000 at all.

Quote:

See, telling a guy as much as I told you now above? Gave him enough ideas to copy RamCharge on me... mistake on my part! But, what I did was original, & the first! That's good enough, & I made quite a bit of cash selling it since 1997, & still do! BUT, not as much... anyone can copy/steal really! Even improve on things... takes original thinking OUTSIDE THE BOX, to create! SuperCache-NT is such a creation & is winning awards for it AND showing its use on benchmarks & also real-world use!

I might be looking in the wrong place, but I can't find any real-world benchmarks on the web site, at all. Again, something that sets off the alarm bells.

Quote:

Business Rule PeterB/DrPizza: A good magician? Never gives away ALL of his tricks... I do hand out alot, why not? Helping folks out is 'good karma' and good for yourself too, good review! But, on programming mechanisms, no... lol, not on ones that have made me money! I did that once with memory mechanisms for RamCharge & told a guy ONE TINY CLUE, he turned around & wrote one like it!

Then there can't have been a great deal to it.

Quote:

I was a Windows Magazine 1997 Shareware of the year winner with that program... next year? He was... after using my idea! I learnt my lesson there! Nuff said...

I have no intention of copying the program. I just wish to read an explanation, even a vague one, of how it works (and, more importantly, how it speeds things up).

[DrPizza]

Quote:

Originally posted by AlecStaar
Look, you're a coder! I make speculations on designs of products with NO clues, and many times? I can design something of equal or better performance & aesthetics than an already existing product because I did not have to spend time on the initial idea designs... I can't give you that, & you KNOW it!

Er, why not? Bits of the thing are apparently patented; you can at least tell me what the patents are for (they're protected by patent, so I couldn't copy them even if I wanted to).

Quote:

True, but when I run 3 diff. ones (Dr.Hardware 2001, SiSoftSandra 2001, & WinTune97) on it?

That's a decent indicator, as 3 diff. tests concur with how my system responds, for ME, in the real-world!

I can tell just running it... try it, you will too. But, up to you!

This assumes that the three benchmarks have different testing methodologies. This would very much surprise me.

Quote:

Does not do that either!

Well, actually, it does. It can't help it. It permits me to turn off NT's caching (it claims), but it doesn't require me to. And if I don't do that, it will harm NT's cache, by taking memory that NT would otherwise use for its own cache.

Quote:

It grabs nothing from the normal cache, & does a better job of things, again... test it yourself, with anything you wish... including your own perceptions (best one) & you will see!

How?

It's a simple enough question.

Does it use a more advanced read-ahead algorithm?
Does it merely use a larger cache?
Does it allow the same tricks as NT's filecache does (for instance, files can be both in the file cache and in an application's address space simultaneously)?
Does it cache CDFS/UDF disks?
It says it lets you alter the page size -- but 4 kbyte pages are a hardware feature (Pentiums and perhaps others apparently have a 4 Mbyte page mode, but this is undocumented and requires an NDA to be signed to receive documentation). So what is it actually changing?

Quote:

Awards are indicators of quality man. When awarded by peers in the field, they are. That's like saying "The Matrix" is a bad film, but 1000's loved it. I guess it depends on your views.

Awards are not indicators of quality. They're indicators of the receipt of an award. No more, no less.

Quote:

It IS winning more acclaim now than it did then in other arenas where our peer judge it!

Try it yourself, test any way you like even, the best & only way, right?

Its documentation makes dubious claims, and what it's doing is a mystery. It fiddles with important parts of the OS, but doesn't explain how. So, no.

Quote:

You will see it does boost performance! I don't know why you resist that... you tried my software, and you say you won't use anything that is not "rock solid" but yet, you used my stuff (and I know a few ways you did alot more than once).

I couldn't get your stuff to even *install*. Let alone get a chance to use it.

Quote:

Look around it, they're there... you do have to dig! Not the best organized site I have ever seen! Your best bet, is to test it yourself though, what else can I say?

You can give me a URL to some real-world testing.

Quote:

You tested my APK Windows 2000 Tools stuff, I know this... & like it or not on your part, why resist testing some more of it then? That program is about 1/5 my work, the rest is the work of driver coders, & other C++ folks too.

APK Windows 2000 Tools didn't mess with anything important (it never got the chance; the installer bombed out persistently). This program, however, does. And until I get some indication as to /what/ it will change, I won't trust it.

Quote:

You're too clever I think for me to give you any insight to this, and I know you hang with JEH... HE DEFINITELY COULD ASSIST in your duplicating this! You work the top end GUI part, he work the driver.

I've never written a GUI in my life, and I have no desire to start now.

Quote:

Sorry man, no can tell! Just try it, how bad can it be right? Bad enough to be featured in Windows 2000 mag every month in ads for years now, & rated well by them too!

Placing adverts requires money, not quality. How bad can it be? It can crash my system on startup. It can corrupt data. So, pretty bad.

[pmistry]

I don't usually trust these types of programs. I remember some piece of lame software called Soft RAM 95 that promised to double your RAM under Windows 9x yeah right, it did nothing. Or what about those programs from symantec like WinTurbo, WinTune, SpeedWin, or other programs like Ispeed. man there all gimmicks.

[brainsoft]

URLs, benchmarks, something, ANYTHING would be more than you've really given them.

Hey guys, you should come to my house and test out my banana! I peeled it myself, and set it on my computer! It's making my Voodoo5 output 137FPS at 1600x1200, 32-bit colour, with 4xFSAA!!! I've tested it out with a wide array of syntetic benchmarks too, and they all agree that that banana is making my Voodoo5 a powerhouse of speed! Honest, I swear!!! I peeled it myself, i think I should know! I won't tell you HOW to peel the banana, or you may try it yourself, and then try to sell it!

Give them something to go on. How do you get paid for your work on the program? Do you get royalties, or did you receive a fixed amount. You're not an employee, so you're not getting paid that way.

If you're not getting paid for it anymore, and you're not under NDA you can tell us how it works.

If you wrote that stuff, and it's patiented, give us the patent numbers, perhaps someone would like to look them up. Some people are VERY edgey about what goes on their computer. It's not usually an issue, most software gives SOME idea of what it does. You are being VERY evasive about the whole matter though. They're not going to install something without knowing what it's doing. All they want is a general idea, they don't need exact details (atleast I don't). You aren't doing anything to belay they're fears!

If you were a friend of mine, one that I trusted, I'd try it out, but you're just some guy. I'd react the same way as they are if some guy came up to me on the street and told me he has magic poweder that can do wonderful things. I sure as hell wouldn't want to try it unless I knew exactly what it did, not just the result!

The ends don't always justify the means.


URLs, charts, details. If you wrote so much of it, then you should have a pretty good idea of how it works. And if you're not being paid, and aren't legally bound to silence, you have no reasons I can think of to NOT tell! (unless you play to rehash the same concepts to another company for more money perhaps?)

[brainsoft]

I don't actually care though, I ws just trying to give them a hand. I computer is fast enough for me as it is, thank-you. =)

[Amnesiac]

I wanna try it but your download links are dead...or at least they are for me. If I can get it downloaded i'll give it a shot.

-Amnesiac

[Amnesiac]

299.00 hmm

-Mac

[Amnesiac]

how about I just upgrade my computer with that 299 bucks lol

-Amnesiac

[Dirty Harry]

Somebody should invoice this guy for advertising.

H.

[snakefoot]

One could buy two 40 gb 7200 RPM IDE drives along with a IDE-Raid Controller for that kind of cash.

God Damn !

[zalem]

I cant find the supercache trial download for win2k only superspeed

[Brian Frank]

Not to rain on your parade, APK, but Systernals has some freeware stuff that may be a better option for many people.

I know for one, that's money I just spent on a couple of 60GXP's for my VP6:D

[zalem]

Well, i dont know if im stupid but i dont see supercache as trial download just ramdisk and superspeed!!!

[zalem]

[QUOTE]Originally posted by AlecStaar
[B]

I can email it to you if needed, I noted that link is broken again... the file is 1,743kb in size. Just over the size of a single floppy disk...

(Easily sent by email until they fix the link zalem)

Yes please, that would be nice!

[*Iblis*]

could you send it this way too? cheers