pointertovoid

Hello dear readers! Ignoring nearly everything about websites, I'd like you to explain me why Google occults a website, or rather parts of it, in its search results. Raw elements: - Forum not listed, nor its "Sciences" section , but does list the "General discussions" section - Other search engines occult the forum as Google does - Said forum describes no drugs, explosives, misconducts... But it does contain my inventions (civilian uses) in the occulted sections - Other forums where I put a few inventions, and links to this Saposjoint, are listed by Google - Until recently, I didn't receive e-mails sent to myself containing an address of the Saposjoint website. Questions: - I browse from the semi-free country I happen to inhabit presently, called France, whose official policy is to censor some websites, under various excuses (explosives and so on). Could you please check Google's answers from your country? - Can the Saposjoint website be programmed (Php) in such a way that search engines occult some sections? Thank you so much! Marc Schaefer, aka Enthalpy, aka Pointertovoid

128MB is very little even for W2k. <1GHz would be acceptable, as I saw with a 800MHz PIII. Could you increase the Ram? W2k will always stay very slow on that capacity. If running alone, W2k gets its full speed at 256MB at is just 10% slower on 128MB - this is better than Xp, right. But as you pointed out, the antivirus takes a lot of Ram (Avast 4.8 free takes 120MB out of 512MB here; this version still gets signatures updates) and the firewall as well (20MB for Comodo v2.4.18.184). Then, Firefox 3.5.2 (more recent ones exist) crunches 100MB before I open many big pages. Avast has the advantage of being slept-in and awakened in two clicks without rebooting, and even selected parts of it: mail, Internet. Very useful for video games if you lack Ram. As well, W2k demands a fast hard disk drive - more so than Xp, whose prefetch is very efficient, and the smaller Win Me. ----- Knowing that, and after having used W2k on many old machines, I believe it will remain unusable on your Ram (and disk). So unless you want and can upgrade at least the Ram to 512MB, I would suggest: - Not to go on the Internet with that machine. Then, remove the antivirus and the firewall, which allows you to put the excellent Win Me on it. That will be real comfort. - Or replace the mechanical disk drive with one or two Compact Flash cards. That is, I suppose you have a single bay with a P-Ata interface. Cheap and thin adapters exist for two CF on a 44 pins P-Ata port. Their quality isn't decisive. The quality of the CF cards is all-important. They must be of type 1, and they must have a stable UDMA transmission, which means Transcend - though Lexar and Adata are not bad: they get one step slower rather often but lose no data. The card's transmission is the point, not the adapter as we users thought for a while. A CF1 card improves hugely a machine with a too small Ram because paging gets so much faster than on a mechanical drive - even compared with a good 7200rpm 3.5" drive. But only a type 1 is fast and will live long in this use. This works also for desktops with a decent disk drive: just paging on a CF1 accelerates all application switching. Very nice for i815 chipsets that limit the Ram to 512MB, which is getting narrow for W2k+security+browsers and is long too little for Xp. Then, the CF can also hold the scratch folder for the browsers. ----- Only SLC (single layer cell, type 1) Compact Flash are good here. MLC (type 2) don't work, as they lack the buffer that now makes MLF SSD acceptable. SLC is to be checked on each and every model, as sometimes bigger capacities of the same product line are MLC. Examples: - Transcend 266x from 2GB to 8GB - Transcend 300x from 4GB to 8GB, but the 16GB is slow and the 32GB even worse - Lexar 300x from 2GB to a bit more - Adata 266x 4GB or 8GB, but the 16GB seems much slower The key point to check is the throughput at writing small files like 4kB. Unless you're willing to spend much time and money on benchmarking everything after bringing it to work, don't experiment about it. Especially, all 400x and 600x speeds use MLC as far as I know. Examples right now are eBay's 230592803121 and 230592803260 (paste in their search window). Not Transcend, not big, but around $20 each when new, and used SLC are acceptable. Such a capacity isn't much... Enough for some laptops, or for the paging file of a desktop. Hence the dual adapter as well. My Bios doesn't see the the second CF, but Windows - starting with W95 - redetects all hardware and uses the second CF as soon as it has a driver for the P-Ata host, which may require you to provide this driver. Some users also take a bigger MLC as the second card and the SLC for Win, the paging file, and selected applications. P-Ata SSD are extremely bad or (MTron) extremely expensive. They look like a worse choice than Compact Flash with an adapter. Of course, you decide if this investment makes sense, as compared with a less old used laptop.

Despite all the admiration I have for the work on uSP5, I have to warn you. Microsoft created patches to improve the functionality, not only security, of W2k sp4. They were intended for an official sp5 which was cancelled on the way. Some of these patches were included in the Rollup1 (though Microsoft claims the R1 contains only security updates), others were not; Microsoft tells in essence that these latter patches "are not fully tested and should be used ONLY if you experience the faulty behaviour a patch improves". These patches are somewhat hidden at Microsoft's site and automatic updates don't load them. After having tried such patches created for USB, I can tell you that "not fully tested" means "seriously bugged": they DO introduce buggy behaviours and bluescreens. You really shouldn't install them without an acute need. Though, the uSP5 includes all these patches, so I expect W2k with uSP5 to have serious bugs and instabilities. Mine got nasty with the USB patches. ----- The two editions of the Rollup1 are nearly identical. Use the second one. None is the uSP5, whose "u" means Unofficial, as it doesn't come from Microsoft. ----- For the rest, I didn't test uSP5 by myself; it is designed to work exactly as a standard patch or service pack, and allow automatic updates. A tutorial exists. I believe to understand that uSP5's creator, Gurgelmeyer, passed over, rip, so you won't get first-hand information.

Adding to the installation disk a driver specific to one controller doesn't prevent installing on an other controller. You just have the advantage for the drivers you have added. As there are very few drivers that cover most of common controllers (say, 3 drivers cover 90% of all mobos) you can integrate these 3 drivers. I don't integrate drivers into my Win installation disks, as I can live with F6 diskettes. There are so many combinations (Bios config "compatibility", Sata, Ahci, Raid - then driver version - then Ahci, Raid during install - bugged disks - etc) that can go wrong! I prefer to keep flexibility here. Yes, Ahci brings experimentally an awful lot and must be used.

IF you only wanted to place existing keys on a different pattern, it would be for free. At least for keys of identical size: you can de-assemble the keyboard and re-assemble it with the keys at a different location. Then a driver or a creator of keyboard description would re-map the key positions, so you ket the letter that corresponds to the key's caption. One drawback: the Bios, and Ms-Dos, and possibly more programmes (bootable disk test, Ram test...) would map the keys according to the original keyboard identity, or to some default mapping - typically Microsoft Qwerty. By the way, this drawback will exist with any custom pattern, until your pattern is such a huge market hit that every OS and application knows it.

Hello nice people! The second generation of WD's Velociraptors, with 450GB and 600GB capacity, have more agile arms - though it's not advertised everywhere, their track-to-track hop is 1/3 faster as are short strokes. Nevertheless, neither noise nor consumption have increased since the first generation with 300GB capacity. This is a hint to a lighter arm. I couldn't read an explanation for this improvement. Can you? I had suggested electroforming to produce lighter arms of nickel, some µm thin, there: http://saposjoint.net/Forum/viewtopic.php?f=66&t=2160 and sent this idea to the manufacturers of hard disk drives, plus some more ideas, like the two-armed hard disk http://saposjoint.net/Forum/viewtopic.php?f=66&t=1793 I'd be happy if at least WD has used this idea of electroforming. Thanks!

Wear-out of writeable optical disks is not much a matter of time but of ultra-violet light. The UV light that writes the disk's thin data-holding layer at the laser also "overwrites" it. The brand can't help radically here. UV comes from Sunlight (it's powerful!) and from fluorescent bulbs as well, which includes the energy-saving lamps - they're nothing else than a fluorescent tube folded into a compact shape. Same story for plain plastics, which get destroyed mainly by UV, though the polycarbonate making the thickness of optical disks is rather stable to it. Much more observable at old polyethylene plastic bags for instance. So the basic precaution with writeable optical disks is to protect them from light, both before they're written and after, by keeping them all the time in an opaque box. You may also consider copying your data after 5 years for instance, instead of forgetting it. But the, why should your data stay on optical disks? A hard disk drive has only advantages for that use.

The "historic instability of Win Me" is probably a myth. At least here, I saw many unstable W98se and only very stable Win Me. Remember something like 3% of all users ever had a Win Me but everyone tells an opinion about it. I can give some hints about silent hardware, but not if W98 or Me can run on them, please check by yourself. The general idea is that you need a low power consumption so that the fans can be silent. This helps the power supply as well. Better: in Summer, your room won't heat as much. And it does make a difference on the electricity bill. For that purpose, a Socket A isn't the leader... And a Pentium 4 even worse! At the same computing power or better, a Core 2 or an Athlon 64 draws far less power than a P4 (remember the BUT with W98-Me). As well, some mobo chipsets may require a fan, so have a check, especially if you want an Amd Cpu. Depending on your graphics needs, you can choose a video card with passive cooling. Many 3D games are perfectly possible, but not all. Depending on your applications, you may prefer a dual or single core over a quad. Radical change in power consumption. Only a few video games and image-video editing applications use several cores in 2010. I feel the varied throttle modes useless, because noise at bootup or at full load disturbs me as well. Just as an example, I wanted an extremely silent computer as well. I took a Core 2 Duo E8600 @4GHz because no single core was available (Core i5 would be more recent) and a P45 chipset: they're still the fastest on single-task applications. I sold the Nv9500 because its fan was noisy at bootup and bought a passive Nv8600, perfect. The power supply is anonymous: it has a big fan. You can get a used 30GB Vertex for 70 euro. More important for system speed than everything else, and silent. Once you have an Ssd, you may tolerate a less agile mechanical disk as a complement, like a 7200.12, which is very silent.

Thank you! This is to my eyes the real proof that W2k does share the work among all cores. And nice performance! 64 bits: - Computing capacity doesn't change radically for most operations. The essential difference is the width of the addresses. The only computing hardware available is the SSE on a Core, with MMX and scalar operations emulated by the SSE hardware. - But the 64 bits mode accesses more registers, and this makes a difference if the code is compiled to take advantage of them, which is definitely the case of 7zip. Only 8 registers in 32 bits mode, needing to swap values with memory too often. - And the 64 bits mode can use the operations on 64 bits integer values. MMX and SSE operations remain the same. I expect that 7zip, which is optimized, uses SSE (=128 bits) wherever possible, and then the 64 bits operations bring little. - 20,000 Mips is far from twice 15,000 Mips. A small hint that 7zip uses SSE in both cases, the difference coming from the number of registers and a few added, more efficient instructions.

It is precisely a matter of size of the registry. The registry isn't corrupt. The worry is exactly linked with Xp's Ntldr and its known limitation.

Hello you all! I saw a few years ago a small free application that reduces the screen resolution and launches (typically) some old game that would display too small on a modern screen. Later, it restores the original resolution. Maybe Xp has it built-in, but I have W2k and stick to it. When I saw that application, I didn't download it, and now I miss it and don't find it again. Do you have clues? Name, address...? Thank you!

Hello everybody! I recently tried Pitaschio (thanks to Lenny_Nero!) and it looks interesting. It brings a huge amount of tweaks supposed to improve the use of Windows' interface, mainly of W2k. Though I've had some surprises or didn't find the functions I hoped in it. Maybe you have a more complete experience? - When Pitaschio runs, Paint Shop Pro 4 complains it doesn't find some entry point (indicated by its ordinal) in Mfc42.dll and won't go further, but works again as soon as I stop Pitaschio. Paint Shop Pro 6 doesn't complain. Bizarre bug, uncommon with W2k. - Explorer didn't actualize the display of folders any more, even after I stop Pitaschio. I had to restart the session. Same for the "save file as" dialog. What's that joke? How deep does Pitaschio fiddle with Explorer's or Windows' internals? - Can I maximize a window (both horizontally and vertically) using Pitaschio? I enjoy much doing it by clicking just once the scroll button with 4dmain, a stone-old mouse driver I try to replace with Pitaschio. Thank you!

Thanks Lenny_Nero! Tried Pitaschio, looks interesting, started an other topic about it. os2fan2: nothing to do with the hard disk, as the beginning of the thread shows.

The latest DirectX that installs on W2k is, officially as well as naturally, v9.0c. Though, some users succeeded in installing more recent dX versions on Win Xp. Has this been done on W2k?

Some software need a minimum OS version only for reasons of licence (sorry for being unseemly), others because they do call functions that appeared with the version they require, still others because they weren't tested will the older version. In the last cases, you may run into troubles that appear late and hence are difficult to link with one particular piece of software. This was especially frequent with W98 when users (or application installers) changed some Dll inappropriately. One precaution would be, using for instance FileAlyzer, http://www.safer-networking.org/en/download/index.html http://www.safer-networking.org/en/filealyzer/index.html to compare what functions your application calls, and which one are offered by your system. This holds for drivers as well! The test isn't perfect because some software calls different functions depending on the OS version it detects. But it's often precise. Such a hand check is slow, so I'd like it to be automated and described my wish in detail there: http://www.physforum.com/index.php?showtopic=24897 this checker would be very useful to me, a user of W95 and W2k.

And how many Mips do you get on 7-zip with the Core i5 760 Quad? http://sourceforge.net/projects/sevenzip/files/ Measured with v4.57 : - My PIII Tualatin @1650MHz brings 1050Mips - My Core 2 Duo E8600 @4000MHz brings about 7600Mips

Is this a bug of NtDetect 5.1 or of Ntldr 5.1, when double-booting with W2k? And why the heck was this never fixed? Sp4 for W2k was issued 2 years after Xp, so Xp could have been fixed! Or would there be a non-official fix?

I have found no "Matrix Storage" for "Nt 4.0" at http://downloadcenter.intel.com/ maybe they existed and were all removed, maybe not. UniAta was designed exactly for that purpose, and originally for Nt4. They didn't work immediately on W2k and I had BlackWingCat's alternative, so I didn't insist. http://www.msfn.org/board/uniata-driver-and-ich10r-t138573.html I understand they work only in IDE mode, not Raid nor Ahci - probably no Ncq then, which is a pity. Alternative solution: add a SiI3124 host on the Pci. This one has Sata3000, Raid, Ahci with Ncq (very important for performance) AND NT4 drivers including F6. On the ich2 Pci, throughput is >130MB/s. http://www.siliconimage.com/support/searchresults.aspx?pid=27&cat=3 3124_x86_win32_base_logo_13016.zip => Driver 1.3.0.16 for Nt4, no Raid, use with Bios 6.2.17 Raid or Base = 3124_x86_bios_6217.zip there http://www.siliconimage.com/support/searchresults.aspx?pid=27&cat=15 Got my used ones from pcland2008pcland2008 at eBay.com $10 +shipping $15 to Europe he had >10 pieces, sold 3, didn't relist them for sale. I would tell his eMail through the private mail here.

On the same W2k with the same DotNet 2.0sp2, SeaTools v1.1.0.3 for Windows works where 1.2.0.1 can't. There are very few versions between both. I found v1.1.0.3 there http://www.seagate.com/support/seatools/SeaToolsForWindows.msi

Quadraman's QuadraVex runs on my W2k with DotNet 2.0. Thank you so much, I really appreciate!

What about SeaTools for Windows? (You know, it tests Seagate's hard disk drives. Several simultaneously, and you can do something else on your computer at the same time, which are advantages over the Dos version I like). I tried to install v1.2.0.2 on W2k, though it officially requires Xp. Installation works but I get the error Not found ThrowWithExtendedInfo(System.Management.ManagementStatus) I have installed DotNet 1.1sp1 and 2.0sp2 (instead of 2.0sp1). Then I tried SeaTools v1.2.0.1 which was officially for 2k and got the same error. First, where should this ThrowWithExtendedInfo come from: Windows or DotNet? And what is the latest version of SeaTools running on W2k?

So, has somebody observed if W2k Pro loads more than 2 cores with real tasks? This isn't the same as displaying the cores in the Device Manager. By the way, last time I looked at the InfInst (probably a Core i7 in socket 1366), they were officially released for W2k. Only the chipset, especially the recent ich or pch were not - but we have BWC's driver, thanks.

I've had trouble over trouble with 98se but nearly none with Me, so my first answer would be Me. Then, (1) I had a Via chipset together with 98se, explaining most troubles (2) I was more experienced when I had Me... - Me starts faster than 98se, provided you have the minimum of Cpu+Ram - If tweaking little, Me is sooner usable than 98se. Comes with more recent drivers (Usb!), system components and add-ins (dX, ie and the like) - My Me were (and are) perfectly stable. All rumours against it sound odd to me. Remember you get both cheaply on eBay, like <10$ with a licence and a Cd.

Phenomic, it seems you succeeded to install a Usb 3.0 driver for the Nec chip with W2k... Could you tell us more? Here or there, as you prefer: The kind of questions your screenshot raises: - Where does the driver installer come from? - Was it the normal installation procedure? - Do you get Usb 3.0 performance? ------ Blackwingcat, I confirm what Syclone0044 saw: neither KB979683-v2 nor CoreChg.zip are accessible.

I certainly respect your favour to W2k for I make the same choice. W2k Pro using four cores would be new to me, based on what I read at Microsoft: W2k said to count each core (and even each hyperthread, as opposed to Xp) as a socket. So up to now I believed as well that W2k Server would be necessary to take advantage of a quad core - one reason why I have a Core 2 duo (other reason: my applications are single-thread). You also have fast dual cores among the Core i5 in socket 1156. I could also find a W2k Server for reasonable money on eBay. I sold it again because it shut down slowly, but other people here told me it shouldn't (other thread). However, I take good notice of other people's experience. Microsoft's claimed limits about its own outphased OS isn't perfectly reliable. For instance, MS tells W95 can't use disks over 32GB, but I installed one on 80GB for over six years that serves faithfully on a daily basis. A reliable way to know if all cores serve is 7-zip. It has a built-in benchmark that allows to run any number of threads and displays resulting Mips. 10% improvement means the same number of cores, 100% means twice as many. I have the best experience with Blackwingcat's drivers for W2k. I use the one for ich10r, it works easily with the usual F6 floppy procedure (as opposed to uniata which needs some doing) and provides Raid and Ahci - Ahci is vital for performance, both with a recent mechanical disk and an Ssd. He also has some W2k drivers for chipsets working with Amd Cpu (which slower than Intel these days, alas) as well as for recent graphics cards, but I didn't try them. Blackwingcat recently told his driver works (with his new Inf file, same Sys) on the Pch southbridge (other thread) for socket 1156. Could something go wrong with the no-execute bit of recent Cpu? W2k doesn't manage it, so one should disable it in the Bios I guess. There was also NtSwitch which "transformed" a W2k Pro into a "Server" (same for Nt4)... - I suppose it doesn't bring Active Directory and some more! - Would it change the maximum core count? - Legality is debatable... One guy on a forum claimed a judge in the US decided it is legal, but I didn't check. Fact is, NtSwitch is difficult to find. Please tell us! Whether W2k Pro uses more than two cores interests more people here!