CLASYS

Umm... I'll say that XP is superior anyday. Windows98 still uses shared memory. This means that the following code would bring down the entire computer: while(1) { *( (char *) random() ) = 62; } If you don't understand C/C++, the code basically picks a random byte of memory and assigns the value 62 to it... forever, until the program is killed or the computer crashes. Chances are you'd never be able to react quickly enough to stop it though. The driver support in Win98 is horrible as well... ever tried using a simple USB flash drive in Win98? The "plug-and-play" device requires drivers... And finally... try leaving your Win98 machine running for more than a week... it'll slow to a crawl in no time flat. If you're saying that Win98 is better than XP because scanning with Spybot is quicker in 98, then you've seriously gotta read up on how things work! <{POST_SNAPBACK}> Gee, I guess you are right, XP is better because it allows as***** programs designed to do nothing useful to run :-) Gee, I guess XP is so good, that it also needed drivers before something that became part of SP1 was applied that it couldn't even recognize my old USB 1.1 Belkin Hub that 98SE never had a problem with even out-of-the-CD. As far as 98SE systems slowing down, I have no idea what he's talking about. Perhaps he is confusing this with XP systems that sometimes mysteriously slow down to NOT MOVING AT ALL with the BSOD. Long before 98SE got patched, we were all occasionally victimized by memory leaks, something that can still happen with XP. Today, 98SE has no such problem; I had a system up since 2004 I had to take down to install the SP. It's running now as a toy-version file server on a Pentium III-600 with 128 MB of memory. cjl (who uses 9x and XP where appropriate; no panaceas exist)

Actually Windows 98 was 16 bit, and emulated 32bit, which is why Win 98 was still dos based. Windows 3.1 was 8 bit, 3.11 for workgroups was 16 bit. I bet none of you have ever used Windows version 1, it was a 4 bit OS, with B&W GUI. Oh crap, I just revealed my age <{POST_SNAPBACK}> I hope this post was TOTALLY in jest, as it is totally factually incorrect. in Win9x, DOS loads the kernel files of Win9x, then largely disappears as it is replaced by real 32-bit drivers for just about anything. One of the nice things is that it is NOT necessary to replace everything, thus some 16-bit apps can still run. Win31S is an overlay for Win 3.1x that enables a subset of 32-bit mode that is indeed emulated. This allows a few real 32-bit apps, such as freecell.exe to run in Win 3.1x. None of this is relevant to 9x. Win 3.11 is indeed the precursor to Win95, and it has been shown to be 60% binary identical to Win95. The reason is that unlike Win 3.10, some of it was upgraded to 32-bit mode. In Win9x, more is so upgraded, but still it allows backward compatibility if you need it, something that XP cannot do. To give credit where due, XP adds some support for quirky 32-bit applications that grew out of 16-bit conventions in terms of .ini files, as opposed to 32-bit environment issues per se. In particular, some early 32-bit apps continued to use win.ini and system.ini etc., just as they did in 16-bit Windows. Thus, while these apps cannot run in 16-bit Windows, they cannot run in a 32-bit Windows without these trappings, and XP is prepared to accomodate such as this better than previous NT family offerings. Of course, if all settings are in the registry, all of this is moot, etc. cjl (who, to show his age, is known for using 12-bit systems)

Yes, on the surface XP does run slower than 98, but do you understand why and the reasoning for it? IMO, the benefits outweigh the costs. I'd rather have a rock solid system that runs a bit slower than one that runs really fast, but crashes... and I loose my work. <{POST_SNAPBACK}> Given the last paragraph, why are you even using XP? I DO NOT ever lose my work in any Windows system. I save it periodically, or I use an app that just does it for me. I have used apps that crash any windows, including XP; the claimed "stability" oftens leaves a lot to be desired. With 98SE as is practiced correctly in 2005 you can have stability and speed. Wth XP you are a Windows Update mistake away from disaster. Note that M$ recommends corporate types use virtual PC's to test the impact of hotfixes and SP's on their operations before committing. Are you telling me you NEVER apply updates, such as the predictable steady stream of them coming out of M$? Basically, other than apps-upgrades, the SP applied to 98SE is perfectly stable. If we don't all work together to add frills to it, it just stays constant. To the extent that we all allow M$ to update IE, all of us are in the same instability boat; this is not an O/S issue however. I guess some of us have a curious definition of "stability". In any case, 98SE properly updated doesn't crash, if that's what's being referred to. And additionally, has the expected value of non-change to its innards, all of which in XP seems to be always up for grabs. cjl

This is MSFN forums, where we have the PRECISE OPPOSITE of what is implied in this post going on. [Note: I picked the above quote at random to illustrate a mindset problem that goes beyond the specifics in it.] About the only true thing stated here is that M$'s support of ANY Windows sucks to some degree, and perhaps their support of 9x family is even worse than the [already unacceptable] support level of XP family, etc. On this forum in the 98SE section, the forces of lots of people's experience largely nullifies all of the imprecise statements I am reading here. About all that can be gleaned here is that you all are M$ lemmings and will undoubtledly dump XP for Longhorne which will undoubtedly run a whole lot slower than XP, much less 98. For clarification, see the 98SE SP forum, but a few quick points first: 1) NO M$ O/S works at initial release. Since XP has had literally HUNDREDS working on THOUSANDS of fixes since "beta-testing" on the public, why should anyone grumble about un-fixed 9x when the fixed version is at hand? Anyone want to run XP sans hotfixes and SP's at this point? 2) 9x is not to be lumped together. Arguably the original 98 [as we refer to it as FE or First Edition] is as dead as Win95. This just means that the next improved system replaces the previously less-improved system. Note that the NT family of systems DOES NOT follow this logic, as it is a separate development. [Note: This notion is contrary to the words of M$'s SALESMEN; this is a technical discussion.] Arguably the best version is 98SE, since it benefits from having few if any restrictions compared to 98FE. WinME has some really nice improvements, but at the expense of having the entire system dumbed-down [a prelude for what was done out-of-the-box for the next released system--Windows XP, which most of us know how to UNDO to get it back to non-teletubbies usage, etc.]. Fortunately, there are MSFN forum members who have embarked on essentially excising just about anything that ME offers and grafting it onto 98SE if anyone cares. Also, there is a whole lot if misinformation here as well: In point of fact, some FIXES first appeared in WinME that were eventually released to fix the corresponding problems in 98SE, That they were in the initial release of ME and needed to be applied as hard-to-acquire hotfixes in 98SE is besides the point since we have all of this resolved in the Win98SE SP 2.01, a product whose discussion was created on this very forum! This allowed M$'s marketroids to attempt to sell ME as a "better" system than 98SE, and the same illogic is used to sell XP to the uninformed. 3) So where are we? in the world of 9x, experts are making it easier for lesser mortals to upgrade their unfixed 98SE systems trivially. Yes, M$ let us down, but that is moot given the easily downloaded SP. The ability to graft on the ME features is at a more embryonic state; you need to be savvy to apply the fixes, but that will undoubtedly change. Because precisely of this forum, as we all exchange information towards the overall common goal, clearly not only are these frills available, they will get easier to obtain and install. Note: None of these updates are as "hard" as applying any variant of the discussed-to-death XP SP2 problems, which apparently the deluded are willing to overlook, etc. I don't mind expending effort to improve my systems, but I do take exception to anyone suggesting that it isn't work to do ANY update to ANY system, etc. 4) XP can never be faster on comparable systems against properly stabilized Win98SE. In fact, XP can't even RUN on some of these systems! That said, it is true that by brute-force techniques, you can get a machine so fast that you don't mind the overhead, but please don't pretend that the overhead isn't present! Please note that parts of XP are "virtualized" to the point that they run on virtual hardware that is in itself emulated. [HAL anyone?] Some of this nomenclature goes back to the days of Win 3.1, but the actual implementation is something only found in the NT family. 5) It IS TRUE that XP can handle 2 GB of user memory giving any additional as the place where the O/S lives. I would hate to think that we are seeing 2 GB apps; that's something quite scary. Rather, it sounds more like incredibly poor program segmentation. If a section of a program is rarely used, it can be linked to dynamically when needed, causing an insignificant delay when loaded and then flushed, etc. I can't believe there are many applications that can justify this large space in any constructive way, nor do I think anyone here is actually pointing this out as a got-to-have feature. Regardless of this, Windows 98SE can "only" use 1.0 GB including for its own usage. This could allow an app to consume all of the additional memory at its own peril, but again, I don't think this is much of an issue. Yes, you do have to apply a few minor patches to the April 23, 1999 release of Win98SE to get this to work. But after you apply the SP it just works. Thus, doing an insignificant amount of work [far less than XP SP2!!!!!] you get a stable, fast, lean and mean Windows machine.] I don't doubt that many can quote now largely irrelevant war stories about how they used Win98 out-of-the-CD and had all sorts of problems. This sort of reminds me of the same when many people more than superficially used XP sans hotfixes and Service Packs. The only point that seems to stick is that most users didn't stumble too much on the hundreds of bugs eventually fixed as compared to the 9X family, which ultimately had far fewer things to fix, and perhaps this is more meaningful-- that now ARE fixed. Anyone want to venture a guess as to how many additional [mostly security] fixes we are still in for regarding the future of XP? I can tell you that for a 98SE system with the SP applied, the answer is likely none except for problems related to IE, which seems to be patched just about every month recently [including for 98SE! Note that every IE fix has three binaries: One for 9x, one for XP before SP2 or Win2K and one for SP2-only]. If you haven't tried Win98SE 2005-fixlevel edition, please don't make time-retro comparisons. cjl

*steps up on soapbox* Personally, I think the XP UI looks like a f**king clown car. Who thought that purple, orange, blue, red, and green was a good color scheme? Hell, while they were at it, they should have thrown in some fuschia. It sure couldn't make it uglier. And what's up with the ridiculously large window close, minimize, and maximize controls? *steps down* <{POST_SNAPBACK}> I routinely turn off all of the LUNA-cy when using XP, so it looks more sane, meaning identical to 95/98. MS themselves admits that the little crap in LUNA is the results of "set for best APPEARANCE" while what is wanted is "set for best PERFORMANCE". Both are available, but the lunacy "serving suggestion" turns on all of the slow-down overhead. Thus, even if you have a fast machine, XP crawls along identically to what it does on a "slow" machine [well, slow for XP standards!]. Apparently, things work sort-of like this: When the start menu goes to another level, schedule a sleep task and the task to get ready the stuff to paint the screen with. When either is done, wait for the other. When both are complete, THEN paint the screen with the results. So, if your machine takes 3 seconds to number-crunch the next menu-level update, then you are just about done when the sleep task is already done, and then you paint the screen after the delay. If you have a **** fast machine, you might calculate the screen update in less than 1/4 sec, but you then wait for 2.75 more seconds for the sleep task to finish. Either machine appears equally crawlly, etc. But if you set the max PERFORMANCE settings, the slow machine appears fidgety, unlike in 98SE, since the overhead shows how overloaded the slower machine can be. On the fast machine, it can resemble the zippy performance we are used to in 98SE, etc. BTW, XP SP2 has added another category of cosmetic silliness to luna, to slow things down even more. When you click on an .exe file, execution, BY DEFAULT, is delayed as the selected icon fades out. Thus, the further disparity between "maximum appearance" and "maximum performance". Remember, LUNA makes it real easy for those who have more fingers on their hands than items they access on their computers. To run the REAL program menu you have to goto "all programs" and wait three more seconds to get into the actual programs start menu. [but you can make any number of desktop icons to match whatever you run often, so why need LUNA at all? Isn't in your face better than having to call up any form of sub-menu? Besides, the desktop itself could have folders on it with sub-menus inside, ad nauseum, etc.] Another user has perhaps a better description of LUNA, it's the TELETUBBIES version of Windows. [bob version 2?] cjl

It looks that way. I know that COMDLG32.DLL, EXPLORER.EXE, and SHELL32.DLL are required for 98lite to install the sleek shell. I looked at those three files at various points of a single clean installation. # After a 98lite chubby install COMDLG32 DLL 176,128 04-23-99 10:22p COMDLG32.DLL EXPLORER EXE 180,224 04-23-99 10:22p EXPLORER.EXE SHELL32 DLL 1,400,832 04-23-99 10:22p SHELL32.DLL # After performing shell swap to sleek COMDLG32 DLL 92,672 08-24-96 11:11a COMDLG32.DLL EXPLORER EXE 204,288 05-16-04 5:46a EXPLORER.EXE SHELL32 DLL 819,200 08-24-96 11:11a SHELL32.DLL SHELL32 W98 1,400,832 04-23-99 10:22p SHELL32.W98 # After installing sesp2.0rc3 COMDLG32 DLL 92,672 08-24-96 11:11a COMDLG32.DLL EXPLORER EXE 204,288 05-16-04 5:46a EXPLORER.EXE SHELL32 DLL 819,200 08-24-96 11:11a SHELL32.DLL SHELL32 W98 1,388,816 12-06-01 11:25p SHELL32.W98 # After uninstalling sesp2.0rc3 COMDLG32 DLL 92,672 08-24-96 11:11a COMDLG32.DLL EXPLORER EXE 180,224 04-23-99 10:22p EXPLORER.EXE SHELL32 DLL 1,400,832 04-23-99 10:22p SHELL32.DLL SHELL32 W98 1,388,816 12-06-01 11:25p SHELL32.W98 After uninstalling the sesp it didnt put the older 95 versions of explorer.exe and shell32.dll. The .w98 file I assume is just created by 98lite to backup the newer version of shell32.dll. I was able to resync those files by shell swapping to chubby, then back to sleek. <{POST_SNAPBACK}> Indeed this is a problem. In a nutshell: 98lite CHUBBY is the same as "regular" 98. 98lite SLEEK [V1] is different. SHELL32.W98 is the ORIGINAL SHELL32.DLL with only the name changed [maybe not; see below]. SHELL32.DLL is the file taken from Win95B. Same for EXPLORER.EXE and COMDLG32.DLL For the SP 2 to be a total success, including de-installation, these files must not be modified. Clearly, the process "broke" the set, etc. The reason for the SHELL32.W98 version of the original file is that some components of 98 and some components of some apps require certain Windows calls that are normally handled by DLL-linking to SHELL32.DLL. The problem is that the Win95 version isn't capable of handling these calls and give a classic complaint in SLEEK [V1] about a shell extension error, etc. For every one of these that can be fixed, the binaries are patched so the calling program links to SHELL32.W98 instead, where the function call can be executed correctly, etc. Indeed, proper installation of IE60/SP1 with SLEEK [V1] requires the LOADWC.EXE file to be patched in this manner. I can provide a copy of the patch to anyone who needs it, etc. The 98lite author is working on a future 98lite release that will create the patched version if needed during the install, etc. My method works now as a workaround. [Note: I was the one that brought this whole matter to their attention; their former position was that the LOADWC browserWebcheck thing was to be ignored and even provided a fix to remove it. However, they now agree that it is NECESSARY to allow LOADWC to execute once you reboot after installing IE60/SP1, etc.] 98lite does all the relevant patches for 98SE as part of the SLEEK install or shell swap. However, it doesn't do any of the additional ones in the O/S area: 1) There is an updated version of WUPDMGR.EXE. Like the standard version in 98SE it also needs to be SLEEK-patched. If you are using this one instead of the standard one, you have to manually patch or replace it for SLEEK [V1] to work. [This file was at least tentatively to be put into the SP 2. To tell them apart, the standard one is about 56K and the updated one is about 68K.] 98lite only knows how to patch the original file. I have the patched and unpatched updated file, which itself is derived from the WinXP installation CD and was sent to Gape, etc. 2) LOADWC.EXE for usage after install of IE60/SP1. [The file comes up in various contexts later; The file is PRESUMED to ONLY be that which looks for IE updates, but actually does far more than that! For example, if you install AOL, or you change a dialer setting from/to dial whenever a connection is needed, the LOADWC.EXE file runs to update certain internal settings. When you install IE60/SP1, LOADWC.EXE is needed to finish the installation by adding "personalizations" and "customized settings" for IE, OE, Media Player, NetMeeting, the IE desktop icon, etc. This also includes creating start menu items for the installation. All of this is in addition to the original function, to make an Internet-based check for new IE versions. The sloppy implementation leads to an MS bug: In Internet settings Advanced, there is a checkbox to turn this feature off. Since the LOADWC.EXE has to run at least once just after the reboot when installing, the default value is to be checked. However, the bug is that you CANNOT turn off the checkbox and get LOADWC.EXE to stop. You can eliminate it with a registry patch or with MSCONFIG. If the checkbox is cleared, setting it will correctly restore LOADWC, just not the inverse, etc.] 3) Due to bugs in 98lite/MElite, certain Windows components aren't patched/unpatched correctly when installing or shell-swapping in WinME. To correct this, I appy a batch fix to just switch all of the known components either way. [Note: Due to file protection, this has to be done from a diskette boot.] Additionally, there are patches available [mostly from me] for a whole slew of apps that allow them to run under SLEEK [V1]. These include the trial and paid versions of Webroot SpySweeper, Adobe Acrobat Reader 6.00,6.01, 6.02, and 6.03, Kazaa Lite 2.40 [Note: 2.43 does NOT need to be patched; the authors fixed the problems themselves.] There are two specific apps not supported: Norton AntiVirus from 2003 forward because if you do NOT patch you get the shell problem, but if you DO patch, you get a Norton internal compromise consistency check error [damned if do and damned if don't]. The other is PowerArchiver past version 8.80. The program seems to have 5 different SHELL32.DLL calls and I haven't yet tried all 32 combinations of patching/not patching to see if a version that works can be achieved. [For the record, no patch and all 5 patches are two cases that do not work. There are 30 other cases to try!] There are complications regarding the SHELL32.W98 file: It indeed STARTS OUT as the unmodified SHELL32.DLL file from Win98/SE merely renamed. However, it is also subject to being replaced by the implementation of Q313829 which upgrades it to a version immune to a security hole. That file in turn, is subject to being patched [see Axcel216's MGDX site somewhere for details] to return compatibility with some sort of transparent icon trick which is binary dependent on the original SHELL32.DLL or the replacement with patches, etc. [Note: While using SLEEK [V1], the whole transparent icon trick, as well as any other cosmetic changes such as the ME icons [see below], are not available since you are using the Win98 SHELL32.DLL which doesn't apply to any of this. A feature handled by the SP is the inclusion of WinME desktop icons patched into SHELL32.DLL. To my knowledge, the offsets where to apply these changes doesn't move relative to either the original or Q313829 versions of SHELL32.DLL. Clearly, whatever changes are enabled, either the standard version, ME icon patches, Q313829 base version change, or the transparent icon trick fixes, the version chosen has to be made to be the file used by 98lite to be either the standard SHELL32.DLL in CHUBBY or be the SHELL32.W98 file in SLEEK [V1]. My reference to SLEEK [V1] is to distinguish it from SLEEK [V2] which is experimental at this point. The idea in SLEEK [V2] is to eliminate the entire problem as outlined above, and instead do the following: Patch all of 98SE proper to expect that SHELL32.DLL is the "normal" file, but the actual shell is SHELL32.W95 [which is the one from Win95b] which is called by a patched version of COMDLG32.DLL taken also from Win95b;same for EXPLORER.EXE. In essence, this means to start from CHUBBY and just change a few files. The problem is that while this eliminates the entire original problem, it adds in about 6 new ones still being worked on by the 98lite folks. In theory, a solution may be at hand, but unfortunately, relatively recently some new more subtle problems emerged, thus there is no current recommendation for using SLEEK [V2]. I don't expect the SP2 to support 98lite's specific quirks; that should be the province of either 98lite itself or someone's after-install file maintainer, etc. However, since the SP2 does modify the files and is aware of all of this, it should keep the original files to be restored if the SP is uninstalled! cjl

This fix is only intended for WMI (WBEM) installs. WMI is not installed as part of original 98 + 98SE. Only ME installs WMI as mandatory component. Just a bit misleading: In any case, you CAN install WBEM. In 98lite installs, these sub-components are broken out and can be installed/not installed at the time of initial system install.WMI requires WBEM, but WBEM does not require WMI. Regardless of how you get there, the hotfix applies as long as the feature is already installed, mandatory or otherwise. btw: In WinME, WBEM is a mandatory requirement of the help and support center that forcibly comes up in safe mode. 98meLITE will prevent you from elimination of the WBEM if the help center is enabled, etc. [Note: Eliminating the help center itself is not recommended as it then makes other things break!] The point is that WMI itself can be eliminated while it makes sense to keep WBEM there, etc. Additionally, you made reference to Steve Gibson's site [GRC] and his tirade against UPNP. As I understand it, he is referring to the UNFIXED UPNP that precedes these hotfixes and their cousins for WinME and WinXP. To my knowledge, there is currently no basis for complaining that UPNP represents an ongoing security risk on properly patched systems. Gibson would better serve common interest pointing out the need to be up to date instead of banning UPNP altogether. cjl

These get installed when you start from WinXP and enable Internet Connection Sharing. You then can also optionally make an install disk for use with 98 or 98SE [ME and up already have UPNP installed independently, perhaps as an option, but clearly locally available as part of their own install]. Once you use that disk on 98/98se, you now have installed the flawed upnp for 98/se and need to appy the above to fix same, etc. Note the files have the mark of WinME, thus versions of the form 4.90.300x. Without the initial upnp installed, these hotfixes give errors similar to how Q329048 gives a complaint caused by not having 98 Plus! installed. The installer is looking for the previous files to upgrade, etc. These hotfix files are not new and have always been available since their respective release, etc. On my own system, upnp allows Windows to notice I have an "NAT Gateway Device" which is my wireless router back to the cable modem, etc. I would hope that the SP 2 could apply all of these option-dependent fix files as the scenario is always the same: If the files exist, just replace them with newer ones. Adding them under any scenario should be harmless at worst. cjl

If such a "magic" system existed, we all wouldn't be here!Just for the record: There is NOTHING WHATSOEVER like the SP which at this point is still undergoing "growing pains" as we add even more frills to it. The fact is that most of the patches provided by it have been "mislaid" by MS over the years. The relatively "minor" collection of files that were ever offered by MS is merely a "component" of the SP which itself breaks down into roughly the following: 1) Hotfixes "freely offered" by MS for 98SE today. This is itself a composite of several sub-components: a) Ongoing available hotfixes that they haven't yet "lost" in terms of making them generally available to just about anyone with a net connection. [Apparently, there are some of us who don't yet grasp just how little this represents as compared to the whole SP project!] b) Files that used to be in a) but no longer are. Made available by certain individuals such as myself who shall remain nameless. c) Relatively recently added fixes actually made available by MS, or at least they are currently until they wind up in the b) category. d) [There is at least one of these, or so it was yesterday] latest updates promised for availability but only by Windows Update, but in fact not yet available; if MS doesn't deliver on this fairly soon, perhaps some pressure can be exerted to obtain this. e) Updates formerly in d) but now available, but only from Windows Update. [Officially, anyway "only" available from Windows Update.] f) Updates available, but they are claimed to only be for other versions of Windows. This in turn involves several sub-categories: I) Files which are seemingly designed for other 9x-type systems [such as ME] that if you "dissect" them you find that the "innards" works fine on 98SE. Thus, the only reason they don't work for on 98SE is because of a deliberately inappropriate restriction of the hotfix's installer that can be gotten around using a non-standard procedure that works not only on the intended Windows system, but on 98SE as well without otherwise any problems. This usually involves something like merely running an executable within the hotfix, which generally is a self-extracting archive with command line options to merely extract, etc. or you can use something like Power Archiver on it to get to the same place, etc. II) Similar to I) except that you have to right-click INSTALL on an "inner" component, generally a .inf file to get the installation going. III) Similar to I) except that the hotfix is for an NT-family version and cannot be directly run even if dissected. May possibly be similar to II), but some can only be implemented by using a key set of files found within the hotfix and manually placed into 98SE usually while in shutdown DOS mode due to file protection, etc. g) Files recommended by MS as workarounds for the fact that there is no general fix for a problem; the files are generally themselves available. An example of this is that MS will NOT FIX the general implementation of h.323 protocol in 9x, but apparently did in most other Windows. However, in their opinion, the only client application for this feature in NetMeeting, for which they re-released a version that is claimed to be immune to the security problem. In this case, the protocol involves sending voice packets over the Internet; whether you get any problems with other programs YMMV with regard to problems, etc. Thus, this sub-category is more of a band-aid than a hotfix as a lame excuse for the latter, etc. h) Files that have NEVER been generally available from MS. This in turn comes in two sub-varieties: Documented and un-documented. The documented ones use fanciful language which adds up to a "stonewall" in effect saying that they have it and you don't, and tough on you. Several posters have indicated some partial success prying some of these updates out of MS and "contributing" them to the SP project. The un-documented ones exist but either have no current documentation or their documentation has been "discontinued". In some instances there are no longer any write-ups on them within the MS KB, but some of us have read the KB articles when they did exist, and in some cases even had a freely obtainable download way back then... In some other cases, they apparently were NEVER available as opposed to "merely" being CURRENTLY unavailable. Yet, some of us, who again shall remain nameless, have "contributed" them to the SP project, etc. There are "interesting" stories about how these updates were "stumbled" upon, but this is not a topic for discussion on this forum, etc. All I can personally recommend is to always look where something could be; sometimes you find the unexpected! In some instances there are documentation "discrepancies" where problems are stated as never to be fixed, contemplating some newer version of something, such as entire Windows [for example, fixed in ME with the implied suggestion to abandon 98SE and instead purchase/upgrade to ME, etc.] thus suggesting there is no available hotfix for the problem, yet the truth is that there actually IS a relevant hotfix, sometimes identified as the binary associated with the article whose verbiage is claiming the binary doesn't exist! 2) There is another wholely different component to the SP that by its nature is NOT available, namely advances to what 98SE actually is. In most instances these are either third-party files that are generally available for download; fair use would seem to be consistent as the SP merely makes it easier to install them; the result would be the same as the more straight-forward installation of these packages, but would require the user's time, bandwidth, or expertise better handled within the SP itself as an attempt to be a "catch-all" for all relevant problems, etc. Alternatively, some of these are MS files that are not "intended" for the purpose put to, but in point of fact they do work and in some way enhance the overall 98SE product. In some case producing what some would describe as a "Frankenstein" of Windows features taken from other versions such as ME and/or Win2000 or even possibly WinXP including SP2. Many in this category are being argued for/against by people on this forum because of their personal opinion about just what the SP ought to consist of, not that the add-ons themselves shouldn't be made available somehow. Some are arguing over the IMHO relatively myopic notion that it makes the overall size of the SP "too big". Proper design of the SP can virtually eliminate this argument by being provided in modular form; in essence, if the component isn't present or you don't want it, either delete it from the SP install directory and/or invoke some form of command-line or other option to prevent it, and you never needed to download it if that bothers you. For others, bandwidth is a non-issue: Please note that WinXP Service Pack 2 consists of a SINGLE EXECUTABLE FILE that is 266 MB long! Sorta makes arguments about whether the SP is 15 MB or 18 MB or even 50 MB pale by comparison! With the work of some people, "master" versions could also be made on a CD image suitable for burning from say, an ISO image, etc. For certain people, this would be the preferred method, totally avoiding download. It would be useful to allow such a method to have the ability to make small updates to the overall image to create more recent descendants without having to do the entire monster download again. There are successful other vaguely analogous projects like ours such as AutoPatcher for XP which support the idea of an update as opposed to a complete download, etc. There is another more intangible aspect of the SP as well: A lot of input goes to testing out interaction between the various hotfixes et al. The SP is clearly an important step to preventing negative interactions as it can implement specific combinations that are known to only work either as the specific collection they have become, or alternatively, it provides a framework to guarantee they are applied in a critical order known to work. There are numerous examples of critical interactions that have perhaps only a few acceptable install orders yet numerous ways to screw up if merely casually installed due to interaction. The SP can fixate the order to prevent this sort of problem. Additionally, the list of problems addressed in some MS KB articles [many of which are already documented on Gape's website] does NOT imply that the hotfixes that the articles refer to are actually present. This is either because to actually implement the hotfix may be impossible because it's not available [to anyone we know] but has been replaced with a superset hotfix that includes what was implied in the article along with additional fixes addressed in an alternate KB article, etc., or it's merely redundant, albeit harmless. In some instances it is actually DANGEROUS to actually implement the stated KB article because the provided hotfix is supplied by MS with a defective installer or flawed registry settings. The SP gives you the intended fixes but sidesteps the negative consequences of the MS-provided binary hotfixes. Thus, an invaluable aspect of the SP is how to judiciously apply WHICH particular updates and in which order, to provide the benefit implied in the MS KB articles, yet avoiding the unintended "side effects" you get doing it MS's way, etc. Much of the schema described here comes from the work of many individuals; this is a form of "community effort" for which we need to thank our forum hosts for providing the communications platform, etc. Without all of this feedback, the SP would be quite different from what it NOW is! cjl

Q840315 updates the ITSS.DLL file. If anyone needs it, I have the "real" MS one which updates the QFECHECK program/info as well, etc. There is a little trick to get updates from Windows Update for your own use later. I also used it to get the more recent Q890175 update [not relevant to SP 1.6.2, but needed for 98SE]. It seems that MS is making it harder to get updates because they will only be available directly from Windows Update. Here's how to do it: Need to have a relatively unpatched system regarding Windows-Update-available updates, etc. Just get a whole lot of updates, including the ones you want to keep/cannot get any other way, and start a really big download. Once the download is finished and installation of the updates starts, hit the reset button on your computer! Start in safe mode and locate a new arrival, namely a folder named WUTEMP. Within it are all of the updates you asked for, including the hard-to-get ones, etc. [Note: Most of the updates are available from many places, including a download link found on the KB article or security bulletin, etc. Windows Update is just an alternate way to get those, but these new only-WU ones... ] cjl

No, it's just not for what you think is important.We discussed this already: MICRO is for extremely teeny systems that can boot on an LS-120 without any usage of the hard disk whatsoever, so you can maintain your hard disk from that vantage point while it is not booted up. No hard disk files are exempted from any form of access/deletion, etc. Incidentally it's fast, but the point is that it is fully-featured *enough* to run defrag, scandisk, anti-spyware/antivirus stuff, etc. that can find problems with your main system that it itself cannot even find due to cleverly-hidden viruses, etc. Every variant of 98lite has a valid purpose, including OVERWEIGHT; it shows you what 98lite does by omitting the improvements temporarily! cjl (have used all 4-5 variants)

I can't afford to use XP, and I'm not talking monetarily. I don't let any machine with XP on it get out of my hands without at least 98SE for a maintainence system where you can run antispyware/antivirus stuff apart from the security-buggy XP. This is an ever-burgeoning problem I am seeing at an alarming rate: The system this happens on is always the same general configuration: I can run WinXP on say drive C: and also dual-boot to drive F: to run 98SE [lite perhaps]. Both systems are prepared to run the same set of anti-spy/vir tools. I have a floppy backup of BOOTSECT.DOS to get 98SE up along with the other few files 98SE needs to have unvirused on drive C: if necessary [the basic DOS stuff that gets 98SE proper up on drive F:]. Run all tools on XP including Ad-Aware, CWShredder, SpyBot S&D, and Trend Micro Housecall free virus scan [i recommend this one; finds stuff Norton and McAfee NEVER find!]. After all of the gauntlet of tests, XP tells me it is TOTALLY FREE of unwanted guests, etc. which have either been deleted/quarantined/removed or were never there, etc. However, this is in fact NOT THE CASE! Now boot to the F:-based 98SE system and do the same collection of tools. Trend Micro Housecall finds that there is a resident trojan in the XP system! One particularly bad case was that SVCHOST.EXE was infected and XP said it was not! Thus, all of the attempts by well-meaning people to get the public to at least run available [and mostly free!] anti-spy/vir tools isn't good enough, since smarter viruses are around that give the intended false sense of security afforded by inadequate testing for problems present but not known to the user, etc. As I said, I can't afford to run [only] XP. cjl (Woody Lenhard referring to Gates' quote: Trustworthy computing, indeed!)

I see a lot of misinformation, disinformation, and lots of ignorance in this thread, as well as an indication that SOME people actually know something! [Gee, sounds like a public forum....] There are all sorts of reasons people use any and all O/S'es for various reasons. DOS is never dead, since there are vital applications that have to run from DOS. Simply put, Windows is NEVER a "panacea" O/S, simply because it's not organized around an in-memory model of itself capable of totally unrestricted file replacement guaranteed because said model would have to be totally unrestricted in design. [Note: DOS is NOT the only system capable of doing this trick written by Microsoft; the other system is OS/2 which can be brought up from the BIOS level by 1-3 diskettes to a non-GUI but totally functional system capable of minimal task switching and 100% memory-resident operation. Once IBM asserted its ownership of the code they paid for, Microsoft embarked on a bashing campaign because the "Chicago" project was 4 years late, not yet done, and IBM was getting some rave reviews from the technical press. Gates ordered the project to get a product out, which became the wretched subset Windows 95 rev 0 released in Aug 1995, soon patched to be known as Win95a, which didn't support hard disks bigger than 32 GB nor FAT32 despite elaborate claims that it would have these frills. Win98SE is merely the fulfillment of that project released April 23, 1999 after Win98FE was released in November, 1997 following two other re-releases of Win95 in 1996 and 1997 that did add FAT32 support, but never larger disk support. MS's bashing proved fatal to IBM's product, but not for technical reasons; totally hypocritical to bash your own product merely because it changed hands after you rave about it while you do own it, etc.] Once you have a memory-resident system, there are lotsa good things you can do with it. One area still being developed constantly is embedded-DOS systems that appear in all sorts of places you never would imagine such as set-top TV boxes by the multi-millions. No hard disk needed, just a ROM big enough to hold the relatively small amount of code of a DOS kernel. Closer to home, I and hoards of others use programs such as Norton Ghost and PowerQuest Drive Image Professional to backup disks/partitions. These programs run in DOS to take advantage of this model. The UDMA interface sounds like a way to get enormous improvement in backup times with Ghost; will try it and see... Shane Brooks, the 98lite guy, also runs a business that sells embedded 9x. He can get it down to as little as 5-6 MB stripped of all frills. This is useful for some who need a GUI system, but it's still about 8-10 times as big as DOS in the same arena and beyond the reach of elegant ROM-based solutions. [On modern motherboards we do see 4 MBIT ROM's, but that's .5 MBYTES uncompressed. Big enough for DOS in one chip, but need 8 chips to hold stripped-down embedded 9x.) Back to Win95 itself: After 95 came out there was an expose book from an ex-employee complete with code examples and provable/reproducible evidence showing that 60% of the code in Win95 was BINARY IDENTICAL to Windows for Workgroups 3.11 [the end-of-the-line product for Windows 3.x] Part of the reasons are for DOS compatibility. Much of this is also true in 98 and 98SE. The part that has to talk with DOS cannot change, etc. But the fact that so little else was different only underscored the fact that MS was farting around with the project for so long accomplishing so little; realistically it took them 8 years past WFW311 to get to 98SE, not to mention years more of so many dozens of patches that are the essence of why we are here for the SP. cjl (read .sig below) ps: Isn't it interesting that there are two O/S's identified as "98" when in fact NEITHER of them was released in 1998! Some of us want to call 98SE Windows '99.

Erp: You are ABSOLUTELY CORRECT that not everyone needs any and every hotfix/patch. That said, the problem is that there are far more updates that are perfectly fine to install in anyone's system than the far fewer list of those that can potentially cause more harm than good. In essence, there are a few updates to be nearly universally avoided; fortunately, for a few of them, they were eventually replaced by better-still hotfixes that not only corrected the original problem, but also the unintended side effects as well, thus now suitable for all of us to install because they are at best helpful and at worst only innocuous, etc. Indeed most of these are what is in the SP and they won't be optionable since there is no reason to limit their installation. However, MS never changed the status of any of them from the "stonewall" position they have always had on them. Fortunately as a group we have "acquired" them and they make up a goodly portion of the SP. Thus, we are making them available despite MS's position of overly restricting their being available only to those who "need" them, etc. To make matters worse, some of the updates HAVE been available that cause the harm to some systems while the larger amount of updates that are truly safe aren't available by "normal" means. Indeed, even Windows Update has distributed "bad" fixes at times, only to be [hopefully] corrected later. For example, take two YOU pointed out, namely that 828026 is broken and should be avoided because they had to release 832353 which both replaces it and fixes the damage that 828026 causes if installed, etc. cjl

Erp: I don't want to start a flame war with you, but a few points: 1) First you chide me for overly long posts [which I believe I did NOT do; what I posted takes 8 short paragraphs to explain!] then you post multiple posts in a row that are each longer and aggregate to far still longer. 2) I stand uncorrected as you clearly still do NOT understand what I said. As such, I will repeat what I said clarifying accordingly: ALL MEMORY AGES AND GETS SLOWER. The circuitry of the memory is CAPABLE of dealing with the memory at its ultimate lower speed; reputable memory does this by not attempting to exploit the "extra" speed the memory has when it is "young". Because of the SPD chip and the way the memory is timed, this "extra" speed capability is not used. Thus, you have a smooth operation of memory that runs as you expect from day one and for years to come, etc. Unfortunately there are disreputable memory chip vendors who sell memory sticks that emit bogus timing information that leads the mainboard to believe that the memory is capable of what ultimately it is not, namely reliable operation at the rated speed down the road. If the mainboard can be manually timed by ignoring the SPD information, you can indefinitely get the memory to run at WORSE than rated speed reliably, but you can only get the rated speed while the memory is "young" enough in terms of powered-on hours, etc. Unfortunately, the trend today is to have BIOSes without manual timing characteristics/settings, so you are at the mercy of the SPD chip on the memory stick telling you a conservative figure that can be trusted IF the vendor is reputable, or you are installing a "ticking time bomb" of a memory stick destined to fail when you have powered on the machine long enough. So, for your example, the point is that say a memory is rated for PC100 service which means that you can complete a memory cycle in 8.x nanoseconds. A reputable vendor uses chips that when actually new COULD have been rated to be able to cycle in say 6 nanoseconds. However, being reputable, he sets up the SPD chip to give back info that says the chip runs in 8.x nanoseconds and all is fine indefinitely since once the memory has say a few thousand powered-on hours, it ages down to only actually being able to do 8.x nanoseconds, etc. Since this is in spec, you have reliable operation. Notice that your machine never changes how fast it goes relative to memory speed, since the chip timing always is the same. The only difference is that the "headroom" the chip has regarding how close to the edge of whether it really can do it or not is getting closer to the edge as the chip ages, but reputable memory never goes over the edge, etc. Mr. Sleeze down the street sells the same chips that TODAY do 6 nanoseconds as "PC133" memory with an SPD that tells you that the stick does 6.x nanoseconds cycles. And today it actually does work. But several thousand powered-on hours from now, your machine is freezing up because at that point the chip is only able to do the 8.x speed as above. Mr. Sleeze also sells you "PC100" chips that are actually more like 10 nanoseconds ultimately but today will pass for 8.x nanoseconds and will work for awhile. Down the road your machine locks up/freezes/crashes, etc. So we have come full cycle as to why I brought it all up. I have had numerous experiences over the years with memory that aged out of spec leading to numerous "interesting" machine failures. Some of them are quite curious, such as: One machine could run MS/PC-DOS 6.x and Windows 3.1 but OS/2 crashes with an error message usually associated with bad memory. Also DR-DOS refuses to load on the same machine. However, the same machine eventually won't even finish the POST with the crap memory in it. Changing over to reliable memory solves all problems, etc. I have a small box here somewhere chock full of crap memory that can be made to work on some machines if you retard the timing enough [assuming the BIOS will even let you!]. But I won't put it back into a machine that has to be depended upon, etc. Some of the chips are DDR, some say PC133 and some say PC100. I wouldn't believe what any of them have printed on them, but someone did when they were new! cjl

You are of course correct. My point was merely to illustrate that there are nasty releases of hotfixes we need here for 98SE that nominally don't work on 98SE, but if you go "inside" you can get it to work on 98SE anyway. Thanks for the post on that IE55-patch specific method. Will keep it in mind if any more overly-restrictive stuff comes out. So far, all 28 of the patches I know about for IE60SP1 don't need any of this kind of "help" etc. Anyone need any expansion on the list of 28 hotfixes? [Or have MORE than 28?] cjl

I have similar sentiments regarding saving time during installation, a seemingly interminable process much of the time. Since I virtually always am installing IE 60 SP1 and ALL of the updates over <no IE there presently because this is 98lite eliminating IE 5.0 before I start> I would want an easier way to install all of the updates than just doing them "the hard way". It would seem that an SP having an optional feature to do this is a reasonable request; making it mandatory is as foolish as not being allowed to ask to have it at all. However, the main event is apparently unclear to many of our readers/posters: 1) You CANNOT GET NO WAY NO HOW some of the updates that are in the SP because MS has decided to make them unavailable for whatever twisted/lazy reason they choose. In spite of that, some of us have come across either the updates themselves or the vital components of the updates that make it possible for Gape to include them in the SP. 2) Even if you have all of the updates in INSTALLABLE FORM [which NONE OF US HAVE ALL OF NOT EVEN GAPE!] it just takes an insane amount of time to install them piecemeal. This I can say from the dubious "experience" of having most of them and doing all of them manually. I really don't look forward to doing something again that is "painful" times about 1.3x assuming I had those 30% more I don't have in stand-alone installable form, etc. 70% of the job is too much of a pain!! 3) Most of them DO NOT COME FROM WINDOWS UPDATE either presently or in some cases ever. Yes, Windows Update has tended to "lose" a few updates over the years. But it NEVER represented the majority of updates some of you have been manipulated into thinking it should/does/ought to be, etc. BTW, that also goes for WinXP where the actual number of post-SP1 updates that were ever on Windows Update was a cruel joke compared to the number actually available if you "dug them up". [Last year, I had a catch phrase: "I have all 167 of the 45 updates you can get from WU".] [Yes, I know XP is at SP2 now, and largely this is moot, as long as you can run all of your drivers/apps under SP2!!!!! ] 4) For unknown insane MS reasons, some of the relevant updates are available, but NOT DIRECTLY INSTALLABLE into 98SE. Sometimes there seems to be an "agenda" to a hotfix, such as "This thingie only installs on ME" even though clearly the update has nothing specifically to do with ME, rather than the 9x "family" which includes 98se. The SP may not be needed to overcome this IF YOU WANT TO DO ALL OF THE WORK to overcome this, but it IS work. And sometimes you can't get it to work at all. The SP can and will overcome this because its vantage point allows the files to be installed anyway, etc. 5) Various people who congregate here [and some "silent" others who aren't vocal/don't have the time to "chat" ]have done work on finding files that enhance 98SE but were never disclosed by MS as to their relevance. For example, Win2K updates contain files that are useful to 98SE, but aren't documented as anything other than an update to Win2K. The SP will provide them; for most of us, there is NO OTHER WAY to put them in because we would have to learn how to trick the O/S into accepting them. [i know, this is more of a problem in ME or newer, but the point is that we really don't have the bulk of people here wanting to wantonly patch their systems to death manually!] This is why we are here folks; it's not for "convenience" as much as a virtual necessity that we need the SP, and wanting it to be as complete as possible is a laudable goal in and of itself. cjl (Go get em, Gape!)

erp-mon: Is this true? [i have no experience; don't use IE 5.5 anymore.] Have you tried to install the hotfix "from the inside"? There are a bunch of 98SE fixes that "only install under ME" unless you goto the inside files and install it manually. [Don't quite remember the specific ones, but 828026 for WMP comes to mind for example.] Use power archiver or whatever to get to what's inside and find the .exe or .inf and see if that works, etc. cjl

Hey, erpman: Who needs 98lite micro? I do, but not for my main system: I make machines with LS-120 instead of a floppy. 98lite micro fits entirely on the LS-120 and boots up with no regard for the hard disk whatsoever. From that vantage point I can diagnose various problems, run anti-virus/anti-spam, ME defrag that rearranges all files with no reservations, etc. on all of the now-looked-upon-merely-as-data hard disk partitions, etc. I posted awhile back all of the differences between 98lite CHUBBY and 98lite SLEEK [V1]. There are people who prefer SLEEK over CHUBBY who have a free choice. If SLEEK [V2] ever gets fixed, it might even tilt the scales a little bit more for SLEEK, since from a user's standpoint, SLEEK V2 works identically but has less "growing pains" due to ill-behaved apps, etc. One particular point is that regardless of what you are running 98se on, SLEEK beats the hell out of CHUBBY with regard to emptying the recycle bin, running more like XP on steroids than the standard lame sloth of all other forms of 98SE 98lite or not, etc. and has no problems with copying lots of/large file manipulations that seems to plague others [Have we resolved the BROWSEUI.DLL and BROWSELC.DLL problem? Can I safely revert to the IE501SP2 versions if I have all of the IE60SP1 and hotfixes installed? In SLEEK I don't care; it doesn't use them]. Believe it or not, there are users who don't care about win2K themes, ME icons, bloated screen windows that put up options on the top they never use, toolbars that waste taskbar space, active desktops, etc. For them, SLEEK is just fine. Also, the SHELL has nothing whatsoever to do with what the O/S does or does not support in terms of hardware. I agree that win95 is dead, that's why I don't use it for a variety of reasons such as lack of FAT32 support [except in 95B/C], lack of large disk support [no bigger than 32 GB] and lots of drivers and apps that won't load there, but are fine with 98SE, and a general lack of hotfixes for problems it shares with other 9x that didn't get fixed there, etc. However, the SHELL is not related to any of that, since using the 95 shell on top of 98SE changes the hardware situation not at all! I have yet to find a machine that cannot install 98lite regardless of shell choice, unless I also cannot get plain 98SE to install either [and there are lots of cases like that unless you are prepared to do a lot of driver research in some instances, and perhaps install some stuff in safe mode, etc. In one case, I had to do the pass one of the install up to the first boot, on another board entirely, then use GHOST to transplant the rest of the install on the problem board, which then worked fine! This was NOT related to 98lite whatsoever, etc.] [On some boards, you have to install hotfixes before you install drivers or it crashes. In some instances, such as on fairly recent TYAN boards, the driver installs on the provided support CD are actually installing hotfixes automagically while they install the drivers to make it even work, etc. Installing Gape's SP can work wonders here BEFORE you install the drivers!] So, yes, by all means install/use what you want, but please no misdirected rants here. cjl

I take a small exception to the "it doesn't matter as long as it fits" attitude expressed here, especially in light of my having to fix an AOPEN machine as recently as 10 days ago [just before my vacation]. I admit I'm seeing the problem a whole lot less in the DDR era, but SDRAM and older memory suffer from a problem that is also shared with DDR [in the case of DDR I have only encountered it twice so far!] in that the memory is fraudulently rated. [Even from companies with "lifetime" warranty on the memory stick.] The problem manifests itself as a machine is fairly new and has worked fine up until now and/or a working machine has had memory added not that long ago and has worked fine up until now. The machine now is flakey/locks up, won't boot, etc. In the case of large memory, the machine won't work when the new memory is depended on for cache or large apps, etc. The actual problem is that ALL memory slows down with age. PROPERLY RATED memory takes this into account and is rated for how it will perform down the road, and not for how it works when it's new! On fewer and fewer motherboards, it is possible to setup the memory timing manually. In many of these cases, setting up p***-poor performance settings allows the machine to work. albeit more slowly than before. Unfortunately, the vast majority of machines run on timing dictated by the memory sticks themselves. They use an "SPD" chip onboard the memory stick that basically tells the mainboard what the memory timing of this particular stick is, so that's what the board sets up and expects from the memory. If the memory is repuatable, this timing is realistic indefinitely, i.e., is still reliable after the memory has been in use for some hundreds or thousands of hours. [Note: This is powered-on hours, not sheer age. Regardless, I have some oldies-but-goodies machines that are both old and used a lot and have no memory problems is spite of being over 20 years old!] But if the SPD chip gives out overly optimistic settings to the mainboard [let's just say the quality of the settings is "sufficient to work through the end of the warranty period" ] then there WILL ultimately be a memory-related failure. While there can always be a random failure, make no mistake about it, the vast bulk of memory failures are PLANNED because of the statistical way memory is rated. If you measure the true max speed of memory made yesterday, you can with great certainty give the maximum reliable speed of the memory at the other end of powered-on time. Sleezeballs know this and commit the fraudulent ratings anyway . cjl

That's what I thought! However, I don't think there is much call for updates to IE55 here, just for IE60SP1. The "correct" answer is to either leave it at 5.0 as it is, or to bring it all the way to IE60SP1 with all known updates, unless you just want to do the long approach of making anyone happy for all possibilities, which would open a time-consuming door: Everyone might have a pet version they want to see: IE501 IE501SP1 IE501SP2 IE55 IE55SP1 IE55SP2 IE60 IE60SP1 not to mention various beta releases of any and all of the above. I can understand doing nothing but updating that which must be there, namely IE50 be brought to the best level possible, since it would usually be there, unless removed by 98lite or IEradicator. But I would also desire updating IE60SP1 to encompass the 28 updates I know of [are there more? ] to get us to as good as it gets, etc. And again, to make this clear: Some of these updates are NOT part of any cumulative "rollup" updates you get from Windows Update which in turn are generally either IE updates or OE updates. Thus, the SP could do us the service of making it possible to be OTHERWISE up to date needing only any follow-on updates that WOULD be available from Windows Update, assuming there even will be many more of these. I wouldn't have a problem where the SP did all the "dirty work" thus all I had to do was goto WU to get some more recent fix and then expect the SP to "trail" WU by some reasonable period of months, etc. And when MS declares all IE versions obsolete other than wherever IE60SP2 lands at that point, then the SP would essentially be final [other than any other ideas we dream up or other unrelated fixes come to light, etc.] As to the problem of size, why can't the SP check to see if the need to update what you have exists and also an optional update file to accomplish it it also exists? That way if you really want it you can download it while if you feel it's either burdensome or not to your liking it can be ignored, etc. Even with all of the versions of IE out there, all fixes to all of them, all of their original releases, the entirety of the SP, and any additional fixes any of us want to dream up, etc., the whole scope of such an item still fits comfortably on a single CD. Distribution of such a disk would be desirable, and I'm sure there are people who would help get such a thing to those who needed it, etc. cjl ["think big"] ps: Gape, glad you are back; I was away on vacation and heard you were sick; hope you are better!

Two things: 1) Not clear just how much longer they will continue to do this considering their cavalier attitude about leaving all but XP behind at IE60SP1+ while XP gets IE60SP2+ etc. 2) To my knowledge, there have been 28 updates to IE60SP1/OE and while many are redundant in the rollup sense, some are apparently NOT ever included in any other IE update, and are thus "stable". If we have a base package that includes all of what gets us to say, Dec. 04, then everyone is at most a recent rollup release or two away from current for the foreseeable future, etc. [One for IE and one for OE.] cjl

No, you are fine: You CAN use the original 95 (950) set of three files, but the preferred are the ones from 95b (which various people are violating MS's copyright and linking to, etc.) and you CANNOT use the 95c files! cjl

Well, you CAN edit the start menu, just not "in situ" as Shane describes it. Remember, this is the Windows 95 shell! You use: click on taskbar/properties/start menu/advanced. This puts you into a specialized version of Windows Explorer that starts and ends in the Start Menu folder [even if you have multiple users, it picks the correct menu, etc.] From there you can edit start menu items. Also, in situ editing cannot add anything while this mode allows any start menu editing including creating new groups. In situ cannot paste an item, but it can copy/cut so you can put on say the desktop, etc. This is a rough summary of the differences between SLEEK [V1] and CHUBBY: SLEEK has the problems of compatibility with a handful of programs, the majority of which can be patched for SLEEK compatibility. [98lite does most of the work, but the list is close-ended.] CHUBBY is the basic 98SE shell so no compatibility issues per se. SLEEK has a bug in that all folder sizes are presented modulo 4096 MB in terms of the properties of a folder. Since most folders aren't that big, this is hardly ever a problem in realworld situations. [i have a 10 GB folder that appears more like 2 GB!] CHUBBY has no such bug. SLEEK has instant recycle bin empty, better than WinXP. 98SE takes a long time depending on how many logical drives/volumes you have. The above is partially influenced by a bug involving the BROWSEUI.DLL and BROWSELC.DLL files and IE version conflicts. SLEEK is TOTALLY IMMUNE to this problem of instability involving large/many file transfers and related hangups. SLEEK lacks many frills of CHUBBY that most users turn off anyway. One particular exception to this is the ability to enable via a registry patch the ability to view attributes in details view. SLEEK seems to lack the ability to "allow all uppercase names" as it is slightly erroneously known as in CHUBBY shell. What this actually means is that if a name conforms to 8.3 short filename conventions, then auto-capitalize the first letter and auto-lowercase the rest of the name. While cute, it can be misleading. CHUBBY clearly has the ability to turn this on and off. Anyone know of a patch to SLEEK to turn this cutesiness off in SLEEK/MICRO? SLEEK is clearly faster as well as more stable. To properly install IE60SP1 without any prior IE files, it is marginally harder to install in SLEEK due to the need to do a kludge I personally stumbled upon that will eventually be automated in a future version of 98lite once Shane figures out what the "nicest" way to do it is. To those that know, it involves the so-called LOADWC.EXE problem. [i can elaborate for anyone who needs to know, etc.] SLEEK ignores all cosmetic WinME icon changes induced by the SP. SLEEK doesn't support the automatic properties of how you can dynamically move/change where "My Documents" is from that icon directly. However, you can also do this successfully regardless of shell from tweakui Version 1.33 or literally by cut and paste from any variant of Windows Explorer window, etc. 98lite 4.6 shell swap to SLEEK included a dummy "My Documents" kludge that is merely an ordinary shortcut to C:\My Documents. This is true even if your system drive is NOT drive C: ! 98lite 4.7 and up has deleted this "feature" allowing you to make your own kludge. [Note: a nuance of the AUTHENTIC My Documents icon found only in CHUBBY/OVERWEIGHT/98SE-ONLY: When you open My Documents, the path label says "My Documents" and NOTHING ELSE. Assuming you asked for the display of the full directory path on all windows, this is curiously inconsistent. Indeed, should you open a subdirectory contained in My Documents, it will CORRECTLY show the complete path for THAT directory, etc. A standard shortcut to My Documents will CORRECTLY obey the shell setting, etc. Thus, arguably a "homemade" shortcut [the only one available in SLEEK or MICRO] is "better" than the standard one of CHUBBY and co.] SLEEK may have a bug regarding seeing "Suspend" on the Start Menu which can be fixed as documented on the litepc website. Implementing a registry patch may cause your machine to be faked out into believe you have Automatic Power Management even if you actually do not and/or have ACPI instead. Ignore it, since all it does is install support for non-existent hardware which is harmless. Once this is done, the patch can be used to remove the "Suspend" or even put it back by "reversing" the logic in the patch. I can provide Suspendonmenu.reg and Suspendoffmenu.reg for those who want this both ways, etc. The following is a SERIOUS LIMITATION of 98lite's ability to invoke Shell Swap: Regardless of choice, you MUST maintain your shell choice forever IF you also install IE60SP1 over no prior IE version AND the update to Outlook Express Q330994 and/or its "rollup" follow-ons! Any attempt to manipulate the shell once the OE updates are installed WILL PERMANENTLY CORRUPT Outlook Express!!!! [Even doing a Shell Swap from/to the same shell currently will cause the corruption!] Any attempt to install said updates will also kill OE if the shell has changed since the point of install of IE60SP1 [including the shell swap from/to the same shell as above, etc.] To make matters worse, all of the above ISN'T GOOD ENOUGH to avoid the corruption! The ONLY WAY to get it to work is to: 1) Pick a shell, any shell, and stop changing it forever. 2) Install IE60SP1 over no version of IE. Use my manual kludge of LOADWC.EXE if using SLEEK or MICRO. 3) Install IE60SP1 AGAIN! Use all the same options you selected earlier even though they are bold in the installer [indicating they are already there, etc.]. This will result in a message asking you to exit the install or use "reinstall all of the components" Select the latter. This will allow Q330994 to install correctly [amazing, isn't it? ] cjl

burkgul Posted on Dec 10 2004, 09:02 AM Thanks for your hard work and an amazing result so far. Please add the 98lite-SLEEK compability to 2.0 RC, it will be very appreciated. I have just finished giving Gape all of the 98lite SLEEK [V1] stuff I am aware of. If you are using 98lite SLEEK [V1] or 98lite MICRO, the SP 2 will not update EXPLORER.EXE or SHELL32.DLL, but will attempt to update SHELL32.W98. Thus, the SLEEK [V1] shell will continue to operate normally after the SP is installed. However, the references to ME-type icons, etc. will not apply, as these are wholely contained in those files which are only relevant to plain 'ol 98SE or 98lite OVERWEIGHT or CHUBBY. Since SHELL32.W98 is the original SHELL32.DLL from 98SE, the update is required to fulfill the security problem addressed in Q313829 regardless of your choice regarding the cosmetic ME-type icons. Other than the bindings-enhanced NOTEPAD.EXE [which has to replace the standard NOTEPAD.EXE as modified by 98lite for SLEEK [V1] operation, thus it also needs to be Sleek-modified exactly the same way, etc.] I don't believe there are any [current] files in the SP 2 that are in conflict with SLEEK [V1]. Anyone running 98lite SLEEK [V1] after the SP2 is applied should report back HERE any new occurences of the shell32 errors that weren't there before in case we goofed on anything else! As an aside, I can provide patches for various third-party apps to be SLEEK [V1] compatible such as Adobe Acrobat Reader 6.00 and 6.01 [haven't done 6.02 yet]. Please note that Symantec/Norton AntiVirus 2003 and following cannot be made SLEEK [V1] compatible due to internal consistancy check spoiled by the attempt to patch for sleek compatibility [damned if you do/damned if you don't], etc. Additionally, I have an updated launcher for Windows Update from MS. [it was provided by Windows Update to anyone who used a Windows XP install disk to check for upgrade compatibility issues; if you allowed MS to download to you updated Windows Update files as suggested, you get this newer Windows Update launcher] As it comes with no surprise, this file ALSO needs to be patched for SLEEK [V1] compatibility, this there are patched and unpatched versions. [Note: the "new and improved" Windows Update no longer provides this service, but fortunately I did capture the relevant files!] I would expect this new launcher to be made part of a future (SP 2.1?) service pack, but I think it needs to wait until we determine what form the SP takes with regard to 98lite SLEEK issues in general, etc. However, there is no problem doing it yourself independently of the SP releases. Anyone out there have any feedback relationship with the author of Power Archiver? Versions past 8.80 are NOT compatible with SLEEK [V1] and it's at least much harder/perhaps impossible to patch for sleek compatibility. A PAYING CUSTOMER perhaps can explain the problem to him/her/them/it and get this fixed, etc. The kazaa-lite versions just before 2.43 needed a patch for SLEEK [V1] but the authors were contacted; Version 2.43 IS sleek compatible. I can also show anyone the CORRECT way to install IE60/IE60SP1 under SLEEK [V1]. Shane Brooks is working on a way to automate my method for future releases, but as of 98lite 4.7 you have to use my manual method and the patched LOADWC.EXE file to get a FULL and UNTAINTED install. [On the 98lite website there are now-proven incorrect references to how the unpatched LOADWC.EXE file is "extranous" along with instructions to eliminate it, which is being updated to be consistent with my method, etc. once it becomes automated/integrated into the next 989lite release; this is a lot of work for Shane because it covers new territory for 98lite not much different from what the SP has to do, namely deal with optional upgrade packages that only MAY be present, compounded with the need to interfere with the MS installer of said package, etc.] cjl