Drugwash

FREE ONGD GOM Player 2.1.40.5106 has been released. I believe the failed registration bug introduced a few versions ago, has been fixed in this version. Thanks goes to the Gretech team!

If you only have one USB controller on that machine, I'd say it'd be safe to just swap the files. There is still some reservation, as I see the root hub on my machine also depends on VMM32.VXD, which is a file buit ad-hoc from other files, so there's a slight chance of incompatibility if they don't play well together. But swapping back the old file should fix it right away, in case anything bad happens.However,if you have more than one USB controler (I don't know the board and can't go searching its manual now), then you'll have to place the new file in SYSTEM32\DRIVERS only after renaming it to something else (make sure the name stays within the 8+3 character boundaries) and after that, you'll have to locate the proper registry keys pertaining to that particular hub/controller and redirect them to the new file. Which is a bit of a dangerous mission if you haven't done registry work before. It's not recommended to tamper with the registry as it can render the machine unbootable. Can't say more, because looking at my own USB-related settings in Device Manager, it's a total mess there. Just... be careful and backup any files you replace, plus the registry (type scanregw /backup at a command prompt).

Hmmm, don't be so hasty. I already warned about mixed controllers. What happens if one has a - say - Intel on-board USB 2.0 controller and a VIA-based PCI to USB adapter in a PCI slot? There's a reason people might want such configuration, mainly due to insufficient ports (some old boards only have 2 USB ports but even 4 may not be enough - that's up to each user's needs). I think there has to be a configuration inf that assigns the correct driver to each controller and that inf has yet to be designed. Otherwise it'll be a mess. Any other opinions, anyone?

According to the official VIA site, the latest available USB 2.0 driver for Win98/ME is 2.70p, which I do have. It's a 15MB package though (the multi-language version), so better grab it from the official site directly. Here's the links: English only Multi-language

Unfortunately I'm not familiar with VB, not to that point; I can do small changes/fixes based on similarities with other languages or by trial & error, but that's about it. These last few days I've been trying to build a small app that'd automatically remove all USB-related devices/controllers/etc from a system in Normal mode, in view of installing NUSB 3.5 (or later). However, so far I've failed, partly due to incomplete/erroneous documentation from M$ and partly due to uncooperative setupapi.dll/cfgmgr32.dll API under 9x. I'll keep trying though, maybe with alternative methods, if possible. Thing is, if any kind of drivers have been installed by Windows upon normal installation, there's chance the USB chipset(s) can be read from the registry; VIA has Vendor ID 1106 and Device IDs are 3038 for USB 1.1 (PCI to USB UHCI) and 3104 for USB 2.0 (VT6202 EHCI). Please note there may be other Device IDs besides these - that's all I have on my main machine though (the VT6202 is actually a PCI to USB adapter card). If Windows did not install any drivers for the USB chipset(s), then BIOS functions should be used; I don't know if there's any Windows API (or which ones they may be) that can read the devices detected by the BIOS. Of course, the Unknown Devices in Device Manager have their correspondence in registry but at this point I'm not sure where exactly and how are they located there - maybe under ENUM\PCI\Unknown device or something. Please bear in mind that a machine may have more than one USB controller (as in my case) and they could be from different manufacturers.

If I'm not mistaken, the documentation reads that at this point one should reboot again. This may be the tiny difference; otherwise please do try the VIA driver rollback suggested above, only in case your USB controller is VIA-based. While I'm here, let me congratulate with PROBLEMCHYLD for his hard work. Hopefully I'll find some spare time & parts to build a test machine; preferrably in this life. Sorry for not being helpful enough, especially lately. Good luck with SP 3.1!

Is there anything you would want me to add to the tool above? I'm not sure if an automatic parsing of the header files would yield proper results in regard to detecting calling convention and whatnot. I may need more details as my knowledge of C/C++ is pretty shallow.

API Parameter Count v1.0.2.0 is out. Just a few bugfixes and improvements. Enjoy! DOWNLOAD executable + source code Homepage

The error number is actually hex 0x80004005 (E_FAIL), which is an unspecified error. Please check this MS KB here (I hope it's right - the board renders the link dialog completely off-screen in SlimBrowser/IE) (yeah, took me that long to fix the link when I could see NOTHING in the link dialog!)

Oh no! I'm sworn not to touch Opera ever again and I mean it. I did try a few versions over the years and could never find the experience I needed. It's been quite a pain to switch from SlimBrowser to K-Meleon, then briefly to Firefox and finally to SeaMonkey and I've had just about enough of switching browsers just because of a bunch of [...] that changed standards so that more and more advertising could be shoved down our throats. As for all those variations of Firefox - no matter their name - they're all built upon the same Gecko engine and that engine itself is the problem for us 9x users, so no matter what else they take out or put in, it still won't work correctly as long as the engine is the same. All those Bookmarks not working are a result of engine updates. Since SeaMonkey 2.1a3 is the last version where Bookmarks work correctly under 9x, just check the Gecko engine version it's been built upon and then cross-reference with any other Gecko-based browsers/variations you might find - if any other has a newer version, Bookmarks won't work and so will be with any other issues that have been reported around. Except maybe when some team has actually reverted any of those changes, but chances for that to happen are very slim. Personally I suspect they started to use Unicode strings and functions that are buggy or not supported/existing under 9x; haven't looked at the sources and probably never will as I'm way too busy with other things and not that knowledgeable in C/C++ (or Delphi, or whatever they may be using).

I used to use K-Meleon 1.7a2 for some time, after SlimBrowser (on top of IE6) started crashing MSHTML.DLL on more and more sites (this one included). However that thing has a serious bug which - on scrolling pages, with or without images - crashes the display driver, forcing me to reboot. Therefore, upon someone's mention here on this board, I installed and configured SeaMonkey 2.1a3 and a few add-ons. Later on, the person reverted to an earlier version of it (some 2.0) due to frequent crashes. Well, I do experience those crashes myself, because after all that tinkering, downloading, installing and configuring add-ons, I just didn't feel up to another challenge, seeing that many of them add-ons are not compatible with that particular alpha build; some of them - having auto-update on - even started to auto-disable themselves as the new versions were suddenly incompatible. That's a model of sloppy programming where there's no compatibility check before installing an add-on. What I wanted to say is that this browser mostly crashes for two main reasons: JavaScript and lack of memory. Actually, the memory thing is more subtle than that, because even with enough free RAM (FreeRamXP Pro - a tool I've been using ever since the Win95 days - doesn't help here), it still crashes silently, which means it's all about something else like too many GDI objects, no room for handles, stack full or things like that which I have no idea how to pinpoint. Only a reboot fixes this issue. As for JavaScript, I have an add-on that can disable/enable JavaScript, Java, Flash, SilverLight, CSS upon a single click (displays a few buttons in the statusbar), so those issues can be temporarily avoided. Together with other add-ons like Ghostify and NoScript, it's a quite pleasant experience, all in all - until resources get depleted, that is. Personally I have no need for Flash, I never ever go near YouTube and I immediately discard any site that requires Flash for its main menu or other important functions. Flash is opaque technology, anything might hide inside, user has no control whatsoever over its content and actions, so to me it's a no-no. I mourn the days when browsing was pure HTML, no link could fool you through redirecting or other modern tricks, there was not even the slightest idea of a video/audio captcha that'd drive one crazy, no stupid flashy ads flying, jumping or sneakily hiding transparent and so on. The Internet of the 21-st century makes me sick, literally and I try to interact with it as little as possible; I'm sorry, however, for the good people I get disconnected with.

PROBLEMCHYLD, it may not be that complicated, after all. Both RP and KEx are standalone, self-contained installers and they both require a reboot after installation. If you decide to include them in the Service Pack, my opinion is that they should be added as-is and if the user chooses to install any of them, then deploy the respective package to a safe location (%windir%\Cabs or whatever), create shortcut to it on the desktop and inform the user that they have to launch it manually. Obviously, this operation will have to take place only after the Service Pack has finished installing its core files (and/or the optional ones). Just make sure the files in the Requirements package recommended by Tihiy for RP, have been installed regardless of the core/optional choice, if RP installation was selected.

Oh well, I had a funny feeling I won't be taking care of the garden this year either... Since it's closely related to my own needs, I may be taking on the task of building an as-updated-as-possible API database, complete with library location, parameter count & type and - most important - return value & type. As a mini-MSDN. The return value is indeed very important, not only in terms of success/failure indication, but in certain situation may have to represent a pointer to a string or worse: a pointer to a pointer. Things get complicated when the original API would return a pointer to a POINT, RECT, COLORREF or more complex structure. But that's already too much to dig for - we may as well build the API replacements from scratch, as is the case with KernelEx. B)

Agreed with Dave-H. I'd still have some reservations in regard to KernelEx; it's quite unofficial alright, however since most recent software (browsers, primarily) wouldn't install/work under 9x even with latest updates in this Service Pack, it may be beneficial to include it, albeit in a separate section called 'Advanced' or 'Expert', if you want. Thing is, KernelEx is not an application or utility, but an extension to the system core, namely the kernel; as such, it should be regarded as a core update. Admittedly, there may be occasional problems with certain applications, fine tuning may be required, but with proper documentation and under a well-defined section that appeals to advanced/bold users, it would be of much use for the future. And definitely would be easier for anyone to have an all-in-one package with options to select from, than to pick bits and pieces by hand, from different places. To be honest, I myself wouldn't conceive a 9x system without Revolutions Pack either, as it makes it more eye-friendly. But I know there are minimalistic guys out there, there's guys that also disable themes in XP and that's fine, because having the option to [...] is most important. Come to think about it, it might be an idea to build a modular Service Pack, so at download time, people could choose whether they want the Optional pack and/or Expert pack in addition to Basic/Core pack. Additional benefit would be that a subsequent update would alleviate the download bandwitdh usage when only one of the packs has been modified, as opposed to current 'compact' solution. But as I said, it's just an idea.

You're welcome! It wasn't much of a challenge, anyway; currently I'm working on parsing the headers in PSDK2003 and building a constant database for use in AHK. Found over 67850 so far, but the parsing code is still buggy. I'd be glad to find someone with a more recent PSDK installed (ideally one that includes Vista/7 or even 8's headers), to pass them the parsing script and retrieve an updated list of constants. When I fix it, that is. I haven't yet looked at your code, but I can't stop asking myself how would it know what are the allowed return values & value types for each API function it patches... unless I misunderstood its usage. If you need any other tool, don't hesitate to ask - I'll try to rise up to the challenge the best I can.

I mocked up a simple tool that returns the number of parameters for an API name given as command-line parameter. Alternatively, a direct launch will display the whole list of known APIs to browse through. The application contains the API list published in this post ; it extracts and uses it as default if there's no other. Hopefully it may be of help to someone; however, no guarantees or warranties are implied. It's free software. Sources are included in the 7-zip package. [EDIT] Updated to version 1.0.1.0 (change log included in the package) Download API Parameter Count v1.0.1.0

I may be asking too much but... could you provide some basic docs as to which is what and the stage the project is in? Admittedly, I know nothing about Pascal/Delphi and also don't have any such IDE installed, but I also see some C sources for a RP10 that I never heard of before. So please, a few pointers (so to speak ) would be more than welcome. Although most (or rather all) of us out here would rather have you back, working on this masterpiece. Coz this piece of your life is also part of our lives, and has been for a few good years. Revolutions Pack is synonymous with Tihiy. (fixed a couple of typos)

Thanks guys, but as you know, to buy something, one needs money first and then access to the "market"; I have neither, in this country. Some of my hardware is pretty old, most BIOS implementations do not know LBA48 and some of the older ones barely accept 32GB HDDs or even less (such as the 486 which only takes 526MB IDE HDDs). Anyway, this discussion is off-topic, let's not deviate. Thank you for trying to help!

Open RegEdit and check this registry key and set it to 0xFFFFFFFF hex (-1 decimal) if it's 0 (zero). HKEY_CURRENT_USER\Software\Microsoft\Windows\CurrentVersion\Explorer\Advanced\TaskbarSizeMove Also monitor this key's value when (un)checking Lock The Taskbar menu item in Taskbar's context menu (right-click in Taskbar empty space). This item only appears when RP is installed. @ LoneCrusader: what you point to in the first screenshot are the Chevrons I mentioned above (RBBS_USECHEVRON style). They should appear when taskbar is unlocked, with a comctl32.dll version of at least 5.80.

@ jds: Please check the version of your Common Controls (COMCTL32.DLL); if it doesn't read as the screenshot below, please update it manually (in DOS mode) and check if the unwanted behavior persists. Programming info: The Rebar control is implemented starting with v4.70 of comctl32.dll. It uses different REBARBANDINFO structure sizes according to IE version (since MS implemented CommonControls updates in IE only - hence stripping IE from Windows is a BAD idea). The problem is probably caused by the control not recognizing the RBBS_USECHEVRON style flag in the REBARBANDINFO structure, since this particular flag has been introduced in v5.80 of comctl32.dll.

@ PROBLEMCHYLD: I'm sorry for not being around lately, things are quite chaotic on my side of the world. Haven't read about your personal problems other than the mentions of other members in this topic, but rest assured I'm with you and I'm glad things are in order now, as far as I can see. As for the pack itself, I still couldn't find a way to test it (or separate components), since the biggest problem I'm facing is the lack of IDE/ATA hard drives for my (very) old machines, then the lack of RAM for those that could hold a Virtual Machine. That is why I didn't wander around here posting nonsense. My apologies for not being of much help! And thank you for your continuous work!

Install GOM Player. Try to play one of those files. It will attempt to find and download the codec for you, if it cannot find it on the system (and if it's a free codec). I got mine like that ( WebMVP8Decoder.exe ) The Comments field reads: WebM VP8 Decoder(0.9.5.0) webmproject.org EDIT: Few minutes ago I decided to check for an update and the new player version 2.1.37.5085 failed to register GVF.ax (error 0x0000001F). Installation went on and completed without other incidents, however I'm not sure what this failed registration may impact on. Dependency Walker shows no missing libraries or functions. Under XP there's no such error. I've submitted a bug report to Gretech Support Center.

To add/replace icons in a library, you may either edit an existing library using Resource Hacker (replacing icons manually one by one or using a ResHacker script) or building the library from scratch through a Visual Source, MinGW or similar project - for the programming inclined people. On another note, I've been experiencing a reproduceable problem: if the K-Meleon browser is running, trying to switch the RP theme through Display Properties >Appearance will just freeze the system. Closing the browser (forcibly, through Task Manager) will immediately resume the operation, completing it succesfully. Related to my previous issues involving K-Meleon and its crash of the video driver (or whatever that was), it may be too early to be sure, but it may have been the fault of Logitech MouseWare (since I'm using an old Logitech mouse) - the browser hasn't crashed yet, after uninstalling Logitech MouseWare (I think, version was 9.79.1, latest I could find some time ago). Fingers crossed! Unfortunately it didn't help much; the browser did "live" a little longer, but eventually crashed the video again. I give up!

Current alpha 1.3.15 has problems using Windows DirectSound as Audio Host. I'm getting an error related to Output Device not found. It does work using MME though. Being an open-source project, I suppose 9x programmers/users can always suggest related fixes to keep compatibility. Thanks for the heads up!

As a long-time Revolutions Pack user, I thank you for your kindness, Gelosea! If possible, I would dare to ask dencorso to add mention of your agreement and a link to your deviantArt page in the first post of this thread, so anyone interested could browse your work, possibly helping in keeping your views counter up.