SlugFiller

Well, it's just a matter of reading the attached resources. I imagine command-line resource hackers/decompilers can do that. Though I'm not sure if that's necessary. If the file's hash does not match anything from the database, it may simply say "Version unrecognized". Argh, my eyes! Why so much green? And also, it's not very descriptive. I agree with Sfor on this one: The files checked can be determined by the ones available in the database. I could probably clear up my schedule and write a scanner+GUI in C++ or Java (according to forum members' preferences), but unfortunately I haven't the time, resources, or expertise to manage the database. If I can find a volunteer for that, I would be more than eager to create a database viewer/editor. I would think a file's language would be a part of the "from" field. That is, you would have "Q239696 English" and "Q239696 Japanese", etc... That way, it could identify the languages of the files you have, and you can compare. On a side note, which format do you think the database should be in? Some form of solid binary (a-la flushed struct array)? Maybe something with indexing, or dynamic tags? Text, maybe (or XML)?

But as I already said, it doesn't work for what I want to do! You are only testing for WM_RBUTTONDOWN when g_hRichEdit is focused. You can just as easily test for WM_RBUTTONDOWN in the window proc of the rich edit control for a similar effect. The behavior would slightly differ: If you right-click on the main window's border or title while the rich edit is focused, currently it opens a pop-up menu, but it won't if the WM_RBUTTONDOWN is done in the rich edit control. This behavior can be mimicked by having WM_RBUTTONDOWN also detected in the main window's window proc. Although I doubt this behavior is desired, correct me if I'm wrong. Also, instead of actually detecting WM_RBUTTONDOWN yourself, you can forward it to DefWindowProc from the rich edit's proc to get it to generate real WM_CONTEXTMENU messages itself. Your code would then be able to detect Shift+F10 and Context-Key as well. Although you would only care about Context-Key if you have a 104-keys keyboard.

Again, my suggestion isn't for an auto-update program, or auto-downloader, or anything of the sort. All the program would do is scan your Windows and System folders, then tell you something like "Your ios.vxd v4.10.2222 is outdated, you should get Q239696", or something of the sort. Then you pop open a certain sticky thread in this forum, and get the appropriate file. The way I envision it, the results window would look something like an eMail program: -On the top half there would be a list of the files you have in the system, each with its version, some hash-type thing, a guess on where the file came from (if it is recognized), and the scanner's guess on whether it's "outdated", "up-to-date", "unrecognized", etc... Obviously, appropriate sorting and filter option will be available. -On the bottom half, for each file you select, you will get a sort of "version history", giving you all the information on each version of the file starting with the one that came on the disk, and up to the latest unofficial patches, including various branches where such exist. As mentioned above, the information on a file's version includes where it came from, which could be "Original 98 CD", or "Original Danish 98 CD", or "Q123456", or "Unofficial ShellEx v4.20.9000", etc etc... The database itself would simply be a large array of entries, each including: -File name -File version (Preferably as seen in the file's resource) -File hash -File source (Short string, a la "Q123456") -Up-to-date flag, stating whether this file warrants an update. May be multi-state ("Update now", "May need update", "Not latest but doesn't require update", "Latest", "Update if IBM portable", etc). Each file would have one entry for each version. So, for example, there would be about 7 entries with the filename User32.dll. Obviously, scanning and tagging hundreds of files for the database is a lot of work, but it's not difficult. Tools for automating at least part of the process can be easily made. Incidentally, by simply using a different database, such a tool can be extended to any Windows version.

If you ask me, placing message handling of a specific control in the main message loop is just plain wrong. It's the sort of hacking you'd expect a Mac-developer to do on Windows, "because it's not quite the same". Creating your own window proc is not that difficult, nor is setting it to your control. It's available for Riched20.dll. I've noticed you're using Riched32.dll which is slightly different. Not sure if it works with 32. At any rate, supporting IME is fairly simple. You just need to load up a small COM object which offers a TranslateMessage alternative (if loading fails, you just use the normal TranslateMessage), plus handling of WM_SETFOCUS and WM_KILLFOCUS in the RichEdit control. It basically enables the creation of multi-byte characters with multiple key-strokes, for languages where 102-keys just doesn't quite cut it. One more thing, if GetMessage returns "-1" then it's probably a good idea to leave the loop, not just destroy the window and keep going. After all, having already failed once, there's no reason to believe it would successfully return the WM_DESTROY or WM_QUIT which would be posted by a normal DestroyWindow. A "return -1;" or "throw;" may be more appropriate there.

You seem to have missed the sentence that followed: True, changing KernelEx would indubitably solve your problem, but the slightest mistake will live you with a non-working system, or worse, cause data corruption. The prior is not that big a deal if you know how to use DOS, though, since KernelEx automatically backs up any system files it alters. The That being said, if you want to push through, here's some of what you'll have to go through: -You need to get the KernelEx source. KernelEx is a sticky on this forum, and the source is linked to in the first post. -You'll need to figure out how to compile KernelEx. This actually has two parts: You'll need to learn how to use GenPatch (provided with the source) to create your own APIs. Then, you'll need to learn how to compile a patcher from GenPatch's results. -You'll need to write an API replacement for RegSetValueExA/W in advapi32. You can look at existing APIs in the KernelEx source in the subfolders of "\KernelEx\src\api". A prime example is "\KernelEx\src\api\Kernel32\GetVersion.c" which also demonstrates how to read custom options from KEXVER.INI. This could be useful to allow you to customize the registry entries you wish to "lock", giving a more robust offer. -You may also need to overwrite RegOpenKeyExA/W to keep track of the key's path as I see no way to retrieve that from the handle. Needless to say, you'll need at least some programming knowledge. If the KernelEx source code looks like utter Greek to you, you probably don't have a chance of making this happen. In that case, your only other choice is to ask Xeno86 for such a feature and hope he agrees to create it (though I wouldn't bet on it). If you do decide to go ahead with this, and at least have enough basic knowledge to get your initial bearing, I'd be more then willing to help you with whatever issues you run into.

I got Inkscape to run, but all the menus, tooltips, dialog buttons, tab titles, combo boxes and lists had no text in them. Even spinners appeared empty. Window titles did appear fine, though. Incidentally, is it supposed to say "Maybe OK"?

I don't follow... Notepad supports localized texts. AFAIK, on native systems, it supports IME. It doesn't seem to support 9x-style installed IME, though. On the topic of subclassing, it simply means using: // In global WNDCLASS richedit_wc; // In init GetClassInfo(NULL, RICHEDIT_CLASS, &richedit_wc); Somewhere in initialization, creating a message proc with: LRESULT CALLBACK EditProc(HWND hwnd, UINT message, WPARAM wParam, LPARAM lParam) { switch (message) { //... } return CallWindowProc(richedit_wc.lpfnWndProc, hwnd, message, wParam, lParam); } The applying it to your control using something like: // Creating the control (replace with your creation code) HWND edit = CreateWindow(RICHEDIT_CLASS, NULL, WS_CHILD | WS_VISIBLE | WS_VSCROLL | ES_AUTOVSCROLL | ES_MULTILINE, rect->left, rect->top, rect->right-rect->left+1, rect->bottom-rect->top+1, parent, (HMENU) id, (HINSTANCE) GetWindowLong(parent, GWL_HINSTANCE), NULL); // Setting the new proc... SetWindowLong(edit, GWL_WNDPROC, (LONG)EditProc); As for WM_CONTEXTMENU, you will need to forward WM_RBUTTONUP and WM_RBUTTONDOWN to DefWindowProc in order for the message to be generated. It may also be worthwhile to test if Shift+F10 works. This may also warrant for appropriate forwarding of messages. For more details see this little bit of info on WM_CONTEXTMENU in RichEdit, and this description of WM_CONTEXTMENU. Also, if you want, I can post my little code snippet for adding IME to a RichEdit control.

@dencorso: The file is here. Someone should probably contact VIA and tell them about mraid not being 9x compatible, though. They still provide it as a 9x driver download.

It works! IT WORKS!! IT WO-HO-HO-HO-HORKS!!!! Okay, getting Q239696 was easy, thanks to this forum. It was also utterly useless. I actually had to use the Wayback Machine to find viasraid. Apparently V-Raid versions as old as 310E already have mraid. I had to go all the Wayback to V220E where it was still called "SerialATA" instead of V-Raid. Incidentally, getting the file directly from Wayback didn't work, as the archive came out corrupt (only the file list was valid). Fortunately, the direct link to the VIA-Arena download mirrors still worked, so I managed to get the file. Well, after updating the driver, and setting the SATA mode to "Raid", my HDD is finally out of compatibility mode! Suffice to say, I really like the increased performance. I'd like to thank all the forum members, and especially dencorso, for all of your help!

Does your editor have IME support? Or is it not that type of editor? I remember when I tried creating a control with IME support. I went to hell and back twice, but managed to get a working code snippet in the end.

Last I've checked, auto-patcher didn't have a database-only version. With >250MB download, I would think it's trivial to understand how looking for specific updates would be preferable to someone with a mostly patched system. Besides, I was thinking of something that gives more of a per-file analysis, rather than attempt a full-system update.

Just thought of one more possibility: It's possible to edit the KernelEx source code to make those two registry entries ineditable. Just a matter of adding a small file in "api/Kernel32". It's sort of like taking out flies with a heat-seeker.

I'm currently using 4in1 v443, the one recommended for 98. Works fine for my IDE controller. However, 4in1 is only vatapi, not raid. Well, this seems noteworthy. Where do I get sraid? Both v-raid 560 and Hyperion v513 only have mraid as far as I can tell. Incidentally, my ios.vxd is only 4.10.2223. Which Qxxxxxx should I be looking for?

Get KernelEx from the members projects. Should solve the problem.

Not exactly since: 1. The Microsoft Windows Update system is online, meaning you have to be online. Not such a big problem this day and age, but still... 2. The Microsoft Windows Update system is online, meaning it sends you system information to Microsoft (and your credit card information to random third party companies). 3. The Microsoft Windows Update system is down. I can give you a screen-shot of it saying "your OS is too old and we don't support it anymore". 4. It only has official updates, most of which require security updates and bugfixes themselves. 5. It does expect to automatically install whatever it detects is missing anyway, so it's not that different from a service pack. 6. As I recall, when a service pack is available and not yet installed, it recommends installing it, regardless of the actual versions of files you do have installed. What I'm thinking about is a program that simply scans your system folders, MD5s or SHA1s or similar all the file there, then compares it to its own (offline) database containing filenames connected to version histories. It then reports what versions your files are at, if there are known better versions, what older versions exist, which files (if any) are unrecognized (unknown version), what systems certain files are designed for (e.g. recent unofficial ESDI version targeted at IBM portables), etc. Also, the database can include marking regarding whether or not a certain version of a file requires update, could use an update, or is sufficiently updated (latest version, or later versions not critical or useful under certain settings), then files detected as being in a version requiring an update could be displayed first. Additionally, the database should contain, for each version of a file, besides the version number, file hash, and, obviously, the file name, also the place from which it can be obtained (Qxxxxxx, unofficial X, or original CD, etc). Now, I don't think a program that checks files and reads such a database, as well as a GUI for displaying the result, would be that hard to create. The database itself would certainly require serious management work, but no more than any service pack, and since there are like 5 or 6 of those, you can't really say there's no one who'd be willing to do it.

Well, it was already suggested that you use regedit at startup, I can't think of anything more automated than that. Changing the settings within the theme description would also do the same, but with less resource costs, and less hacking. Aside from those two options, changing manually every time, and changing manually on less occasions, I seriously don't see what other options are there. Maybe if you can give a hint as to what sort of alternative you're expecting, exactly, because I think the entire combinatorial range has been covered here.

There are plenty of current projects in this forum which simply install plenty of updates to the system regardless of what you want or need. It would be nice to have, instead of something that installs the upgrades, something that scans your Windows folder, checks the versions (or MD5s, or whatever) of the files, and tell you which updates you actually need. And maybe which ones you already have installed... Then you can go to this forum and look up updates in one of the stickies. Would be much more useful than a massive service pack for people who are already mostly updated. It can also be a really small download, since it doesn't actually contain updated files, only version information.

Okay, new thought: Is it possible for a SATA controller driver to be incompatible with daemon tools? While messing with settings, drivers, bios, etc (it would take a really long post to describe all the combinations I tried), it seems the d343port used by daemon tools got screwed up. After repairing the daemon tools installation, the device I set in place of the SATA controller now shows d343port.mpd as its driver. Except that I've set it to "VIA Bus Master PCI IDE controller", it shouldn't have d343port as its driver, it should be using vatapi.vxd. This is quite strange. Maybe I've just messed with the settings so much I got Windows confused...

You don't need this registry key to be correct while "debugging" your themes, you can just set it after you finish, no? That aside, if you're already "debugging" your themes, wouldn't it be "within your scope" to just look up those two values and set them (and then debug to make sure it works)?

Well, I've switch the SATA to IDE mode, went into Safe Mode, and changed the driver about 4 times until I got it to use ESDI_506.pdr instead of viamraid.mpd. Only notable difference was, it hung on startup loading ESDI_506.pdr instead of viamraid.mpd. For what it's worth, that was right after successfully loading ESDI_506.pdr for the CD drives. ESDI version used: 4.10.2226 from LLXX. Next, I'll try dencorso's suggestion to use 4.10.2225 instead. Incidentally, I did read the 1TB thread. It seems the problem there was only with formatting and using scandisk, not with start-up or drivers. Strange that no one else seems to have any similar problem...

The LBA48 thread is, unfortunately, not exactly reach with testimonials. The few that are there discuss disk-sizes rather than partition sizes. Usually, their large disks are divided into many small partitions. I did find 98 Guy's post regarding his large partition. Unfortunately, he doesn't really go into detail as to how he did it. It does give me a few things to try before I resort to repartitioning my drive.

@dandnsmith: The first thread is about limitations of the "format" and "fdisk" utilities. It doesn't mention what happens if you use third-party software to partition the drive, then attempt to use it. The second only seems to talk about file sizes, not partitions. @Sfor: ESDI does not appear in the required files list for the driver: C:\Windows\System\Iosubsys\viamraid.mpd C:\Windows\System\Vmm32\IOS.vxd C:\Windows\System\Iosubsys\viamvsd.mpd Also, I think I already have the unofficial ESDI installed, though I could re-install the latest version to make sure. Links please? Did they mention chunk size, etc?

Well, I don't actually use themes. In fact, I'm not sure my 98SE is theme-able. So I'm no expert on how to re-adjust the skin configuration settings. Closest thing I've got is Appearance Schemes in Display settings. Well, I would think for the same reason people don't switch their desktop images every Friday - Why would you want it to keep changing appearance? Really, I would think this sort of thing is a "pick your favorite and stick with it" sort of deal. I would think the pain in your eyes for having to deal with a new color scheme every short period of time would be reason enough to cut down.

Read the article. Half about it was applicable to "Why 98 and not 2000". The remainder was pretty much "It doesn't have IE". Fair enough, I presume, although I have seen the occasional site which requires IE.

Have you actually ever tried doing that? None of the standard utilities to actually install programs enable choosing a destination folder. Synaptic, aptitude, rpm, they simply don't. So right there you have to compile your own program. Even then, most programs expect to find their own executables and data-files in the default location, and crash if they're not there. Same as what happens when a program looks for, say, GDI32.dll. Poor example. GDI32.dll is part of the Kernel. I can accept the Kernel and its data-files being in a specific location. That being said, the presence of "/etc" should invalidate the need for "/usr/bin" [code[There is NOTHING that stops me from creating a folder in that location. mkdir /boot/Apps. mkdir /boot/setup I called C:\ similar to /boot because it contains files analogous to autoexec.bat, command.com, etc... are located. No you can't. You need to be root to even try something like that. "/" is not "C:". C: is a partition. /, and all folders "in" it, are actually representations of where to look for files in a partition-independent manner. I said "~/" not "/". Although, incidentally, "/" is a partition, although it can sym-link to other partitions for its sub-folders. Still, I've yet to see a distro where the sub-folder of "/" don't at least have a matching sym-link file in the boot partition. If you want to make a folder there, put a folder there. Nothing, except root permissions, stops you. except root permissions That's not nothing. Besides, I can not not put a "/usr" folder or at least sym-link in there. If it must, it should put all of these under a "/linux" folder. And it does. Launch the program already. Can't, have to first move it to "/usr/bin". And give it executable permissions. And it will still only have the subset of the harddrive linux allows it to access, rather than the one I would want it to access. No, it makes for an easy recovery should you screw something up. /home partition of it's own can be used by several Linux distributions. Should you choose to format it as FAT32, it makes it much easier to share files between Windows and Linux. Woo, server thinking all the way, huh? It's my personal computer, not a bank server, I don't need 100 distros, multi-boots, or 1000 layers of redundancy. The fact that, worst case scenario, I can F5 at boot-up then reinstall Windows, is all the redundancy I need. And quite frankly, it's all the disk space I'm willing to invest for redundancy. Translation: I'm willing to use Linux, but not if I might have to spend 20 minutes learning something new, because I'm stubborn. No, it means: "I'm willing to install KernelEx or some unofficial DLL to improve my system speed, but not to repartition my harddrive just so that I can put my files on it". I did that testing Seamonkey Ooh, great example. Except that I have that one on my Windows as well - it needs to support minimal footprint by-design. More common Linux-only programs, like Gnu, tend to have a much larger footprint, andn ot too much versatility. , since Ubuntu doesn't offer it in their package manger. The package manager's job is to make sure you have the necessary libraries installed for your programs, such as So let me get this straight: you want each program in it's own folder on the C: drive, but you can't tolerate the same chaos in your home directory. Technically, I have them in "C:\Apps\program_name", but that's not the sort of chaos I'm talking about. What I'm talking about is my ability to create "C:\Docs" and "C:\Docs\document_type", instead of having "~/program_name" and "~/document.txt" sharing the same home parent folder. Or how about having "C:\Apps\program_name\config.xml" instead of "~/.program_name.log"? Why should folders and files even have the same parent directory? Why shouldn't files generated by a program not be present in the same folder as the program, assuming the program isn't shared by 100 users who need separate configurations? They're configured somewhere, aren't they? You can log on to them, or run programs as them. And even if they don't actually own any files, they can, can't they? Then what makes them different from an actual account? Not being used? No. Because files have absolutely no footprint at all on system performance unless they are being accessed. Hello? Diskspace? File-system structure? Files are by definition foot-prints. A few bytes in the RAM or a few cycles in the CPU, I couldn't care less about those, but bytes on my permanent media? I can't even think of anything else that could be called "foot print" (You know, save for an actual print of a foot). Even when they are being accessed, they are only attributes, just like "hidden" or "archive." Permissions may be flags, but ownership is not. If it was just the flags, it wouldn't be such a problem. And what, exactly, is the difference in performance between a permissions system that checks "Is this a system file? No? Then he can touch it." vs, a system that says "Is this a root file? No? Then he can touch it"? Overhead? Seriously, some of your questions show a lack of basic perception of computer concepts. Then again, when all low-level stuff are abstracted and locked-away for you, it's understandable if you then fail to understand them. By "workarounds" I meant people need to "add stuff" to get access to the stuff they want to do. The "stuff" they add end up opening security hole bigger than those in a completely open straight-use system (You know, stuff that also allow remote access). Once that happens, the insecure system is definitely better. Yes it can. If you want, a minimal kernel can easily be installed. USB, graphics, etc... can all be excluded from a kernel and loaded as modules at boot time as needed. And all you need is to set up a dual boot to enable this "safe mode". Seeing as how Linux doesn't exactly have "step-by-step confirmation" on its boot process. There is nothing that stops you from having a kernel with almost no drivers compiled into it and selectively loading modules in an interactive startup. I dare you to make that happen. Seriously, just try. No it doesn't. Almost every program I have ever used in Wine either worked without any configuration, or didn't work at all. With the latter being the more common case. That's no different than what many people have to do to run programs under compatibility mode in Windows XP, or patch other things in Windows 98. Well, I've never actually had to do any of those, so I wouldn't know. My programs "just run" whenever I want them to. "sudo apt-get install program" is hardly a difficult process. You can't apt-get Java 1.6, you need to alien the rpms for it to work. Getting alien wasn't that simple either. The multi-standard environment of package managers really makes things difficult. Suddenly one manager doesn't realize which components were already installed by another. What ever happened to "where is your library_name installed"? Your example shows the opposite. The only slowdown Windows has for locating and installing a program already endorsed by the Linux repository is clicking "Next". Using the repository is like using Windows' "Add/Remove Windows Components". Quick, but a limited choice. Your example shows Windows' slow method (actually having to look for the software) is still faster than Linux's fast method. Mind you that if I Googled for Linux software, it would take me a whole lot more than a few "next"s to get it working. Downloading and installing Java took me far less time in Linux. Hmm... Let's see. java.sun.com -> Download -> Accept license agreement -> Run downloaded file -> Restart. On the other hand, java.sun.com -> Download -> Accept license agreement -> Save downloaded file... -> chmod -> Synaptic -> Password -> Install alien -> Sudo -> alien Java -force. Not to mention the prior allowed me to choose a folder first. Alternately, I could run Windows, use WinRAR to unpack the binary into a folder of my choosing (under the root, [i]not[/i] stuck in some "home" sub-folder forced upon me by the OS), then open Explorer, and go click the exe. Plus, how much "specialty" software do most people run? I am not "most people". "Most people" should by using Macs anyway - they're comfortable, stable, give all the basics (and little more), and work perfectly for people only interested in reading their eMail. The very things Windows can do that a Mac can't, are the things Linux can't either, and I'd dare you to claim Linux is easier to use than Mac. You missed the point. "Out of the box" does not apply to Linux since it does not come in a neat little box. First, you have to choose one from hundreds of distribution, then download one of the 5 ISOs it offers (without explaining the advantage and usage of each), probably using a torrent client, then burn the CD-- But wait, you're not done yet. Even if you got the CD from a friend (or, *gasp* purchased it, a la RedHat), you still need to start up the package manager, choose the GUI server you want to use, go screen after screen of selecting "module or built-in" for kernel drivers, and about 4 days later you may have it installed. Or, you could pop Windows in the CD and just click "Next". On the plus side, Linux has Live CDs which are better than installation. It's what I use whenever I need to test something with Linux. Still, that's just another reason to keep Linux on the CD rather than my harddrive. And that's Ubuntu. Ubuntu is designed to include what the vast majority of people want. Other installers do allow you to pick and choose which software you want installed. Like Debian? Don't even get me started on that one. There are plenty of DVD rippers for both platforms. The fact that you desperately cling to one is an artificial limitation, not a practical one. Ripper5 is not a DVD ripper. I don't know where you got that silly idea. If I want to "rip" a DVD, I just use Nero like any normal person. At any rate, I've yet to see a suitable replacement for Ripper5 on any platform. Linux does not hide any hardware information. You just don't know where to look. It abstracts the very fact that you have hardware. You may be able to "look up" what processor you have, for example, but Linux will do its best just to make sure you don't even realize whether it's x86 or SPARC. Macs have every single "problems" you have listed, and then some. File system structure, where programs are installed, etc... Exactly, which is why I don't switch to Mac. Mac is perfectly suitable for any end-user that is not actually a low-level user. That's pretty much its gimmick. Linux doesn't even have a chance of stealing Mac's flame in that regard. Windows tries, but it's also not much of a match. I hope I needn't explain to you why Mac is perfect for the spoon-fed, and how far other OSs fall behind and did respect. So, once you exclude the spoon-fed from the market, Linux suddenly gets the short end from both sides: Not as comfortable as a Mac, not as software compatible as a Windows. Linux can and usually does corner the market for servers and embedded software. That's its domain, and that is where it must and will inevitably stay. It's just not an end-user system. For the non-spoon-fed end-users, Windows becomes the main, if not only, choice. You might put it like this: -Windows is C++. -Linux is Python. -Mac is Visual Basic. Again, I'm basing it on long articles comparing use-cases in various OSs. Just Google "why linux isn't for the average user" or similar, and you'll find plenty. Nothing speaks louder than use-case to me. How many does it take? 5? 100? 1000? Saying you can't play games on Linux is like saying you can't play games on a Wii, because Final Fantasy VII won't run on it. You make a general statement, and point out very specific examples as to why it's not true. And the PlayStation can't run Smash Bros. But while very few commercial game companies make Linux ports, I've yet to heard of one that skipped the "Windows port" on a PC game. Sure, Klondike for Linux only runs on Linux. Same with its Galaga clone, and "Shoot Bill Gates" game. But games you'd hear about in IGN or GameSpot, or see in E3 are simply not Linux targeted. They may be Windows targeted, or console targeted, but never Linux. It's just not a target market. Since when does Linux count as "throwing out your computer?" If I'll only be using it for e-mail and web-browsing, it would be the same as just using a cell-phone, no? In that case, it would exactly count as "throwing out my computer". The alternative is to double-boot, but why should I when Windows reads e-mail just fine? And I know of several programs where the variety (or at least free variety) programs do not exist: audio workstations. There are far more programs for that than for Linux. I dunno, my GoldWave works just fine. Same with my Acoustica, and all AnalogX software. Also have a bunch of graph-styled synthesizers, although I find them a tad difficult to manipulate. The FSF has nothing to do with either ease of use or the availability of programs. And there I thought the "free" in FSF stood for "freedom" (isn't that their motto?). I would think availability of all-purpose programs on PCs, as in contrast to, say, consoles, would be the main goal of the open-source community. You know - you want something, just develop it. That Java, Sun, and IBM have been, so far, doing better in that respect than the zealot open-source community is another matter. In another repository. Medibuntu, for instance, exists because Ubuntu legally can't bundle some programs due to copyright restrictions. Adding these repositories is rather easy. Assuming they are compatible with your package manager. Again, standards are abundant as far as Linux is concerned. Alpha / beta software. You probably don't want it unless you're a developer, in which case you are assumed to be competent enough to install something. Isn't the greater part of open-source software defined as "alpha / beta" by default, even years after its release? Many of the programs are used are dubbed "beta" somewhere on some level, but they are still fully working stable software. Besides, that I have an easier time installing VLC beta on Windows than on Linux definitely raises an eyebrow, especially considering it doesn't even run properly on Windows. By contrast, Windows makes it very difficult to install something because it is either A. Shareware B. Infested with Spyware. C. Doesn't exist Are you kidding? I can get one-click install software from source-forge, for crying out loud. Do you even realize how much easier it is to release a binary for Windows, compared to releasing something for Linux? And then you claim Linux can ever have more specialty software? How could it when everything that is not "big" and "mature" enough to be in some major repository is too difficult to install? I agree that you should be allowed to choose what software you want to run. However, I also think that you should be allowed to know alternatives exist, instead of perpetuating FUD. One need not find, learn, and configure a replacement for each of their programs just for the questionably valuable purpose of "switching an OS". You can't honestly claim alternatives from other OSs are "better", even if you use "it's for that 'better' OS" as an argument. The fact that you will need to switch every last one just to switch an OS - a goal with questionable benefit at best, makes the entire thing not worthwhile. A. a minimal footprint in many cases. Although I'm sure Windows 95 fits into 3 MB of RAM and Linux does not, performance on 9x does not scale well, so an outdated machine of, say, 64 to 128 MB will perform just as well in most cases with Linux as with 9x. Plus Linux can easily utilize much more memory and partitions than 9x could ever hope to accomplish. The lack of overhead is negated by the fact that it can't do anything with the free resources. That Windows, in its entirety, is held in one folder on my partition, and that it doesn't touch all other folders on that partition, is what I consider "minimal footprint". The Linux requires rearranging the entire folder structure of the partition is what I consider to be the exact opposite. B. A flexible file structure. Your insistence that you are forced into /, /bin, /usr, etc... is completely flawed. Those aren't limitations of Linux, but rather an agreed standard. Why can't I install Windows on my Ɣ:\> drive? You can't install Windows on "Y:"? Why not? Never stopped me. You'd still need the MBR and MBR-helper files on the main partition if you want to boot from it. As for standard, to me, the only standard should be that my files can go in "C:\", and my OS can stay the hell away from it. Linux just doesn't allow that. Instead of Linux being in my file-system, my file-system is in Linux. An OS should limit what it forces upon my file-system to the MBR.