Talk:Control-Alt-Delete

From Academic Kids

I'm not sure the term "hook" is used here correctly. A hook is usually something a piece of software provides to allow it to run other software that wasn't anticipated when the first bit was written. The second part is "hooked into" the first. Ctrl-Alt-Del may cause a routine in BIOS to run, but that wouldn't be a hook. In fact, the modern events that occur on Ctrl-Alt-Del is more accurately described as a hook into this original routine than the other way around. Clear? Probably not... GRAHAMUK 13:15, 27 Aug 2003 (UTC)

Hi, I defer to you on whether hook is the correct term here. Since you seem to be well informed about such things, perhaps you create a stub on hook.

Also, the line:

This command is a hook to the BIOS of a PC running DOS;

is the qualification "running DOS" really necessary? The CAD "hook" is there in the PC, in the BIOS, no matter what OS is running. Or not? Presumably it is also there under Linux (but what Linux does with it I don't know). (FWIW, I use OS/2 where CAD still has its original function.) Cheers, -- Viajero 13:44, 27 Aug 2003 (UTC)

I think technically, the command causes an Interrupt in the BIOS. But as Tarquin says below, this whole page might be better off merged under Three-finger salute - although I note that doesn't refer to BIOS interrupts either. I may even be wrong that that's what it does... - IMSoP 19:01, 31 Jan 2004 (UTC)

---

What do we do about this article and Three-finger salute? Merge, or keep this as a particular case of Three-finger salute? -- Tarquin 13:49, 27 Aug 2003 (UTC)

Hard reboot?

I'm pretty sure that control-alternate-delete causes a soft reboot, and the page on soft reboot seems to confirm this.

TJSwoboda 23:14, 28 Nov 2004 (UTC)

Some rather verbose justifications and apologies.

• In an attempt to avert any conflict between myself and Uncle G, I'd just like to explain why I made some of the edits that he has "reverted" (in reality, I consider them more in the line of "corrections", and am thankful for them):
  • The distinction I was drawing attention to was between a "MS-DOS and real-mode systems" and "Windows 95 and its successors". While I couldn't quite remember, I had the feeling that Windows 3.x was sufficiently reliant on MS-DOS that it would fall in the former category; as such, I renamed the latter. Obviously, with the inclusion of Windows 3.x in the latter section, the original heading ("DOS-based Windows") makes sense, but the distinction I labelled wasn't "false", I was just labelling the distinction that previous editors had made. Now you mention it, I think I do remember blue screens in relation to ctrl-alt-del, so thanks for that.
    • Windows 95, Windows 98, and Windows Me all rely upon DOS, too. The problematic section title is, in some ways, the first one rather than the second one. There's simply no way to explain the complexities of DOS extenders and Virtual 8086 mode in a section heading. Uncle G 19:43, 2005 Jan 17 (UTC)
  • The "32-bit" mention was, in retrospect, unnecessary, and I know it's a vaguely debatable point; I just wanted to draw a clear line between two very different generations of Windows, and that's one short-hand commonly used for that distinction.
    • "vaguely" is an understatement. I was there in the mid-1990s when these debates were in full swing. It's very debatable. There were then (and still are now — I encountered someone making the same old debunked arguments a couple of months ago.) people who blithely believed the Microsoft marketing blurb, which to those who understood operating systems was quite inaccurate. Uncle G 19:43, 2005 Jan 17 (UTC)
  • My point about "normally only the superuser" having permission was intended to convey the rather subtle point that there are all sorts of ways of programmatically granting superuser rights - as a simple example, if I type "sudo halt", I may have been granted the right to issue that one command as "root" in /etc/sudoers. If so, I can programmatically halt the machine without "being" the superuser, except in the most [almost pedantically] technical sense. Like I say, it's a subtlety, and my edit didn't really convey that.
    • That sudo is involved doesn't change the basic fact that it is a process running under the aegis of the superuser account that is performing the actual task. In contrast, one doesn't even need to be logged in, and running any processes at all, in order to use Control-Alt-Delete. Uncle G 19:43, 2005 Jan 17 (UTC)
  • I only removed that link to Microsoft Windows because it "felt like" a duplicate - I now see that that exact article hadn't been linked before, but it still feels a bit redundant with all the earlier mentions. Plus, now I look, it mentions "DOS and Windows" in a paragraph about Windows NT, so it's only really referring to "non-NT versions of Windows". I hope the new wording seems better.
    • Actually, it's referring to a rather woolly notion of "whatever Microsoft operating system the Windows NT user used before". Windows NT users will have upgraded from operating systems where Control-Alt-Delete soft rebooted the system. I know that some Windows NT users upgraded from DOS. The structure of Wikipedia currently is that Microsoft Windows is the main article and there are articles on Windows 95, Windows 98, Windows 2000, and so forth. It's unfortunate that Wikipedia does not make a higher-level subdivision between "DOS-based Windows" and "Windows NT". (Which reminds me: I must correct the errors that implied a single line of development, instead of two that ran in parallel for over a decade, that I saw in DOS the other week.) Uncle G 19:43, 2005 Jan 17 (UTC)
  • Finally, the introduction: I feel it's important that we introduce what this article is actually about, and how it relates to the old use for soft reboot; hence I've reinserted the mention of "More advanced operating systems" (an ugly phrase, but I can't think of anything better) doing more complex things with it. Otherwise, there's no justification why there isn't an article discussing how different applications treat "Control-A" (OK, that's a deliberately silly example, but without the explanation the "introduction" doesn't do much "introducing" of what the article's there for).
    • It is an ugly phrase. "Operating systems that don't use the BIOS to drive the keyboard" is more correct, albeit that I am loathe to put that into an introductory section. The original point was merely that the keyboard combination comprises the keys that it does as a safety precaution because the action invoked as a consequence of it was a drastic one. Of course, that led on to the point that with the change to the keyboard, this deliberate difficulty went out of the window (which evoked comment at the time). Uncle G 19:43, 2005 Jan 17 (UTC)
• On a slightly seperate note, a quick look around turned up some slightly odd things about system.ini - one website stated that "LocalReboot defaults to Off on Windows 95" (which seems rather bizarre); and another referred to a different setting called something like "KeybdReset", which seemed to have exactly the same purpose. I'll have to look again, and work out what's going on with that, but neither was a Microsoft reference, unlike yours, so they could just have been erroneous.
  • Microsoft KnowledgeBase articles aren't always reliable. (One of my most widely cited articles contains a harsh criticism of a Microsoft KnowledgeBase article for its inaccuracy.) However, I have little reason to suspect the accuracy in this regard of the one cited. There's a lot of other, contradictory (and self-contradictory), information circulating about those settings, which I turned up as you did, but a lot of it appears to be distorted due to the chinese whispers effect. Uncle G 19:43, 2005 Jan 17 (UTC)
• Oh dear: you can tell when I'm tired, because I become more verbose in discussions! Sorry. :-/ - IMSoP 01:27, 8 Jan 2005 (UTC) (where UTC==localtime)