Talk:Operating system: Difference between revisions
imported>Martin Baldwin-Edwards |
imported>Karl D. Schubert |
||
(11 intermediate revisions by 7 users not shown) | |||
Line 1: | Line 1: | ||
{{ | {{subpages}} | ||
}} | |||
== Bundling of drivers, GUI and apps == | == Bundling of drivers, GUI and apps == | ||
Line 54: | Line 44: | ||
::: I've got a couple old books here on vintage operating systems, SCOPE, CTSS, Multics and UNIX. Not sure how much I can help on the history part though. The book is, Operating System Concepts (Peterson, Silberschatz) 1985--[[User:Paul Derry|Paul Derry]] 23:36, 26 April 2007 (CDT) | ::: I've got a couple old books here on vintage operating systems, SCOPE, CTSS, Multics and UNIX. Not sure how much I can help on the history part though. The book is, Operating System Concepts (Peterson, Silberschatz) 1985--[[User:Paul Derry|Paul Derry]] 23:36, 26 April 2007 (CDT) | ||
:::: A couple of comments, here. MULTICS was *not* an IBM system. IBMs systems in this "interactive" arena were: VM (the original one called CP-67 and CMS (for Cambridge Monitor System) and a whole series of them to today's present-day VM system. There was also a an interactive component on MVS called TSO (Time Sharing Option) and a few predecessors such as Call-OS on MVT. There was also ACP (the Airline Control Program) and TPF (Transaction Processing Facility, it's successor), and then a slew of Unix-based systems such as AIX (Advanced Interactive Executive), and IX/370 (based on ). And, for the record, Multics (based on CTSS) was of MIT origin and ran on an IBM 7094. But, IBM was not directly involved. (Here's a short history: http://www.multicians.org/history.html , unverified) And, for those who are into the smaller and even more specialized OSs, there's MUSIC. There's a whole slew of other operating systems from those times that had major impacts on follow-on interactive and personal systems such as GECOS (GE Computer Operating System) that was used by both GE and HISI (Honeywell Information Systems Inc.) on their mainframe computers and it's time sharing facility, TSS (Time Sharing System). Then, on the smaller side, IBM had projects that put both VM and MVS into PC-sized chassis with one "side" of them a PC and one "side" of them a mainframe (XT/370 running VM or "Percheron" or "Washington" were the products). They were used by developers to develop mainframe applications without a mainframe. And, of course there were other midrange systems such as DOS/VSE and DEC's VAX and predecessor PDP-11 OSs. And, the supercomputer OSs for the CDC and Cray systems -- to name a few on the other end of the spectrum. | |||
:::: As an editorial comment, I don't believe that articles should have a "slant" to them and I would be concerned that we get "...[anyone's -- you put in the name] huge failed attempt...." is prejudicial. Not all systems were designed for commercial use (e.g., MULTICS) and some were but did not last the test of time for whatever reason. [[User:Karl D. Schubert|Karl D. Schubert]] 15:21, 19 September 2009 (UTC) | |||
==brain dump== | ==brain dump== | ||
Line 99: | Line 93: | ||
:Or is it about how doctors perform surgery? We should also have an article Operating Room:-))--[[User:Martin Baldwin-Edwards|Martin Baldwin-Edwards]] 04:23, 28 August 2007 (CDT) | :Or is it about how doctors perform surgery? We should also have an article Operating Room:-))--[[User:Martin Baldwin-Edwards|Martin Baldwin-Edwards]] 04:23, 28 August 2007 (CDT) | ||
::I will think about this, but I must say that in my 50+ years, I haven't heard very much conflicting use of the term "operating system" that would not be clear from context, and the vast majority of use pertains to this one. Would any of these others require a standalone article? If so, let ''those'' get modified. To turn the phrase "operating system" into "computer operating system" would seem very awkward. In addition, many links would need to be adjusted. So let's think about it a little more, before doing anything. In the meantime, I have modified the opening sentence to make sure that the computer context is clear.[[User:Pat Palmer|Pat Palmer]] 19:58, 28 August 2007 (CDT) | |||
:::Another thought--the acronym is OS, not COS. Modifying the title not to match the acronym seems odd. But again, I'll think on it. I'm just not sure at present.[[User:Pat Palmer|Pat Palmer]] 20:00, 28 August 2007 (CDT) | |||
Right, the relevant question here is not whether one can find ''some'' alternative use for the title phrase, but whether there is any use that is ''actually'' so common that it might actually be confused with this one. This requires actual linguistic knowledge, not mere guesswork. Moreover, on the supposition that we ''did'' want to disambiguate, then if "computer operating system" is rarely actually used, the better title would be "operating system (computers)". --[[User:Larry Sanger|Larry Sanger]] 21:23, 28 August 2007 (CDT) | |||
:I thought the expanded tile would be clearer from the start. I suppose any disambig will come when and if required. —[[User:Stephen Ewen|Stephen Ewen]] [[User talk:Stephen Ewen|(Talk)]] 22:29, 28 August 2007 (CDT) | |||
::ON reflection, I think the most standard usage must be for computers, and operating system (computers) removes any possible ambiguity. There must be some people who don't know that computers have operating systems, and this title would be helpful for them, too.--[[User:Martin Baldwin-Edwards|Martin Baldwin-Edwards]] 01:17, 29 August 2007 (CDT) | |||
:I'm not happy about using a name with parentheses in it (difficult to get links right). I'd prefer "computer operating system" to "operating system (computer)" if we must rename the article. Also, someone will need to deal with all the existing computer-related links to the current name so they don't just go to a dismbig. page.[[User:Pat Palmer|Pat Palmer]] 22:01, 30 August 2007 (CDT) | |||
== history details == | |||
* In "The dawn of the Personal Computer in the 1970s and 1980s" how can you leave out the C64? It practically defined "home computers" in the 80s. | |||
* Did Unix really come before Multics? I was always certain it was the other way around. | |||
--[[User:Tom Vogt|Tom Vogt]] 16:58, 1 February 2008 (CST) | |||
::Good spotting! Do you have time to fix it? We need help authoring articles like these, which are often still in a state of some dishevelment. Dig in, please![[User:Pat Palmer|Pat Palmer]] 21:04, 2 February 2008 (CST) |
Latest revision as of 09:21, 19 September 2009
Bundling of drivers, GUI and apps
I disagree with the second sentence of the article "These components are almost always bundled together". Linux does not bundle GUI nor apps. Windows does not bundle many drivers (which are instead distributed by the device manufacturer). Alternatives:
- "These components are often bundled together".
- "These components are almost always distributed together".
--Markus Baumeister 14:18, 14 March 2007 (CDT)
Most Linux users don't build from scratch, they use distros, which do in fact bundle GUI (KDE/Gnome on X11) and apps with the kernel and drivers. --ZachPruckowski (Speak to me) 14:32, 14 March 2007 (CDT)
- Well, as you say "they use distros". That's why I made the second alternative proposal. "Bundling" in my opinion has way to much connotation of "made into one". And GUI and specifically apps are not in general part of the operating system. --Markus Baumeister 15:02, 14 March 2007 (CDT)
- OK but you're blurring the distinction between "Operating System" and "kernel." Linux by itself is just a kernel. A distro has to be made for Linux to be considered an "operating system." --Eric M Gearhart
A distinction that's been all that lost with the ascendency of Linux is the distinction between an operating system and an operating system kernel. I don't think anyone would seriously argue that a desktop environment belongs in the kernel, but it might be considered part of the operating system (though I'm not so sure that would do so). Drivers are more interesting in that I think they clearly do belong to the OS. Consider that at one point it was necessary to re-compile the kernel to install certain drivers. Today, operating systems have become more modular. Now applications like editors that are often bundled with operating systems cleaarly are not part of the OS. I think this pretty much uncontroversial. Greg Woodhouse 14:18, 27 April 2007 (CDT)
Unix user to the front
I tried to fix some of the MS Windows centered definitions from the beginning of the article but there are some more in the GUI and especially the Apps section (Breaking down the App-OS barrier with "Browser as part of the OS" was the MS defense line in their monopolistic trials). I'm running a bit out of time, maybe someone else can do it. --Markus Baumeister 15:02, 14 March 2007 (CDT)
I can take a look, as I use Linux exclusively as a desktop --Eric M Gearhart 08:10, 7 April 2007 (CDT)
web browser importance?
The last section about the "fifth generation of computers" seems like someone just made it up, or lifted it out of some optimistic hypey pop article. browser-based apps are still a relatively small niche, compared to actual local apps. computers do /not/ come with less and less software, but rather with more and more (linux distros include a bunch of productivity apps with default install, windows now includes firewall, antivirus, antispyware, and many vendors, such as dell, will install a bunch of trialware for office, ISP, antivir, etc.) so basically, i don't buy it, but before i go and delete or reword, i wanted to get some comments to make sure i am not missing anything. :) --Dan 01:44, 15 March 2007 (CDT)
Nah, I see what you mean. Really, I'd like to see the article rewritten, but right now it's too early in the morning for me to do anything about it. ZZzzzz... --Paul Derry 01:51, 15 March 2007 (CDT)
changed status back to 2
This article is a good start, but I feel it still needs work (and I for one won't get to it right away). So I've revised the status from 1 down to a 2.Pat Palmer 22:41, 9 April 2007 (CDT)
I think we need "history of operating systems" too
I'd like to see a section on evolution of operating systems, starting with batch systems (from the 1960's with tape input, then card input, and only paper output. Followed by IBM's huge failed attempt to write an "interactive" OS (Multics), trumped notably by the smaller Unix system developed soon afterwards at Bell Labs. Finally, the advent of personal computing and the resulting tightly coupled hardware-software wars (Apple, Microsoft, and a few others). There's some real high drama here if we can bring it to the forefront. Perhaps even a graphic with a timeline could be created if someone really goes crazy.Pat Palmer 18:54, 23 April 2007 (CDT)
- Pat, I've got the beginnings of that "in my head" so to speak. As I researched the evolution of Unix I came across a lot of info on the evolution of operating systems in general, from the "Compatible Time Sharing System" (CTSS) and "INcompatible Timesharing System," to MULTICS, to UNIX. If the mood strikes me I'll try and get a History section going here. Eric M Gearhart
- super!Pat Palmer 10:17, 24 April 2007 (CDT)
- Well I ended up rewriting or adding a lot to a bunch of parts... not as much to History as I'd like. I'll try and get some more content going tomorrow. The framework is there though. I did add a lot to the Kernel and Drives section though Eric M Gearhart
- I've got a couple old books here on vintage operating systems, SCOPE, CTSS, Multics and UNIX. Not sure how much I can help on the history part though. The book is, Operating System Concepts (Peterson, Silberschatz) 1985--Paul Derry 23:36, 26 April 2007 (CDT)
- A couple of comments, here. MULTICS was *not* an IBM system. IBMs systems in this "interactive" arena were: VM (the original one called CP-67 and CMS (for Cambridge Monitor System) and a whole series of them to today's present-day VM system. There was also a an interactive component on MVS called TSO (Time Sharing Option) and a few predecessors such as Call-OS on MVT. There was also ACP (the Airline Control Program) and TPF (Transaction Processing Facility, it's successor), and then a slew of Unix-based systems such as AIX (Advanced Interactive Executive), and IX/370 (based on ). And, for the record, Multics (based on CTSS) was of MIT origin and ran on an IBM 7094. But, IBM was not directly involved. (Here's a short history: http://www.multicians.org/history.html , unverified) And, for those who are into the smaller and even more specialized OSs, there's MUSIC. There's a whole slew of other operating systems from those times that had major impacts on follow-on interactive and personal systems such as GECOS (GE Computer Operating System) that was used by both GE and HISI (Honeywell Information Systems Inc.) on their mainframe computers and it's time sharing facility, TSS (Time Sharing System). Then, on the smaller side, IBM had projects that put both VM and MVS into PC-sized chassis with one "side" of them a PC and one "side" of them a mainframe (XT/370 running VM or "Percheron" or "Washington" were the products). They were used by developers to develop mainframe applications without a mainframe. And, of course there were other midrange systems such as DOS/VSE and DEC's VAX and predecessor PDP-11 OSs. And, the supercomputer OSs for the CDC and Cray systems -- to name a few on the other end of the spectrum.
- As an editorial comment, I don't believe that articles should have a "slant" to them and I would be concerned that we get "...[anyone's -- you put in the name] huge failed attempt...." is prejudicial. Not all systems were designed for commercial use (e.g., MULTICS) and some were but did not last the test of time for whatever reason. Karl D. Schubert 15:21, 19 September 2009 (UTC)
brain dump
Eric, I just dumped a few possible new sections into the article. You've made a very good start on this huge huge topic. Keep up the good work, and change any of the stuff I added that you need to.Pat Palmer 00:06, 27 April 2007 (CDT)
- Pat I cleaned up Operating_system#Booting_Up some... take a look at the easy way to do lists. I stole that from CZ:How_to_edit_an_article#Examples Eric M Gearhart
potential snag
The whole problem with this is, that to understand any of OS, you need first to understand virtual memory. In fact, I fear that the memory management and virtual memory articles ought to be in place first. It's all snarled together in my mind right now, and I can't see straight enough to explain it even to myself, let alone to anyone else. It's that problem of thinking I understand something, but then, when I try to write about it, I find I just don't know enough after all. So I quit (for now anyway!).Pat Palmer 00:26, 27 April 2007 (CDT)
Yes I hit this same wall when I try to get the Unix article going. We need a hierarchy of interrelated articles, something like:
- History of Computing
- Mechanical computers
- The Vacuum Tube "Age"
- The Transistor "Age"
- Mainframes
- The dawn of the PC
- History of Operating Systems
- CTSS
- Multics
- Unix
- Operating system
- Virtual Memory
- Memory Managemenrt
- Kernel
All this stuff is swirling around in my head as well... I feel your pain Pat, and I don't hold the Masters degree you do lol! Eric M Gearhart
history of operating systems
I want to break this part out into a separate article. How about we make this article about "understanding what an OS is and does" and then have a separate article, history of operating systems, that can be pointed at both by this article, and also from history of computing?Pat Palmer 10:52, 13 May 2007 (CDT)
what if we moved the image down?
Eric, I like the diagram. How about we try moving it down below the boot/operation/shutdown phases part, and put it where we start explaining what a kernel is? There are terms in the diagram and a lay reader will not understand at the top of the article. Pat Palmer 11:22, 13 May 2007 (CDT)
- I only like an image at the top if it adds something to the introduction. Otherwise, it narrows the reading width and distracts from the opening. I (vainly perhaps) think we have a pretty good opening, and I'd prefer to keep it highly readable. Pat Palmer 11:25, 13 May 2007 (CDT)
- P S - If it's not too much work, could the diagram just say "processor" and not "CPU or processor"? I've been going around changing occurrences of CPU to processor (keeping to link to CPU for now) because, again, lay people may find CPU more intimidating that processor, and technically they mean pretty much the same thing. Pat Palmer 11:25, 13 May 2007 (CDT)
- I only like an image at the top if it adds something to the introduction. Otherwise, it narrows the reading width and distracts from the opening. I (vainly perhaps) think we have a pretty good opening, and I'd prefer to keep it highly readable. Pat Palmer 11:25, 13 May 2007 (CDT)
- Pat I'll take a look at changing it to 'Processor' when I get back to my laptop... I'm at work right now and I don't have OOo Draw --Eric M Gearhart 08:16, 14 May 2007 (CDT)
Operating system - of what?
Seems this should be called Computer operating system. Heavy equipment operators have operating systems on how to handle their machinery, for example. —Stephen Ewen (Talk) 01:59, 28 August 2007 (CDT)
- Or is it about how doctors perform surgery? We should also have an article Operating Room:-))--Martin Baldwin-Edwards 04:23, 28 August 2007 (CDT)
- I will think about this, but I must say that in my 50+ years, I haven't heard very much conflicting use of the term "operating system" that would not be clear from context, and the vast majority of use pertains to this one. Would any of these others require a standalone article? If so, let those get modified. To turn the phrase "operating system" into "computer operating system" would seem very awkward. In addition, many links would need to be adjusted. So let's think about it a little more, before doing anything. In the meantime, I have modified the opening sentence to make sure that the computer context is clear.Pat Palmer 19:58, 28 August 2007 (CDT)
- Another thought--the acronym is OS, not COS. Modifying the title not to match the acronym seems odd. But again, I'll think on it. I'm just not sure at present.Pat Palmer 20:00, 28 August 2007 (CDT)
Right, the relevant question here is not whether one can find some alternative use for the title phrase, but whether there is any use that is actually so common that it might actually be confused with this one. This requires actual linguistic knowledge, not mere guesswork. Moreover, on the supposition that we did want to disambiguate, then if "computer operating system" is rarely actually used, the better title would be "operating system (computers)". --Larry Sanger 21:23, 28 August 2007 (CDT)
- I thought the expanded tile would be clearer from the start. I suppose any disambig will come when and if required. —Stephen Ewen (Talk) 22:29, 28 August 2007 (CDT)
- ON reflection, I think the most standard usage must be for computers, and operating system (computers) removes any possible ambiguity. There must be some people who don't know that computers have operating systems, and this title would be helpful for them, too.--Martin Baldwin-Edwards 01:17, 29 August 2007 (CDT)
- I'm not happy about using a name with parentheses in it (difficult to get links right). I'd prefer "computer operating system" to "operating system (computer)" if we must rename the article. Also, someone will need to deal with all the existing computer-related links to the current name so they don't just go to a dismbig. page.Pat Palmer 22:01, 30 August 2007 (CDT)
history details
- In "The dawn of the Personal Computer in the 1970s and 1980s" how can you leave out the C64? It practically defined "home computers" in the 80s.
- Did Unix really come before Multics? I was always certain it was the other way around.
--Tom Vogt 16:58, 1 February 2008 (CST)
- Good spotting! Do you have time to fix it? We need help authoring articles like these, which are often still in a state of some dishevelment. Dig in, please!Pat Palmer 21:04, 2 February 2008 (CST)