osdev

[yomexzo]

Recently, I was challenged on my knowledge about operating system development. My lecture gave me an assignment. It says "how does computer architecture affects and informs the development of operating system? In giving your answer, specify with as much detail as possible the relationship between the computer hardware architecture and the operating system development". I have searched virtually everywhere but no luck. friends, please help me i have to submit the assignment in a couple of days..........................

[leledumbo]

I don't think there's a direct answer to this, instead you must gather articles (e.g. from the wiki), read, and connect them as you need.

[tantrikwizard]

Recently, I was challenged on my knowledge about operating system development. My lecture gave me an assignment. It says "how does computer architecture affects and informs the development of operating system? In giving your answer, specify with as much detail as possible the relationship between the computer hardware architecture and the operating system development". I have searched virtually everywhere but no luck. friends, please help me i have to submit the assignment in a couple of days..........................

I regret that these sort of questions would come from a C.S. professor and not an OS developer. The difference between the two should be well clarified. One must take into consideration that those that can do...those that cannot teach. It should be well understood that your professor is a professor because he cannot be a proficient programmer. It is more often the case that the illiterate failures, who cannot contribute in any form to the field of OS design ,resort to some form of 'educational' aspiration rather than it's practical implementation and you likely must take the stupidity of your instructors misguided informed ignorance into account when presenting any paper.

With that said, and complete contempt of 'educators' outlined, consider the illiterate nature of the 'college educators' ability. Your professor is a professor because he/she does not understand the topic which he/she is trying to teach. Therefore any comprehensible explanation will suffices in satisfying the illiterate mind of your professor. Make something up, it doesn't matter, you're dealing with an illiterate educator, not an OS developer. If imagination eludes you then seek the advice of the local childhood daycare center, you could contrive enough nonsense to keep your illiterate professor dumbfounded for weeks.

[JackScott]

I'm not sure I agree with that, tantrikwizard. All the people I know personally who have PhDs and/or teach in universities are very intelligent people. Based on my personal evidence, I can't see where your claim is coming from. Do you have any evidence to the contrary?

A good well-known example is Andrew Tanenbaum. He has both taught at a University (IIRC, still does), and has written a very nice microkernel implementation. He seems the sort of person who contradicts your claim.

[Combuster]

This wouldn't be too far from an exam question for systems architecture. Try going around thinking what an OS does (or needs to do), and what the hardware does, and how these two have been influencing each other over the years. (write down a list of new instructions and cpuid bits since the 8086 and then reflect on why they added that)

And you can always write to your lecturer about what he wants you to learn from it.

[JamesM]

Hi,

Firstly, I don't agree with tantrikwizard in the sligtest. Yes, lecturers are normally employed on their research ability as opposed to their teaching ability but this does not in any way make them "illiterate" or "substandard".

To the OP: Have a think about different types of hardware and the OSes that run on them. Contrast between:

• A desktop computer (an x86), which runs a web browser and office applications.
• A handheld device (ARM, possibly Atom) - These chips normally do not have a memory management unit, meaning paging can't be used. That changes the OS somewhat.
• A loaded server (possibly x86, possibly SPARC, possibly POWER); under high load all the time and running several virtual machines via the hardware virtualization that each of these chips offer. How does virtualization change the OS?
• A tiny ARM machine in a car, attached to the brake system. This is (a) safety critical, (b) real-time and (c) low power consumption. How do these (mainly the first two) constraints change OS design?

Hope this helps.

James

[tantrikwizard]

Do you have any evidence to the contrary?

This was, of course, a joke, though not unfounded. The age old adage of "those that can do, those that cant teach" is not without merit. However, there are always exceptions to the rule (like Tanenbaum) but in general, people who seek 'education' as a profession score lowest on high school exit exams, lowest on university acceptance exams and lowest on university exit exams. Just like people who seek "Social Work" as a profession are likely to have dysfunctional relationships, professional 'educators' are likely to be illiterate on the topic they teach.

[JamesM]

Do you have any evidence to the contrary?

This was, of course, a joke, though not unfounded. The age old adage of "those that can do, those that cant teach" is not without merit. However, there are always exceptions to the rule (like Tanenbaum) but in general, people who seek 'education' as a profession score lowest on high school exit exams, lowest on university acceptance exams and lowest on university exit exams. Just like people who seek "Social Work" as a profession are likely to have dysfunctional relationships, professional 'educators' are likely to be illiterate on the topic they teach.

Professors are professional researchers, not teachers. This is often why their teaching ability is so poor, as a rule.

[stephenj]

yomexzo, I'm not going to do your assignment for you, but I'll give you a hint. Start researching "Hardware Abstraction Layer(s)" as a base. You'll need to do more research, but it should be a good starting point.

As for the off-topic subject of CS profs' level of knowledge. One of my (many) pet-peeves in this field is how a lack of education is romantized.

One of my favourite quotes is from Jef Raskin:
"His [Steve Jobs] attractive creation-myth—swallowed whole by susceptible reporters—wherein Apple’s computers were invented exclusively by college drop-outs and intuitive engineers flying by the seats of their pants became legend. To hear him tell it, the Macintosh had practically been born, homespun, in Abe Lincoln’s log cabin. That it had been spawned by an ex-professor and computer-center director with an advanced degree in computer science would have blown the myth away. A good story will often beat out the dull facts into print."

Yes, you can be a really great programmer without an education. This point shouldn't surprise anyone, the story of great mathematicians and physicists teaching themselves with only Euclid's books has been repeated many times over the centuries*. But you shouldn't assume that someone's job title or education precludes them from knowledge. Yes, I've had a CS prof or two that wasn't a great programmer, but they were more than capable of kicking @$$ at their research topic.

As for CS profs being bad teachers, I agree with JamesM, and want to make a further point. I'd bet a lot of PhDs just grok their topic with ease, which can lead to trouble explaining it to people who don't have the same inclinations. Math is a subject that most people, to some degree, can see this attribute (it is evident in high school).

* It should be noted that a degree in CS really just prepares one to learn the indepth stuff on their own. Take what I said to yomexzo as an example. I've had... four courses... where this topic came up. It was talked about in passing every once and a while, but always in context with something else. And the reason should be evident to most of us here. HALs are an architecture specific, detailed subject.

[Schol-R-LEA]

Perhaps it would help, Tantrikwizard, for you to know the original form of the proverb (often misattributed to Shaw or Menken, both of whom made witty versions of it themselves):

''Those who can no longer do, teach those who cannot."

It's a military principle, one which means that a soldier who is too injured to fight can still train new recruits. It also applies to traditional societies where the elders, too old to work the fields, could still explain the tribal lore to the children. In other words, it means that you should listen to those who were competent enough to survive the ordeal, which is virtually the opposite of that which most people think it has.

[M-Saunders]

If one of you guys gets your OS into the wild and OSNews lists it, you'll get lots of patronizing praise that is actually insult. Look at any of the nonLinux OS's articles. Everyone is fiercely partisan. I guess I've done it too. The Linux numbskills freeloader cultists are particularly hard to bear.

Not true; I was lucky to get my mini OS as a story on OSNews, and the discussion was very pleasant: http://www.osnews.com/comments/18626

(There was only one person modded down, and that was you for being typically abrasive. Cut the ego, and stop freaking people out by claiming that God personally chose the graphics mode for your OS as per the latter stages of this thread: http://www.devmaster.net/forums/archive ... 13165.html )

M

[LoseThos]

Yeah, but He said the resolution wasn't supposed to be permanant.... after I made a promise... and you have to keep your word. Sucks.

My OS redraws the entire screen each refresh (60 Hz) and has layers that get merged with a mask. I did a test with 256 color instead of 16. My CPU load went from 5% idle to 57%. With the 16 color thing, it's bit planes what work well with bit mask layers... the 256 color, single byte values screw-it-all up.

I had some goals when I set-out. I want beginners to very easily put graphics on the screen, so I made a persistent layer and I wanted a text layer that was obnoxiously effecient... That was updated 60Hz and it didn't cost much more to put graphics on top, updated each refresh. Then, I made a persistent layer for each core. Masks are used to merge them. I do all of that with a fixed cost of 5% that goes-up, obviously if graphics are done each refresh. I'm screwed if it's not bit planes that can be merged with 64-pixel bulk operations.

Mike, You didn't notice them insulting you, I'm afraid. I've seen them insult you.

[tantrikwizard]

Perhaps it would help, Tantrikwizard, for you to know the original form of the proverb (often misattributed to Shaw or Menken, both of whom made witty versions of it themselves)...

Interesting, thanks for the history