[an error occurred while processing this directive] COMP9242 S2/2000 On-line Survey Results
[CSE]  Advanced Operating Systems 
COMP9242 2013/S2 
CRICOS Provider
Number: 00098G

PRINTER Printer-Friendly Version

On-line Survey Results S2/2000

Answered by all 15 students enrolled

Quick Evaluation

1. Give a low ranking if you have a bad opinion of something (e.g. too slow, confusing, too difficult). Give a high ranking if you have a good opinion (e.g. interesting, useful, effective).

1. ExcellOkPoorN/A
Lecturer 1: Gernot Heiser   66.6%     26.6%     0%     6.6%     0%     0%  
Lecture material   26.6%     73.3%     0%     0%     0%     0%  
Project   86.6%     13.3%     0%     0%     0%     0%  
Reference material   20%     33.3%     33.3%     6.6%     0%     6.6%  
Computing resources   13.3%     13.3%     40%     13.3%     20%     0%  
Course web pages   26.6%     53.3%     20%     0%     0%     0%  
Exam   60%     26.6%     6.6%     6.6%     0%     0%  
COMP9242 overall   66.6%     33.3%     0%     0%     0%     0%  


2. Which factors most influenced your decision to enrol in this course?
2. Interest in operating systems as an area of study11
Chance to build a system7
Chance to get fingers _really_ dirty7
Would like to do systems research5
Looking for a challenge5
Looking for easy 12 UoC0

Other factors not mentioned above?
Comments      6 comments

3. Would you recommend this course to another student such as yourself?
    Yes     100%             No     0%

4. The course is heavy on design and implementation issues. It also tries to reamain close to present research issues. What do you thing about this?

4. Too
Theory/general principles   0%     6.6%     80%     13.3%     0%  
OS design and implementation issues   0%     13.3%     73.3%     13.3%     0%  
Current research issues   0%     6.6%     40%     53.3%     0%  

5. What were the best things about this course?
Comments      14 comments

6. What were the worst things about this course?
Comments      12 comments

7. Keeping in mind that it bears double credit points and is therefore expected to impose a double workload, how does the workload in this course compare to workloads in other ...

7. Much
Elective level COMP courses   0%     0%     6.6%     73.3%     20%  
COMP courses in general   0%     6.6%     0%     53.3%     40%  
Courses in general   6.6%     0%     0%     33.3%     60%  

8. How does the overall quality/value of this course compare to other ...

8. Among
the best
the worst
Elective level COMP courses   80%     20%     0%     0%     0%  
COMP courses in general   80%     13.3%     6.6%     0%     0%  
Courses in general   73.3%     13.3%     6.6%     6.6%     0%  

9. What background knowledge do you think you were missing that would have helped you in this course? Is credit in COMP3231/9201 and a co-requisite of Computer Architecture a suitable preparation?
Comments      11 comments


10. Please rate the relevance/appropriateness of the lecture topics.

10. Very
L4 general and L4 API   66.6%     33.3%     0%     0%     0%  
Microkernels in general   73.3%     20%     6.6%     0%     0%  
Microkernel/L4 implementation   53.3%     26.6%     6.6%     13.3%     0%  
Protection, part. capabilities   40%     53.3%     6.6%     0%     0%  
Caching & TLBs   80%     20%     0%     0%     0%  
Page tables   73.3%     26.6%     0%     0%     0%  
Multiprocessor issues   40%     40%     20%     0%     0%  
Real-time issues   33.3%     40%     20%     6.6%     0%  
File systems   73.3%     20%     0%     6.6%     0%  
Mungi   40%     26.6%     26.6%     6.6%     0%  
Cool stuff from SOSP   33.3%     33.3%     33.3%     0%     0%  

11. Please tell us how interesting you found the lecture topics.

11. Very
L4 general and L4 API   46.6%     20%     20%     6.6%     6.6%  
Microkernels in general   73.3%     20%     6.6%     0%     0%  
Microkernel/L4 implementation   13.3%     26.6%     33.3%     20%     6.6%  
Protection, part. capabilities   40%     33.3%     20%     6.6%     0%  
Caching & TLBs   46.6%     33.3%     13.3%     6.6%     0%  
Page tables   46.6%     40%     6.6%     6.6%     0%  
Multiprocessor issues   46.6%     26.6%     13.3%     13.3%     0%  
Real-time issues   40%     26.6%     20%     13.3%     0%  
File systems   46.6%     33.3%     6.6%     6.6%     6.6%  
Mungi   60%     26.6%     13.3%     0%     0%  
Cool stuff from SOSP   46.6%     40%     13.3%     0%     0%  

12. Which material do you think will be most useful to you in the future?
Comments      12 comments

13. Which material, not currently in this course, would you liked to have seen covered?
Comments      10 comments

14. Which of the current topics would you like to see scaled back or excluded?
Comments      10 comments


15. Please comment on the mode of lecture delivery.

15. Very
Lecture delivery using slides from laptop   60%     33.3%     6.6%     0%     0%  
Looking at kernel code using source browser   6.6%     26.6%     46.6%     13.3%     6.6%  

16. Lecture attendance remainde high throughout the session (contrary to many other courses). What factors caused you to keep attending lectures?
16. I had enough spare time3
The lectures were too good to miss9
Given the pace and lack of a textbook I couldnt' afford to miss the lectures9
It was as good a place as any to take a nap.1
I wanted to be seen to be there3

17. How could lectures be improved?
Comments      8 comments


Please give some feedback on the project.

18. What was the level of difficulty of each part of the project?

18. Too easyJust rightToo hardDon't
Milestone 0   13.3%     20%     60%     6.6%     0%     0%  
Milestone 1   0%     0%     93.3%     6.6%     0%     0%  
Milestone 2   0%     13.3%     80%     0%     6.6%     0%  
Milestone 3   0%     13.3%     73.3%     13.3%     0%     0%  
Milestone 4   0%     6.6%     66.6%     20%     0%     6.6%  
Milestone 5   0%     0%     53.3%     33.3%     0%     13.3%  
Milestone 6   0%     6.6%     53.3%     26.6%     6.6%     6.6%  
Milestone 7   0%     6.6%     53.3%     20%     13.3%     6.6%  
Bonus: Shared memory   0%     0%     40%     6.6%     0%     53.3%  
Bonus: Clock driver   0%     6.6%     26.6%     6.6%     0%     60%  
Bonus: File system caching   0%     0%     26.6%     13.3%     0%     60%  
Bonus: Dynamic file system   0%     0%     26.6%     13.3%     0%     60%  
Bonus: OS paging   0%     0%     26.6%     6.6%     6.6%     60%  
Bonus: NFS   6.6%     13.3%     20%     6.6%     0%     53.3%  
System Documentation   0%     6.6%     73.3%     0%     0%     20%  
Project overall   0%     0%     60%     26.6%     13.3%     0%  

19. How well was the project specified?

19. Very clearOkConfusingDon't
Milestone 0   20%     33.3%     46.6%     0%     0%     0%  
Milestone 1   13.3%     26.6%     53.3%     6.6%     0%     0%  
Milestone 2   13.3%     26.6%     53.3%     6.6%     0%     0%  
Milestone 3   6.6%     26.6%     53.3%     6.6%     6.6%     0%  
Milestone 4   13.3%     20%     60%     6.6%     0%     0%  
Milestone 5   20%     33.3%     40%     0%     0%     6.6%  
Milestone 6   13.3%     26.6%     40%     13.3%     0%     6.6%  
Milestone 7   6.6%     40%     40%     6.6%     0%     6.6%  
Bonus: Shared memory   6.6%     6.6%     33.3%     6.6%     0%     46.6%  
Bonus: Clock driver   6.6%     13.3%     26.6%     0%     0%     53.3%  
Bonus: File system caching   6.6%     6.6%     33.3%     0%     0%     53.3%  
Bonus: Dynamic file system   6.6%     6.6%     33.3%     0%     0%     53.3%  
Bonus: OS paging   6.6%     6.6%     33.3%     0%     0%     53.3%  
Bonus: NFS   13.3%     13.3%     33.3%     0%     0%     40%  
System documentation   6.6%     13.3%     33.3%     20%     6.6%     20%  
Project overall   6.6%     33.3%     46.6%     0%     0%     13.3%  

20. What do you think about offering alternative projects (as some students have chosen)?

20. Great idea!Don't careAbolish!
Alternative projects   60%     26.6%     13.3%     0%     0%  

Do you have any comments about alternative projects?
Comments      9 comments

Anything Else

21. Any other comments/suggestions/etc, that might help us to improve this course in the future?
Comments      8 comments

Other factors not mentioned above?
Back to Question

1. Heard it was good, also heard it was tough...
2. good lecturer :)
3. Hope can gain a thorough understanding of OS.
4. OS was really cool. Chance to use something other than boring old x86
5. None.
6. It would have been "looking for an easy" 12UoC last year, but that wasn't enough of an incentive to do it back then. "Easy", meaning I could hit 12UoC in one go without juggling subjects. I knew AOS was interesting but very difficult. And if I was to be suicidal, it was nice to at least be familiar with people who think the same but do it anyway. This year ... well, I only needed 6UoC, so uh - yeah. Probably the most stupid thing I ever did.

5. What were the best things about this course?
Back to Question

1. Chance to play with L4 Really good lecturer Interesting material
2. Building a real operating system, in a real operating system development with so little debugging tools, and eventually watching it run successfully.
3. Able to build a simple operating system almost from scratch.
4. Just the hands-on experience provided rather than just theory. The weighting given the assignment emphasised this
5. I can get my hands really dirty, and gain real understanding of OS.
6. The whole project was cool The lectures were great fun and full of cool content
7. Definitely the project.
8. OS implementation, good critical view of OS research, lecturer knows what he is talking about.
9. Project
10. The level of freedom allowed for the project. Unlike most other computing subjects, the project allowed us (the guinea pigs) to laregely make our own decisions and either reap the benefits, or the face the consequences.
11. Challenging
12. L4 & Mungi lectures ... they were really inspiring.
13. The chance to build a simple OS from scratch is a good idea, I had heaps of fun doing the project. The small lecture style was cool too, I love the fact that I can pop questions at you when things just pop into my head (You probably can guess who I am now).
14. The project

6. What were the worst things about this course?
Back to Question

1. Lab machines dodgy
2. The assignments were quite hard, and allocated to little time to complete. Balnacing the subject with my thesis was indeed challenging.
3. Have to spend all my time building the simple operating system.
4. lack of reliable machines in asyst, l4 bugs - not necessarily a bad thing in itself but just the time wasted due to them
5. The documentation on L4 is not enough, and the last milestone of the project is much harder than normal, so it's too hard to finish the last milestone on time.
6. The dodgey network setup and the broken asyst machines got to be real pain.
7. Massaging some of the machines into working and generally fighting over the machines that worked well became tedious.
8. Not enough support for people that don't have a strong OS background or a high level of intelligence to strengthen themselves.
9. Wednesday 6-9
10. It would have been beneficial if relevant lecture material been presented before certain mile stones. The in depth look at the ipc mechanisms in lecture 10 (I think) was good, however I think it would have been worth while spending some time looking at how other things work in l4.
11. The hardware and having to come in to do assignments. I really wanted the luxury of doing things at home and run X at home didn't really cut it because of the poor speed and that some machines corrupted their timers somehow during the running of SOS that can only be relieved by a cold boot.
12. It's too short! I think this course could be better if it was a year-long course, but I guess that's what an OS thesis is for.

9. What background knowledge do you think you were missing that would have helped you in this course? Is credit in COMP3231/9201 and a co-requisite of Computer Architecture a suitable preparation?
Back to Question

1. I don't think I was 'missing' any background knowledge. At this level, we should be expected to be able to find any extra knowledge we need! I also think that the credit/co-req is suitable for these reasons.
2. The mentioned prerequisites are definitely required, and also a strong working mind and interest in OS is defintely needed to fulfil the requirements of all the milestones. I think maybe even a high CR in OS should be the prerequisite, and a CR is suffice for Comp Arch.
3. Yeap, I think that is sufficient. Cos unless we have enough background on L4, we still have to go through the painful process of learning and playing around with it.
4. suitable, although having done computer arch previously would probably help (I wasn'd doing them concurrently)
5. I think some knowlege on computer architecture may help.
6. I think the computer architecture was useless in AOS. AOS covered any topics needed in much greater depth and to better understanding. AOS helped more for comp arch than the other way around. OS is a definate pre-req, however.
7. I believe a credit in COMP3231/9201 is a suitable preparation.
8. I don't think either of those courses provide a suitable preparation for the subject as nothing is taught on multithreading and concurrency in early years. C programming is an important strength also which isn't assured in every student at the moment.
9. It might have been better to make computer architecture a pre-requisite rather than a co-requisite.
10. Suitable mostly, but for people who haven't touched OS for a whole year, things do get rusty.
11. The prereqs for this course is sufficient I think. I think I need more general background about things such as how programs how compiled and how to program in C. But that's mainly me not knowing too much low-level stuff.

12. Which material do you think will be most useful to you in the future?
Back to Question

1. All of it - a general grounding of knowledge in how the darn thing that manages your apps works is a good thing. More specifically, I think that the low-level C hacking is the part I'm going to make most general use of!
2. The knowledge of the requirements of an operating system.
3. The appreciatio of a good operating system.
4. hmm... hard to tell at the moment :)
5. the project really helps.
6. All if it. This subject opens your eyes to how wide a field OSs is and that it is not just a windows/linux fight.
7. Caching, TLB and pagetables.
8. That depends on whether I am capable of performing research in the area. If I don't, probably only the low-level coding experience.
9. File systems, caching and TLB'S
10. Microkernels - possibly. I'm not terribly sure. I'm not sure if a career in OS is for me either.
11. I think Multiprocessor and Real-time issue. The other stuff are cool if you planning to be a OS hacker, but I'm not :), So I guess I shouldn't be in this course?
12. Hardware (cache,tlb,PT's etc) and Real-time Some Mungi topics applicable to MMU-less hardware

13. Which material, not currently in this course, would you liked to have seen covered?
Back to Question

1. None. One point I would like to make, though, is that students could be actively encouraged to take part in an ongoing research project, rather than reconstitute an existing SOS... I would suggest that ALL projects should be 'special' in this sense.
2. Nothing really.
3. benchmarking
4. monolic kernel architecture like linux would help to compare from miroc-kernel architecture.
5. Just a bit more on older systems that influenced modern ideas, I think... only if it's an interesting topic... AFAIK it could be very boring :)
6. Not sure.
7. It would have been nice to see more about current research around the world. It would have definitely sparked some interested research nerves inside me.
8. More current research
9. Never did mention much about exokernels. Only read a little about them and I was interested in whether you think that sort of philosophy is any good.
10. We see alot of how OS and hardware interact, but we never seem to touch on how user/apps programmer and OS interact? Like obviously a Good OS has to satisfy the application programmer (besides being fast).

14. Which of the current topics would you like to see scaled back or excluded?
Back to Question

1. None. It was all cool.
2. Source code sessions.
3. Not relevant.
4. they're all pretty cool :) digging through the l4 code wasn't that exciting but I think it was still worth it
5. I have no idea on this one.
6. None. I think an extra hour or two of lectures a week would also be cool :)
7. None.
8. None, all topics are relevant although some aren't particularly interesting to me.
9. None actually. Liked them all.
10. None really, Like I said, there's not enough lectures.

17. How could lectures be improved?
Back to Question

1. Heh - less source code browsing :) Seriously, I recognise the importance of this but it's BORING! This is really nit-picking, though. The enjoyment + learning that I derived from this course far exceeds just about every other course I've done at this university. It's really nice to see that somebody (read: Gernot) is willing to put in the time and effort to avoid AOS following the general trend I've seen in subjects (easier assignments, less catering for advanced students, lower workloads, less material covered -> less overall usefulness). Thanks for a brilliant subject!
2. It really good.
3. can't think of anything really
4. L4 implementation should also include more on structure not just coding.
5. I don't know... no matter how tired I was after work I always enjoyed the AOS lecture (this is very contrasting to the 6-9 lecture on thursday we have for algorithms where I spent most of the time in the bar instead)
6. Definetely required dancing clowns (sorry this survey is too long to maintain seriousness :) ... I can't think of anything though.
7. Also the time meant that it was _very_ difficult to miss by sleeping in :)
8. Not much really, Besides the source code part of the lecture, it is a bit dry. It's probably made all the more dry since the source code is pretty hard to read on screen. (I printed Inside L4 out, which helps).

Do you have any comments about alternative projects?
Back to Question

1. I think EVERYBODY should do an alternative project. That way, we get to have fun comparing projects, and the school gets more useful research done. Alternatively, make the project one HUGE group project encompassing the whole class, and we can try & build a halfway decent system!
2. Nope
3. nope
4. That really depends on what the alternatives are.
5. I was interested in the idea but I didn't really know where to start thinking about another project or to find a list.
6. Must be difficult to maintain fairness in marking.
7. Great idea .... but too scary and the marking scheme well, I had no idea really.
8. Not really, I don't know much about them, but it's a good idea to let the smart people do something cool
9. It would have been interesting to do one....

21. Any other comments/suggestions/etc, that might help us to improve this course in the future?
Back to Question

1. Just re-capping. This subject was brilliant. That isn't to say that there mightn't be improvements, but this subject is so far above 'average' that it isn't funny. PLEASE don't decrease the difficulty of this subject to 'cater for the masses', because I think it would lose a hell of a lot.
2. Keep up the good work....
3. In question 7 when I rated it as 'Much Heavier' that is not too say it is bad! Yes there is more work, but everybit of it is worthwhile. In Q8 I rate it as 'amongst the best' because there is no flat out 'the best' option.
4. None.
5. The lecturer can at times be intimidating. Eliminating that and providing a little more support for the weaker students would generate a more positive feel for those with a lesser knowledge of operating systems.
6. Many times I really wanted to concentrate on the implementation of a particular component and do it _well_. The milestones don't allow that if time is an issue. Doing an unrelated thesis and AOS at the same time may not have been such a good idea. What does it take to join the Mungi team? Of the whole class how many would you welcome to join this team? etc. etc. Some ideas about career prospects in this field would be really nice.
7. I think the weekly milestone, if you're late you're penalised is silly. I like the milestone idea, but it is very inflexible. Like sometimes you try to work ahead, but other times the other subjects swaps you and you tend to fall behind. Sometime you just fall behind a week and already you lost a mark. Maybe have the same amount of milestone, but extend the deadline?
8. L4/Alpha and gdb for debugging :)

Last modified: 17 Dec 2002.