Slashdot Mirror

I can't see which prior posts you're pulling that from, but it is out by at least a couple of orders of magnitude.

Here is a very old analysis of the process creation overhead (I can't find anything more modern after a bit of googling). By old, I'll point out that they call windows NT, and do their experiments on a P200-MMX. Even then, process creation was about 100th of a second. Remember that processors are roughly 20-50x faster depending on how much architectural improvements effect a particular benchmark. The work in spawning a process is roughly constant, although windows may be doing it more efficiently now. I'd guess an upper-bound of 0.2ms on a modern machine.

Well, imo? Not really - how does "pervasive message passing" function as a COMPLETE OPPOSITE, in terms of design

It doesn't. I will give this one more try...

"Pervasive message passing" is a feature of language, or the framework. It's like object-orientedness, or garbage collection. It's a property of the tool.

"Coarse multithreading", as you've explained it, is a methodology. So is "fine-grained multithreading". These are like, oh, model-view-controller. You can implement model-view-controller with an object-oriented language, or, if you really want, you can implement it in a purely-functional language (like Haskell), or in a language without even a concept of a subroutine (BASIC with GOTOs), etc... you can do it in a language with garbage collection or without...

What's more, "coarse" and "fine-grained" are relative and a matter of opinion. How do you define a "task"? And keep in mind that while having thousands of threads sounds like "fine-grained", they are not sharing data, except by passing messages back and forth...

This is all & again: IF NO CHANGES ARE MADE? This NEVER happens, period.

Exactly.

Threads additionally (again), are much "lighter" to instance, than an entire process (as forking does, vs. spawning extra threads).

Sigh...

Apparently, you still don't get it, even though you just tried to explain copy-on-write to me. fork() on Linux is implemented as copy-on-write, and has very little overhead. Linux sees threads and processes as essentially the same thing.

The overhead of using processes instead of threads is pretty much irrelevant, especially if you were going to use message-passing in the threads, even moreso if you're intentionally doing "coarse multithreading". Of course, you could also use shared memory between processes.

And yes, all of this is Linux-specific. Forking a process on Windows is much slower.

A few - but, the point is NOT so much about them

No, the point is that, originally, I said that I would not see any real improvement if I had a quad-core system instead of dual-core.

it's more about what happens when you have a "CPU Hog" of an app...

You seem to be operating in a fantasy world where:

1. I have more than one app that wants to saturate a core
2. I do not have the ability to set the priority of these apps lower
3. I absolutely need my one app to run at 100% of that one core, rather than 90% of it. (Especially ludicrous as you keep bringing up Amarok -- if Amarok is using 100% CPU because of a bug, forcing it to use 90% isn't going to cause problems.)
4. I was saying that dual-core is actually useless to me -- it's not. It just means I don't have to do item #2, above.

I never said or implied that multicore would never be usable, or that it was useless to everyone. I was merely speculating that most people are probably in the same situation I am.

Furthermore, most of the points you bring up, I knew, but didn't think they were relevant to the discussion. Had you been paying attention, you might have noticed that I was specifically making a statement about how completely irrelevant the deep technical details are to the point of whether dual-core is useful to the general population -- most people I know would never notice a machine they were using was dual-core until you pointed it out to them.

Nearly the exact same thing can be said about NCLB as a response to "failing schools." Maybe I'm biased because I teach high school. Aside from a vague notion of school failure, nobody has specific things they think schools (read: teachers) are doing wrong. Still, NCLB provides a suite of tests and measurements to detect failures. Since the tests are concrete we can point to them as specific criteria, but some of the requirements are as absurd as 3 oz of hand soap. Let me find an example:

3.1.1 The student will design and/or conduct an investigation that uses statistical methods to analyze data and communicate results. Assessment limits: * The student will design investigations stating how data will be collected and justify the method. * Types of investigations may include: simple random sampling, representative sampling, and probability simulations. * Probability simulations may include the use of spinners, number cubes, or random number generators. * In simple random sampling each member of the population is equally likely to be chosen and the members of the sample are chosen independently of each other. Sample size will be given for these investigations.

This is from Maryland's NCLB compliance test for mathematics. It looks nice but there are a few problems in the implementation. The one that sticks in my craw the most is the use of stem-and-leaf plots as a method of visualization. There are other examples like line plots. I have yet to see one of these in actual use and they're not especially interesting as a math topics. A bigger problem (that somehow doesn't stick in my craw) is the one of simple random sampling. It's called "simple" so people think it's simple but there's some subtlety to the concept. I've been at workshops where we review and revise potential test questions. The concept of simple random sampling is subtle enough that it's very easy to come up with multiple choice questions about it that have no correct answer. At the workshop that day, we brought this up and the state representatives didn't really understand and resorted to "well, write that on there and we'll review it later."

The idea of tests to measure progress is a very old one and not terrible. It's a problem of bureaucracy that leads us to tests with low validity and pointless questions. There is a completely separate group of issues surrounding special education and these tests. They don't have an analogy with DHS that I can see.

One might ask what the interest is in creating (another?) huge bureaucracy with an impossibly broad mandate in education. It's generally accepted in some circles (my bias here) that NCLB is designed to take money out of public schools by making the system more intrusive on regular classroom education and thus disruptive and distasteful. Either people object to the level of testing in favor of education and take their kids out, or the schools start "failing" and parents want to take their kids out. Fewer students = less funding and rightly so. Opting out of the system for private interests is something our government does (see: Kyoto Protocol) so why not do the same thing with your children?

The last thing I'll say about the Maryland Mathematics High School Assessment is that it's caused problems for the math sequence. That's my bias I guess is towards the math sequence leading to Calculus. Our NCLB test is tied to the Algebra I course, which used to actually be Algebra I. Now it's about 60% math and the rest is "data analysis." The course is usually taken in eighth or ninth grade. One of my colleagues who teaches Algebra I was telling me that she feels very bad about not preparing the kids for Algebra II, and she knows they're going to run into trouble there. I had some of those kids th

I had a similar problem to the OP except mine was that I lacked the constitution to back up on a regular basis and when I did back up, it was rarely to the same place twice. This resulted in many partial backups and differing versions of the same file which got confusing and messy. My solution was to buy my buddy's old 1ghz AMD Thunderbird, a nice fat hard drive and install Linux making sure Samba was included. Three times a week (read: crontab), the Linux box in my closet runs a script (I don't guarantee linked is bug free) that uses smbclient to log into my WinXP machine. Now, my script TARs all the files, but it could be easily rewritten to simply copy.

...found this on Digg yesterday and tweaked the colors in photoshop to make a 10x7 desktop. Assuming anyone cares, I can host larger versions as well. Just be sure to specify the exact dimentions you want.

The CVS disposable digital camera - possibly the cheapest/meanest of such devices - has a plastic shutter that covers the light sensor at all times except during the split-second exposure.

Northern Michigan University has been doing something like this since 2000: http://www.nmu.edu/academics/tlc.htm

In fact, they even reference adapting the idea from similar programs in Wake Forest University and University of Minnesota-Crookston.

University of Michigan Art School
http://art-design.umich.edu/it/laptop.html

Northern Michigan University
http://www.nmu.edu/academics/tlc.htm

I was directly involved in the distribution process at one school. I didn't think it would work, but it did.

To address some of the comments from other posts:

-Most freshmen are buying laptops anyway. These programs provide a structured method for all incoming students to buy a machine that meets their needs. They are informed well in advance so they don't go off and buy a different machine.

-People complain, but it all works out in the end. Some don't like the vendor/price/model/etc. but it's just one more thing about college to complain about.

-IBM and Apple are the best companies to get machines from. They are reliable and have good support systems in place

-Computers are a part of life for any college student. So long as financial aid considers it part of the base expense, it isn't a hardship for anyone.

-Theft is a problem whether all students have the same laptop or not. You have to take the same steps (documenting MAC Addresses, having system passwords, using a physical lock) as you would with any notebook

To answer some of the original questions:

-For software, contact each vendor and they will usually work out a good educational license deal for each machine (No server/key setup). You can cite the above schools' programs.

-Make sure the machines have adequate RAM for the job. Get RAM from a 3rd party if the laptop company won't give you a good deal. Get small, fast HDs rather than huge slow ones. Uses external 3.5" drives for backup

-Make antivirus/antispyware part of the base load/image. Have all updates automatic

-Port block and track IP/MAC addresses for bandwidth usages. Have graduated punishments for connection misuse

-Use a student staff for technical/question support. They are cheaper and more knowledgable that trying to train middleaged staff. Use a mixture of students and fulltime staff for physical support issues. You can't trust kids to do all the warranty work by themselves.

Good luck!

I've thought of this quite a bit myself, as I come from a school which furnishes laptops to all students and faculty and has complete wireless coverage. Also, I hope to be a professor someday myself (hopefully, grad school here I come!)

Personally, I wouldn't mind if students waste their time in my class, but as you say, if they start wasting others' time (more often than not against their will, I've been sucked in by someone playing Quake 3 across the room) I'd be royally pissed off. There are plenty of reasons to allow unrestricted access, but no reason to disrupt class.

I think the method I'll have to use at some point is just to be vigilant for abuse like that, and ask troublemakers to leave class for the period. I guess I will have to make sure I can wield power like that before doing so however, I don't know what the policy usually is for asking students to leave classes at the university. (Though I do know that at my school, my fiancee is asked to leave quite often for coughing because the professor is a germaphobe. Heh.)

the fastest box that i "own"(really i'm renting it from nmu) is a 1.7ghz pentium m and the slowest was a 100mhz pentium with 128MB or ram.... so yes i know hat it's like and i used that box untill about 2 years ago as my desktop. and now my primary box is a 876mhz transmeta. now the pentium is my server box with a 10mbit card. it's my print/web/nfs/smb/ftp/gentoo distfiles mirror/etc my web site was being served by it for a while. now that i'm back on 26.4 dial-up i don't have a web page... so yeah i know about old hardware even though i'm only 18

My school Northern Michigan University already does. I was provided with a not quite stock IBM R40... At acost of something around $300USD a semester. I get the laptop for 4 semesters and 1 summer, and free repair(as long as it looks like normal use) and "tech support"(usually just being told that you have to back everything up and reimage). At the end of my lease I may purchase the laptop. Everything is works under Suse 9 pro. Not sure if the modem is supported elsewhere. I have no use for the modem.

My school Northern Michigan University already does. I was provided with a not quite stock IBM R40... At acost of something around $300USD a semester. I get the laptop for 4 semesters and 1 summer, and free repair(as long as it looks like normal use) and "tech support"(usually just being told that you have to back everything up and reimage). At the end of my lease I may purchase the laptop. Everything is works under Suse 9 pro. Not sure if the modem is supported elsewhere. I have no use for the modem.

I attend Northern Michigan University. We have a campus resnet that has live IP's and DNS names My_Tower_for_example and if you go look port 80 is open, 8080, and 22, but port 21 and the samba port and the nfs port are closed. that and they have blocked ping packets on campus, so i would ping google for you but i can't but it probably would be in the neiborhood of 500ms+ which is jsut a little bit higher than one would expect for a collage campus. some of the biuldings on campus are behind NAT as well. so i guess over all it's not the worse system but there are some odd things that happen.

We also have very few if any computer labs on campus because the school "provides" us with laptops to use, running a cripled(as in rdesktop does not work) version of Winblows XP.

p.s. please be nice to that server it's a 100mhz pentium..... :-)

Broke through! download 1.5 here

Well according to this page, in Debian 2.2:

This kernel was 2.2 and was in 2001. I also found that as of 2.4.0, there were about 3 million lines of code. A full 1/3 of the code is ripped from SCO? hmm.. I bet.

With around 95% of the market it makes sense that Microsoft hasn't really been adding new features to Internet Explorer.

But I'd guess that with the growing dissatisfaction with pop-up advertising and the growing popularity of Mozilla's (or Firebird pop-up blocking they might have to rethink this soon.

doesn't look like it's necessarily needed - but a mirror is up (just dir listing) at

longyear.nmu.edu/~cs/wtc

Substantiated claims? Ever heard of Jochen Liedtke and the L4 microkernel? He and many other researchers wrote the following papers if you'd care to read them:

The performance of microkernel based systems

Achieved IPC performance

Microkernels must and can be small

On Microkernel construction

Improved Address-Space Switching on Pentium Processors by Transparently Multiplexing User Address Spaces

If you'd like a breakdown, here it is: x86 sucks for context switch times for two main reasons, a)user/supervisor transition times are an order of magnitude slower than other architectures and b) a poorly designed TLB cache results in a flush whenever a context switch occurs(only if the address space changes of course). Any advantage the x86 architecture gains by having few registers is lost(and then some) by these other factors. This can be readily seen in operating systems and kernels that rely heavily on context switching such as microkernels.

For quick and dirty evidence: here. That's a link showing context switching times for Linux running on an 850 MHz PIII. The times typically hover around 12 microseconds. The papers I linked to above show achieved IPC times for L4 which are steady around 3 microseconds on much lower-end hardware. That's IPC, not just context switching, ie. context switch and copying data to another address space. The L4 teams have tweaked their implementation as fast as it can go on x86, and have achieved performance an order of magnitude higher than Linux(at least in this area). This demonstrates some pretty solid expertise. Given this, they say admit that x86 is very poor in this respect and a great hindrance in designing a good operating system. I believe some of the papers briefly discuss other architectures, but most of the focus is on the x86 platform because it's such a performance problem.

P.S. the fact that the x86 is register poor is not a good thing. Having two few registers to manipulate data is often a hindrance. Saving 3 times as many registers doesn't take that long anyhow.

Linux just gets a new FS here, a little cleaning up in the code there

Those little bits and pieces add up to a better OS. Thats getting faster

check out the quote on http://euclid.nmu.edu/~benchmark/index.php?page=nu ll_call:

"As mentioned in our methodology section, this is due to a bug in the kernel code that lead to a feature freeze in subsequent kernels."

if a bug in the kernel code can cause a feature freeze, someone better debug the developers! :)

jon

$ strings total_growth.gif | head -2
GIF89a
Software: Microsoft Office

http://euclid.nmu.edu/~benchmark/null_call.gif

This shows why computer guys are not scientists. My first year phys chem prof would tear his own arm off and beat you to death with it if you gave him a graph that looked that ugly.

The Excel defaults may be ugly, but you can change them.

According to this graph page fault latencies suck in kernel 2.2. Is this true? I think I'm running a 2.2.17 AC kernel though and if I'm just doing development and not causing swapping then it doesn't matter though right?