Jane Margolis and Allan Fischer had a close look at the situation of women in computing at CMU. Based on their study, CMU developed a program to attract and retain a larger percentage of women, which actually worked.
Some of their methods are not easily transfered to other schools. For example, they increased the percentage of women admitted to computing by changing the selection criteria - not by watering them down, but by unbiasing them (e.g., women tend to have less previous experience with computers). CMU could do so and still maintain their standards due to the large number of excellent applicants, which might not work as well for schools that are less popular.
Nevertheless, the CMU study produced interesting results, some of which suggest ways to remove bias against women where it exists.
All our student computer labs, except two, are running a customised version of Debian GNU/Linux. One exception is a Windows lab used for a course that depends on some Windows software. The second non-Linux lab is a Mac-based HCI lab. Overall, there are 20 Linux labs; see this overview for details.
In addition, almost all of our servers run on GNU/Linux.
CSE.UNSW has a long Unix tradition, part of which has been publicised by the Salon article about John Lions.
People have been saying for years that Slashdotters don't read the article, so I thought that I'd test the theory.
You are jumping to conclusions to quickly. I actually did read the article before looking at the comments (or even posting one). Nevertheless, when reading the article, I focused on what you call the "insignificant" bits of the article. After all, the summary on the/. frontpage is what made me read the article, in the first place.
So, if you can conclude anything from your experiment, it is that people concentrate on their first impression. Not really a new insight.
of Gnome fame is very good. He has done a lot of stuff - check his homepage - and he can give interesting presentations. (I attended one of his talks once.)
So.. acording to Sony, the "internal workings" means the... programing? The contestants are not alowed to change the iner mechanical workings of the robots or the external apearence.
In fact, the teams weren't even allowed to rip out the whole OS and replace it by something better - Sony doesn't provide the necessary details to directly address the hardware and forbids reverse engineering. Interestingly, it seems that RoboCup teams that rip out as much of Sony's software as possible get better results. According to a member of the team from Berlin this was a consistent pattern this year and last year.
Now, the article seems to indicate that this will be a seperate company with backing from various larger interests: Anyone know if this is true or will this end up as some kind of 'holding company'? Also, if it is a seperate company, any word on an IPO?
The article says that this will be a nonprofit organisation. So, an IPO doesn't make any sense here.
They say you can use any language you like, functional or not, but look carefully at this quotation from the contest page:
Although the precise task chosen will not be revealed until the contest begins, algorithmic cleverness matters. Performance may matter. Programming languages that help programmers to build complex systems quickly may allow contestants to attempt particularly sophisticated implementations in the 72 hours allotted for programming.
In other words, the criteria for winning are completely subjective. The judges can make whatever arbitrary decision they choose and back it up with subjective talk about "sophistication" and "cleverness", without having to take performance into account at all.
You obvioulsy didn't read the announcement very well (not even the quote that you are citing). It says that the task will be choosen such that algorithmic cleverness etc matters. That's all. The judgement won't be on subjective, but on objective criteria - like does your program work. Mind you, last year, most C, Perl etc. submission loosed just because they had bugs in them - they just produced the wrong result or crashed. That's a pretty objective test, I would say (if you don't believe me, look at the Web pages from last year).
Now, you can say that choosing a programming task that requires some non-trivial algorithms is a bias in favour of FP languages. And, yes, you are right. But that's the point. When programs and algorithms become more complex, more sophisticated programming languages buy you something. If you don't believe it, then demonstrate that you can produce a working implementation to a algorithmically challenging problem in 72 hours and get rid of all your bugs.
Nobody says that you can catch all errors with a good type system, but my experience with writing (lots of) Haskell is that a good type system can catch a lot of errors - in particular also a large number of logical errors. Personally, I would say that Haskell catches at least half (probably much more) of the errors during compilation that would lead to a core dump in a C program.
This of course implies that you make good use of the type system and choose good type abstractions and annotate all top-level functions with type signatures.
Chilli
PS: Look at my Web page, I have written a lot of Haskell code and it works very well for me.
Check out the Glasgow Haskell Compiler. It can generate C code and that's how it is bootstrapped on new platforms. A Java backend is just in the works. The RTS is statically linked and it just got a nice interface to call C code (and call Haskell from C). It generates good code and has a sophisticated profiler to help you speed up things if need be.
And of course it is open source (with CVS and all).
You can only wonder why B. Meyer wrote this article. Considering its length, it probably took a considerable amount of time to write and he seems to be a busy man. I doubt that the free Eifel compiler is enough of a "threat" to ISE's Eifel system to be the (sole) reason. Maybe that commercial software developer who allegedly had an unhappy encounter with RMS was a friend of him, or maybe he himself met RMS and got ripped to pieces? Or is he worried that if open source produces high quality software, nobody will listen to him anymore?
I recently had the "pleasure" of attending two talks of B. Meyer. I essentially learned two things:
He is a vocal speaker who tries to manipulate his audience.
His tactics fail on a significant part of the audience (I gathered this from talking to other attendees afterwards).
His main technique is very simple. He starts out with statements of which he thinks that the majority of the audience will agree with. He uses the "validity" of these statements to prove an unrelated point, which on first sight might seem related.
It is interesting that he attempts the same technique a couple of times in this article. The most obvious case is probably how he uses ESR's standpoint on gun ownership to discredit ESR's statements about open source software. It seems, as if his discrediting of C/C++ programmers mentioned in another comment falls in the same category (everybody believes that software quality is important, but does that mean everybody trained on C/C++ is a danger to a software project?)
The problem here, as in the talks I listened to, is that his strategy is just too obvious.
That's a mistake often made by people who still didn't grok the spirit of free software and identify it with commercial software that you can download for free (like various browsers). I actually suspect that is one of the reasons why these people cannot get to terms with the fact that they can download the software for free (see B. Meyer's longish argument explaining why the production of "free" software still costed money).
Free software breaks the consumer/producer duality, instead of with money, you pay by getting involved. If you don't have the skill or time to get involved, you can pay a third party like RedHat to do the "getting involved" bit for you and in return you expect them to make sure that you get the service that you pay for.
Ok, let me make an attempt at a very rough translation - not so much for the info, but because part of it is really funny (and we know already that it is no real product yet):
Title: ZDNet's bold prediction, does Sony's palm design look like this?
What might be the appearance of the Sony palm, which is scheduled for release in the middle of 2000? What are its features? And the price? ZDNet is proud to say that the answer of the fearless PDA-manics - after studying the design of the VAIO and Sony's AV-technology in detail - is the ``VAIO GEAR InfoPilot'' (Sony, please don't be angry about our choice of name, ok?) The VAIO's never changing body colour and colour monitor go without mention, and we want a CCD and a memory-stick slot. The logo is of course embossed in the back.
This time, our imagination (or wishes) had no limit. As we didn't get our hands at any internal material from Sony, we ask our readers to refrain from sending us questions (and it is not possible to reprint the drawing).
Positioning: A palm device is an excellent portable terminal, but being an organiser, if you don't have a PC or a Mac, you can't install any software. Consequently, the forthcoming device should be a VAIO peripheral; or in other words, be part of the ``VAIO GEAR'' series.
Design and Features: Sony is considering a VAIO-coloured body, but as the Sony MusicClip is silver, this is also a very likely design.
At the side, let's have a dial that you can operate one-handed. It might end up looking like the CASSIOPEIA, but we should copy good ideas.
The slot for the memory stick is at the top. This takes into account the ability to connect future GPS units. Maps can be downloaded from the Web via a build in mobile. As the coming Palm will make use NTT's DOCOMO communication infrastructure (there is a definite announcement scheduled for the beginning of 2000), this might actually become reality. Furthermore, to store downloaded MP3s and ATRAC3s, we probably won't be able to do without an MG-memory stick.
Although we don't really see the need for a CCD, if there is a built in mobile it would make for a simple video phone and at times it might be convenient to memorise train timetables.
[..]
Ok, that should be enough to give you an idea of the tone of the article. At least ZDNet Japan seems to have humour (although, I am sure it got all lost in my aweful translation...)
I believe the sound on the newer Sony models is provided by the NeoMagic video chip (no, I'm not kidding), which has a few problems under Linux at the moment.
You are right that Sony is using the NeoMagic chip. Unfortunately, although the chip is well supported for its video operations, its audio component is rather badly supported (and NeoMagic refuses to publish specifications). In fact, Sony uses different versions of the NeoMagic chip, which have different audio CODECs. There are at least three versions: AC97, CS4232, and SB16; but I suspect that there is at least one more (the one contained in the PCG-N505) and I tried, but didn't get it to work at all. When you buy a Vaio, Sony doesn't even tell you which version of the chip they used in that particular model - so, you buy it without knowing whether it will work in the end, or not. This is extremely annoying. AFIK, Sony is the only of the Japanese laptop manufactures who completely refuses to cooperate with the Linux community (others have a least a page listing known in/compatibilities and Toshiba even provides information on how to install Linux on their machines - at least their Japanese page is quite good, I didn't check out their English resources).
Brett, your whole argument relies on a a completely unsubstantiated premise, namely that you can't make money with GPLed code. Unfortunately - for your argument - Red Hat is making a fortune and I bet the programmers working for VA Linux writing GPLed code all day are getting a decent salary. Last I checked, Linus Torvalds also didn't seem to starve.
This is not about eliminating the software industry (as you claim), it is about transforming the software industry to produce better software more cost effectively. I am sure, the society as a whole will appreciate this.
I had two machines from Sony's VAIO PCG-505 series, which I liked a lot, but Sony's recent politics and increased incompatibility with Linux[1] annoyed me sufficiently to switch to Sharp.
I am now using a Sharp Mebius PC-PJ2-X4 (that's at least the name it sold under in Japan) and I love the machine. The LCD screen is even better than Sony's. Video, Sound, APM, etc all works with Linux - only the winmodem is annoying. With a 333MHz PII, 128MB, 12GB disk, and an XVGA screen, the machine has enough resources for serious hacking. Furthermore, Sharp provides a 1 year international warranty on machines bought in Japan, but used outside of Japan (doesn't cover all countries). Currently, the only Japanese laptop companies with that service are Sharp and Toshiba.
Chilli
[1] The major problems where that the new machines (eg, PCG-N505 series) use winmodems, problematic sound chips (and Sony doesn't specify which they are using), and USB for external mouse and floppy (the latter is a problem for the B5 machines that have no builtin floppy).
As I understand the question, the author of the letter is not saying that he is planing on doing something illegal. He (or she?) is just afraid that others might try to harm him by trying to sue him, and he wants to be prepared for that. Every company accumulating a patent portfolio for defense purposes is doing exactly the same and nobody is calling that illegal.
Face it, whether you are right or wrong doesn't really matter too much in such a case if your opponent has the resources to scare you away by intimidation.
I don't think that anybody is ignoring the contribution of all the fine engineers and operators who ensured the safe transition of our computer systems into the new year. What was clearly overhyped IMO was all that fear that we will be thrown back into stone age and should better spend new year's eve in a bunker. That media hype is what is critised.
Happy New Year and thanks to all who have helped squashing y2k bugs.
The important thing is that expressing the program in two distinct languages results in two distinct sorts of thinking as the two expressions are written. Most bugs tend to occur only in one of the thinking modes, and are thus exposed when the diverse expressions of the program are compared.
It is not only this. The formal specification of a program/algorithm also is usually much more concise and closer to the domain language used for informally talking about the problem. As a result it is (if the right framework was chosen) easier to find conceptual flaws in the formal specification.
But you are of course right in saying that there is no such thing as a 100% verification.
Let's make the others follow us through the mine field! The open source/free software community and the (computer) science community already have a lot in common and heavily overlap. We should strengthen this trend and make them blend. This way the front line of research and OSS will be the light that industry follows through the mine field.
...over the word free for a moment: Look at the order in which the operating systems are mentioned. Sure it is alphabetically, but this didn't happen by chance, and usually people don't care that a certain legacy OS starts with W.
The issue here is that the Blackdown team has been working porting new versions of the JDK source to work on Linux for YEARS in good faith with the understanding that Sun would consider the results of their efforts the "unofficial but blessed-by-Sun Linux JDK". Blackdown trusted Sun to play fair and Sun turned around and stabbed them in the back.
The point is that it is foolish to trust a company. You may be able to trust an individual (but even then things easily get hairy when money is involved), but there is no point in trusting a company. They have to make profit and they have to make their shareholders happy (and you can't blame them for that). If you are standing in the way, they'll wipe you away. Even when you found an individual in a company who you think, you can trust, the person may be leave, and again, you are at the mercy of people who don't care about you personally.
The only time you can trust a company is when you have a solid contract with them. RMS understood that a long time go.
Chilli
-- ``The issue is not whether you are paranoid, look around you Lenny; the issue is, whether you are paranoid enough.'' -- Max in Strange Days
Seems like I have missed something - thanks for the tip. (BTW, it also took me quite a while to find in fvwm2's man page how to do it properly - seems the window manager authors, or at least the documentation writers, are not so interested in these kind of features.)
I immediately agree that it is possible to imagine alternative input devices that would make life much easier. Glasses following the eye focus would be wonderful, indeed. I am not sure, however, how precise these are using current technology (and how expensive).
Hating the rodent myself (for everything except surfing the Web with Netscape, which could also be made more keyboard friendly), I have tried various window managers - none being really satisfying:
twm is hopeless, dumped it long ago.
fvwm2 is quite good to use without a mouse. The reason: you can make it switch to a specific window with on key stroke based on the WM_TITLE and WM_CLASS of that window.
WindowMaker can largely be operated with the keyboard, but to switch to a specific window, you have to get the window menu (usually F11) and then select the window based on the first letter of the window title (using arrow keys is far too slow).
I am currently trying E, and having to use ALT-TAB is annoying.
What I really miss most in all window manager (that I tried) - except fvwm2 - is selecting specific windows with one keystroke (or an ALT-whatever combination).
I am currently pondering whether it would be worth the effort adding this feature to E - but E 0.16 still has serious memory leaks, so I don't know whether I won't have to switch again anyway.
This is seriously troubling me and if there were a cool looking (theme-aware) WM that can be easily used with the keyboard only, I would immediately switch.
Slashdot Mirror
Jane Margolis and Allan Fischer had a close look at the situation of women in computing at CMU. Based on their study, CMU developed a program to attract and retain a larger percentage of women, which actually worked.
Some of their methods are not easily transfered to other schools. For example, they increased the percentage of women admitted to computing by changing the selection criteria - not by watering them down, but by unbiasing them (e.g., women tend to have less previous experience with computers). CMU could do so and still maintain their standards due to the large number of excellent applicants, which might not work as well for schools that are less popular.
Nevertheless, the CMU study produced interesting results, some of which suggest ways to remove bias against women where it exists.
Chilli
CSE.UNSW has a long Unix tradition, part of which has been publicised by the Salon article about John Lions.
Chilli
So, if you can conclude anything from your experiment, it is that people concentrate on their first impression. Not really a new insight.
Chilli
Well, vimacs really exists!
Chilli
Chilli
In fact, the teams weren't even allowed to rip out the whole OS and replace it by something better - Sony doesn't provide the necessary details to directly address the hardware and forbids reverse engineering. Interestingly, it seems that RoboCup teams that rip out as much of Sony's software as possible get better results. According to a member of the team from Berlin this was a consistent pattern this year and last year.
Chilli
The article says that this will be a nonprofit organisation. So, an IPO doesn't make any sense here.
Chilli
You obvioulsy didn't read the announcement very well (not even the quote that you are citing). It says that the task will be choosen such that algorithmic cleverness etc matters. That's all. The judgement won't be on subjective, but on objective criteria - like does your program work. Mind you, last year, most C, Perl etc. submission loosed just because they had bugs in them - they just produced the wrong result or crashed. That's a pretty objective test, I would say (if you don't believe me, look at the Web pages from last year).
Now, you can say that choosing a programming task that requires some non-trivial algorithms is a bias in favour of FP languages. And, yes, you are right. But that's the point. When programs and algorithms become more complex, more sophisticated programming languages buy you something. If you don't believe it, then demonstrate that you can produce a working implementation to a algorithmically challenging problem in 72 hours and get rid of all your bugs.
Chilli
This of course implies that you make good use of the type system and choose good type abstractions and annotate all top-level functions with type signatures.
Chilli
PS: Look at my Web page, I have written a lot of Haskell code and it works very well for me.
And of course it is open source (with CVS and all).
Chilli
I recently had the "pleasure" of attending two talks of B. Meyer. I essentially learned two things:
- He is a vocal speaker who tries to manipulate his audience.
- His tactics fail on a significant part of the audience (I gathered this from talking to other attendees afterwards).
His main technique is very simple. He starts out with statements of which he thinks that the majority of the audience will agree with. He uses the "validity" of these statements to prove an unrelated point, which on first sight might seem related.It is interesting that he attempts the same technique a couple of times in this article. The most obvious case is probably how he uses ESR's standpoint on gun ownership to discredit ESR's statements about open source software. It seems, as if his discrediting of C/C++ programmers mentioned in another comment falls in the same category (everybody believes that software quality is important, but does that mean everybody trained on C/C++ is a danger to a software project?)
The problem here, as in the talks I listened to, is that his strategy is just too obvious.
Chilli
Free software breaks the consumer/producer duality, instead of with money, you pay by getting involved. If you don't have the skill or time to get involved, you can pay a third party like RedHat to do the "getting involved" bit for you and in return you expect them to make sure that you get the service that you pay for.
Chilli
Chilli
You are right that Sony is using the NeoMagic chip. Unfortunately, although the chip is well supported for its video operations, its audio component is rather badly supported (and NeoMagic refuses to publish specifications). In fact, Sony uses different versions of the NeoMagic chip, which have different audio CODECs. There are at least three versions: AC97, CS4232, and SB16; but I suspect that there is at least one more (the one contained in the PCG-N505) and I tried, but didn't get it to work at all. When you buy a Vaio, Sony doesn't even tell you which version of the chip they used in that particular model - so, you buy it without knowing whether it will work in the end, or not. This is extremely annoying. AFIK, Sony is the only of the Japanese laptop manufactures who completely refuses to cooperate with the Linux community (others have a least a page listing known in/compatibilities and Toshiba even provides information on how to install Linux on their machines - at least their Japanese page is quite good, I didn't check out their English resources).
Chilli
This is not about eliminating the software industry (as you claim), it is about transforming the software industry to produce better software more cost effectively. I am sure, the society as a whole will appreciate this.
Chilli
I am now using a Sharp Mebius PC-PJ2-X4 (that's at least the name it sold under in Japan) and I love the machine. The LCD screen is even better than Sony's. Video, Sound, APM, etc all works with Linux - only the winmodem is annoying. With a 333MHz PII, 128MB, 12GB disk, and an XVGA screen, the machine has enough resources for serious hacking. Furthermore, Sharp provides a 1 year international warranty on machines bought in Japan, but used outside of Japan (doesn't cover all countries). Currently, the only Japanese laptop companies with that service are Sharp and Toshiba.
Chilli
[1] The major problems where that the new machines (eg, PCG-N505 series) use winmodems, problematic sound chips (and Sony doesn't specify which they are using), and USB for external mouse and floppy (the latter is a problem for the B5 machines that have no builtin floppy).
Face it, whether you are right or wrong doesn't really matter too much in such a case if your opponent has the resources to scare you away by intimidation.
Chilli
Happy New Year and thanks to all who have helped squashing y2k bugs.
Chilli
It is not only this. The formal specification of a program/algorithm also is usually much more concise and closer to the domain language used for informally talking about the problem. As a result it is (if the right framework was chosen) easier to find conceptual flaws in the formal specification.
But you are of course right in saying that there is no such thing as a 100% verification.
Chilli
Chilli
Chilli
The point is that it is foolish to trust a company. You may be able to trust an individual (but even then things easily get hairy when money is involved), but there is no point in trusting a company. They have to make profit and they have to make their shareholders happy (and you can't blame them for that). If you are standing in the way, they'll wipe you away. Even when you found an individual in a company who you think, you can trust, the person may be leave, and again, you are at the mercy of people who don't care about you personally.
The only time you can trust a company is when you have a solid contract with them. RMS understood that a long time go.
Chilli
--
``The issue is not whether you are paranoid, look around you Lenny; the issue is, whether you are paranoid enough.''
-- Max in Strange Days
Chilli
Chilli
- twm is hopeless, dumped it long ago.
- fvwm2 is quite good to use without a mouse. The reason: you can make it switch to a specific window with on key stroke based on the WM_TITLE and WM_CLASS of that window.
- WindowMaker can largely be operated with the keyboard, but to switch to a specific window, you have to get the window menu (usually F11) and then select the window based on the first letter of the window title (using arrow keys is far too slow).
- I am currently trying E, and having to use ALT-TAB is annoying.
What I really miss most in all window manager (that I tried) - except fvwm2 - is selecting specific windows with one keystroke (or an ALT-whatever combination).I am currently pondering whether it would be worth the effort adding this feature to E - but E 0.16 still has serious memory leaks, so I don't know whether I won't have to switch again anyway.
This is seriously troubling me and if there were a cool looking (theme-aware) WM that can be easily used with the keyboard only, I would immediately switch.
Chilli