It's interesting to read the responses to the "source code is useless" comment... mostly concerning OSS/bug-fixing and so on.
Actually, I didn't hear the talk but my interpretation of this comment was a more commercial one. Source code is obviously very important, but often (especially in the last few years after open/proprietary software has become technology-flavour-of-the-month with investors) given too much weight, especially in disputes.
Let me give an example project that I was involved in.
I was team leader of a group of programmers delivering a transactional internet application on a very complex software base including WinNT, Solaris, Java, Corba, J++, XML (let's not talk about why there were so many redundant technologies involved, it will just make me angry...) when a "dispute" arose concerning payment of large amounts of money from the client to ourselves. As often happened in those halcyon days, the client was backed by investors with big pockets and little brains. The dispute got to the point where we refused to develop any more code without payment of current arrears and some payment in advance, and they refused to pay without the source code that we had produced so far - *although* the contract stated clearly that source code would be handed over only on completion of the entire contract.
It was pretty obvious what these guys were planning - break the contract, take the code, and finish it on the cheap with the remaining $ that they hadn't blown already. The concern from our management was that they would be able to get their hands on our IP which might be worth more than what we had managed to collect from them so far.
I knew these guys pretty well, having worked with them for months together in the same office. I knew roughly what their time and money budget to completion was, and that it would cost way more to finish. More importantly, I *knew* how ugly some of that code was, because we had just finished a functionality push and needed to bullet-proof some of the underlying architecture or the thing would fall over. And... I strongly suspected that our client already had a few defectors in his pocket, and quite possibly already *had* the source code but was looking to legimate it.
So... I suggested to management that they don't worry too much about it. Even with that source code, there was no way they could find a new team to understand the project and the code base and get it to completion in time.
In the end... they cut and run, they never paid, they somehow got a copy of the source, and... they ran out of money before the new guys could finish it.
So, the source code which was the focus of the legal threats flying here and there turned out to be useless, in this case.
You seem to indicate that this is not so -- please explain.
Sorry maybe my post wasn't clear. I have nothing against service contract models.
My post was really just to say, let's not confuse "business (for a profit) as usual" with "a noble and illustrious goal".
Also, they made a buzz about supporting the open source community. As I said, they could give the hardware hackers what they wanted too, by providing the alternative option of a no-service hardware-only box. Obviously, not for the same price.
I'm no History and Philosophy of Science expert, but my understanding was that one of the main reasons for "the rise of the conference" in fields like computer science was that publishing in journals is such a rigged game. Nowadays a lot of new results are first reported in conference papers, and when the theory is rounded out, a journal paper is prepared.
No reason why the money (which are, as mentioned in other posts, truely disgusting) given to these publishing companies cannot be used by universities and research institutes to build a distributed knowledge record for research.
The Academic Publishing Industry is not long for this [brave new] world.
A lot of posters let rip on every topic like this saying: "When will everyone realise you can't sell something that's free". Selling technical support and technical consultancy is not "selling something that's free". What do you think doctors, lawyers and accountants sell?
Whether or not DotComCompanyX has a good business model or not is another matter. But never doubt that people will pay for support and advice when it matters to them.
Of course, the other viewpoint is that it's attempt to drain Grandma's bank account so dry she'll need to hock her teeth to stay online.
It's not so noble to start up a business locking people in to a service contract. Maybe it's not ig-noble either, but it's not any less business than IBM, Apple or Microsoft. (Or AOL).
If they really wanted to "support the Open Source community" they could always offer a no-service hardware-only version, at a higher price-point. I suspect they've had a think about that, but they couldn't figure it to pay.
You've brought up a topic I'm very involved in, since I'm Australian (by the way, that means I speak English) living in Germany. Some of what follows is definitely going to be flamed, so put on protective gear before you go in.
Out of interest I've looked at esperanto but not tried to learn it - most people I bounced the idea off seem very unenthusiastic about it because they think it is either (1) too simple to express complex concepts and subtle nuances, or (2) not backed by an interesting "culture", and hence not as rewarding to learn. These arguments don't convince me much, especially after reading a few Esperanto sites, but I do think that (3) you gotta think carefully about learning a World Language that almost noone speaks (2 million or so). Put in all the hard work for how much benefit?
A historical point which may be of interest (and is almost certainly flamebait;-)) is that Esperanto was suggested to the League of Nations or somesuch early in 20C as a candidate inter-language. The French didn't support it, probably because the language of diplomacy at the time was French. Things have changed since then. ASCII and standard computer technologies have made it harder for languages with accents, non-english characters, or (much worse!) pictograms like Chinese, Japanese, Korean. Unicode is much more than a nice-to-have for these people.
My experience is that people tend to learn a second language which is spoken in the land(s) they aspire to become or go to. A lot of Germans learn English. A lot of Spanish learn English. Dutch and Scandinavian people seem to be born with at least 4 languages. A lot of Eastern Europeans seem to be learning German. (Not sure what the French do, but they sure have a lot of different cheeses.)
Now, if DARPA decided to fund the development of a decentralised language which could survive and flourish under heavy cultural attack, and then this were to slowly snowball over the course of twenty years, spreading through universities and research institutes, to become The Interlingua... that would be cool.
The suggested solution of a quote on patents issued in a time period is interesting but IMHO suffers from two major weaknesses:
1) The solution requires that the patent office select the "best" patents. The current US patent problem is mainly due to the fact the the patent office has no clue.
2) Say I submit an application for a patent this year and 99 have been granted already. If you're the patent office do you grant me the patent, thereby barring all further patent applications this year, or do you reject? A sensible decision would require you to know all of the patnet applications you haven't received yet. You're not going to be able to tell.
Imagine the horror of Patent Rush on 1st January each year!
Luddites were not scared of technology, although the word is often misused in this way. They were simply opposed to the humanistic changes that technology was bringing about.
So they were opposed to the changes that were coming right? Since they hadn't come yet, these were imagined changes yeah? So when they imagined these changes, they were moved to disruptive and sometimes violent actions - not really a calm reaction.
It does sound a lot like they were scared, really.
To whit, if mobile phones had been available at the time, they would doubtless have used them as a tool for mobilising against the factory bosses.
I guess they would have to leave the mobile phone factory for last then;-)
Stochastic processes as better? Maybe I'm out of the loop, but once I read this article I said "the factor that limits the usefulness of this is that it's stochastic! I don't want probabilities that my "computer" MIGHT compute, I want to compute!"
Damn right stochastic processes aren't better than hard answers. But what about if:
a) The hard answers is really, really hard to get; and
b) The limit or probability of error is so incredibly small that it may as well be zero (i.e. it's not a hard answer but it's really, really FIRM:-) )
- then maybe you might like a stochastic process. I don't have a good example handy fulfilling these conditions but someone reading this is probably doing a PhD on just such a problem, right?
Maybe Love figured he wasn't popular with the mainstream of the Linux group anyway, so he may as well try to play for the sensible-IT-manager-who-wants-to-get-a-little-bit- pregnant.
Sorry Love [they made me say that] but I don't think your strategy works. When the average IT guy has no idea which products are good, he asks someone who does. Someone who does, is a "Linux enthusiast". You have now alienated (even further) just the folks that you need.
Due diligence refers to the company's responsibility to it's shareholders. By law, the company MUST PROTECT the interests of it's shareholders, and letting others do for free what the company has a right to collect money for is NOT taking care of your shareholders. The officers in a company have a legal responsibility to protect the shareholder's interests.
Huh?
Does this mean that Unisys has a legal responsibility to raise the price of all products and services they provide by $1 today, because they MUST PROTECT the interests of their shareholders?
Does that mean they have to raise the price tomorrow by $1 again because they MUST PROTECT the interests of their shareholders?
Hey - why stop there - let's demand that Unisys double the price of everything they sell - because they MUST PROTECT the interests of shareholders! That's shareholders value they're wasting - every day!
But what if copyright laws were repealed? What then? What would now protect software? What prevents people from statically linking their closed source programs to readline? What prevents people from distributing emacs itself closed source? What prevents people from distributing binary-only encrypted derivations of gcc that require monetary payments to obtain the key?
Nothing I guess. But my counterquestion is: who cares? If there is no copyright behind a piece of software, it will likely be leaked or reverse engineered before very long. If encryption mechanisms like compilation were effective for copyright-like purposes, I doubt copyright would have been introduced.
More philosophically, I don't believe anyone should be compelled to tell a secret (e.g. divulge their own source)... but I don't believe anyone should be able to compel another to silence (e.g. licence software such that the buyer cannot examine, copy, redistribute what (s)he has "bought").
I really don't know why this got moderated up. For the umpteenth time.
Free software is about freedom and not about cost. Your scenario will never happen. Corporations don't value freedom they only value profits.
I didn't read the original poster as meaning the software was cost free.
Also, just because corporations don't value freedom doesn't mean we won't end up with a largely free-software driven industry. What matters most is what the consumers value, and how well they can translate those values into contracts and agreements.
For instance, if most consumers of IT products were tomorrow to decide that the security risks of closed source software were too high to be borne any longer, then companies that could deliver open source software, support and consulting would make a bundle - whether their highest values were freedom, profits or the care and feeding of unwashed ponytails.
Sorry, "most" everything in Java is an Object. Java really tries to push (even enforce) OOP practices on anything that it is justified (I don't even think Sun could justify an int being an object).
Also, comparing this to a Comp Sci degree is a bit silly. This is essentially what all of the "computer colleges" out there are offering - intensive skill-based training. It may be free, and taught by some of the best people in the world, but that is still what it is.
If you look at the curriculum and then compare to a computer college I reckon you'd say it's more theoretical. It's basically a Graduate Diploma (to allow non-CS grads to move into IT). "Skills-based" learning often means MSCE-like "Windows 2000 Administration" courses...
About the burn-out... it'll be real tough to last the distance. They mention one catch-up day per week. I hope Saturday's catch-up day and you don't have to go if you were there all week.
No doubt a major reason for all this is to pull in bright young things for ArsDigita's business side.
Since when is a "master of science in computer science" degree regarded as a trade certificate? I agree with you if you are referring to MCSE etc., though.
I read (and wrote) MSCS for MCSE, sorry. But then... what is the original post talking about? I don't know too many MSc.'s who spent a lot of time worrying about what version of windows was popular while they were studying.
The thing I like about this is that it only takes one year. It seems with a computer science degree, half of anything you learned more than a year ago is obsolete.
I mean, can you imagine what the people that have spent 6 years getting a MSCS have gone through?
MSCS and other "trade certificates" are not so useful, I agree. But virtually everything I learned about Comp. Sci. at university remains useful. MS will have a hard time obsoleting recursion...
It's difficult to argue that securing patents-that-should-never-have-been increases shareholder value in any case. A lot of Amazons' customers have strong feelings about what they did, maybe strong enough that they'd like to buy elsewhere now. Where's the shareholder value increase in that?
Chasing patent rights is a business strategy - with costs (both direct and indirect) and risks. I find it very unconvincing to say the business case is so riskless and obviously beneficial that Amazon would be expected and required to procure and sue on such a small piece of technology.
Management excuse #247: The shareholders made me do it.
I couldn't find a reference for this, but not long ago I heard the Australian Army was priming soldiers before battle exercises by pumping them full of extra blood. (Their own blood taken out awhile back). Made 'em faster and so on. Haven't some athletes been disqualified for this too?
As soon as we start messing with our SELF instead of adding on, we have crossed the line into oblivion. There will be no turning back from that one.
You should bury this post in the backyard. In forty years your grandkids (or "the young generation" if you don't have 'em yourself) will probably have the same reaction to this, as most young people do today to the idea that "credit cards are too dangerous to use... the only real money is cash in the hand".
(Note: I'm not saying they're wrong about the money...)
1) We define the problem and the successful outputs. In other words, we humans may be terminally screwed up in how we perceive and analyze our environment or computational problems, but we will *not* accept a machine as "thinking" until it is approximately as screwed up, and in the same peculiar ways as we are.
Someone famous said intelligence was (something like) "patterned behaviour which we can't explain" didn't they? At least this account allows a little leeway for a creature to be intelligent without following our (possibly screwed up) patterns. Of course, it still has to be sufficiently familiar that we recognise it as a pattern...
It must be really hard to go under the Slashdot lights with your product, but this response did set my bogometer off big time.
It's hard to swallow the idea in the opening comment that Slashdoot was a high hoop for Asynchrony. The tone of the letter is advertising AND THERE IS EVEN USE OF CAPS. And the token "we love open source" sounded... token.
As I said it must be difficult to juggle the personal tone and represent the company as well. Maybe some training would be helpful.
Slashdot Mirror
It's interesting to read the responses to the "source code is useless" comment... mostly concerning OSS/bug-fixing and so on.
Actually, I didn't hear the talk but my interpretation of this comment was a more commercial one. Source code is obviously very important, but often (especially in the last few years after open/proprietary software has become technology-flavour-of-the-month with investors) given too much weight, especially in disputes.
Let me give an example project that I was involved in.
I was team leader of a group of programmers delivering a transactional internet application on a very complex software base including WinNT, Solaris, Java, Corba, J++, XML (let's not talk about why there were so many redundant technologies involved, it will just make me angry...) when a "dispute" arose concerning payment of large amounts of money from the client to ourselves. As often happened in those halcyon days, the client was backed by investors with big pockets and little brains. The dispute got to the point where we refused to develop any more code without payment of current arrears and some payment in advance, and they refused to pay without the source code that we had produced so far - *although* the contract stated clearly that source code would be handed over only on completion of the entire contract.
It was pretty obvious what these guys were planning - break the contract, take the code, and finish it on the cheap with the remaining $ that they hadn't blown already. The concern from our management was that they would be able to get their hands on our IP which might be worth more than what we had managed to collect from them so far.
I knew these guys pretty well, having worked with them for months together in the same office. I knew roughly what their time and money budget to completion was, and that it would cost way more to finish. More importantly, I *knew* how ugly some of that code was, because we had just finished a functionality push and needed to bullet-proof some of the underlying architecture or the thing would fall over. And... I strongly suspected that our client already had a few defectors in his pocket, and quite possibly already *had* the source code but was looking to legimate it.
So... I suggested to management that they don't worry too much about it. Even with that source code, there was no way they could find a new team to understand the project and the code base and get it to completion in time.
In the end... they cut and run, they never paid, they somehow got a copy of the source, and... they ran out of money before the new guys could finish it.
So, the source code which was the focus of the legal threats flying here and there turned out to be useless, in this case.
Anyone had a similar experience?
Sorry maybe my post wasn't clear. I have nothing against service contract models.
My post was really just to say, let's not confuse "business (for a profit) as usual" with "a noble and illustrious goal".
Also, they made a buzz about supporting the open source community. As I said, they could give the hardware hackers what they wanted too, by providing the alternative option of a no-service hardware-only box. Obviously, not for the same price.
No reason why the money (which are, as mentioned in other posts, truely disgusting) given to these publishing companies cannot be used by universities and research institutes to build a distributed knowledge record for research.
The Academic Publishing Industry is not long for this [brave new] world.
Whether or not DotComCompanyX has a good business model or not is another matter. But never doubt that people will pay for support and advice when it matters to them.
At least RMS signs his name to what he stands for...
It's not so noble to start up a business locking people in to a service contract. Maybe it's not ig-noble either, but it's not any less business than IBM, Apple or Microsoft. (Or AOL).
If they really wanted to "support the Open Source community" they could always offer a no-service hardware-only version, at a higher price-point. I suspect they've had a think about that, but they couldn't figure it to pay.
Out of interest I've looked at esperanto but not tried to learn it - most people I bounced the idea off seem very unenthusiastic about it because they think it is either (1) too simple to express complex concepts and subtle nuances, or (2) not backed by an interesting "culture", and hence not as rewarding to learn. These arguments don't convince me much, especially after reading a few Esperanto sites, but I do think that (3) you gotta think carefully about learning a World Language that almost noone speaks (2 million or so). Put in all the hard work for how much benefit?
A historical point which may be of interest (and is almost certainly flamebait ;-)) is that Esperanto was suggested to the League of Nations or somesuch early in 20C as a candidate inter-language. The French didn't support it, probably because the language of diplomacy at the time was French. Things have changed since then. ASCII and standard computer technologies have made it harder for languages with accents, non-english characters, or (much worse!) pictograms like Chinese, Japanese, Korean. Unicode is much more than a nice-to-have for these people.
My experience is that people tend to learn a second language which is spoken in the land(s) they aspire to become or go to. A lot of Germans learn English. A lot of Spanish learn English. Dutch and Scandinavian people seem to be born with at least 4 languages. A lot of Eastern Europeans seem to be learning German. (Not sure what the French do, but they sure have a lot of different cheeses.)
Now, if DARPA decided to fund the development of a decentralised language which could survive and flourish under heavy cultural attack, and then this were to slowly snowball over the course of twenty years, spreading through universities and research institutes, to become The Interlingua... that would be cool.
1) The solution requires that the patent office select the "best" patents. The current US patent problem is mainly due to the fact the the patent office has no clue.
2) Say I submit an application for a patent this year and 99 have been granted already. If you're the patent office do you grant me the patent, thereby barring all further patent applications this year, or do you reject? A sensible decision would require you to know all of the patnet applications you haven't received yet. You're not going to be able to tell.
Imagine the horror of Patent Rush on 1st January each year!
So they were opposed to the changes that were coming right? Since they hadn't come yet, these were imagined changes yeah? So when they imagined these changes, they were moved to disruptive and sometimes violent actions - not really a calm reaction.
It does sound a lot like they were scared, really.
To whit, if mobile phones had been available at the time, they would doubtless have used them as a tool for mobilising against the factory bosses.
I guess they would have to leave the mobile phone factory for last then ;-)
Damn right stochastic processes aren't better than hard answers. But what about if:
a) The hard answers is really, really hard to get; and
b) The limit or probability of error is so incredibly small that it may as well be zero (i.e. it's not a hard answer but it's really, really FIRM :-) )
- then maybe you might like a stochastic process. I don't have a good example handy fulfilling these conditions but someone reading this is probably doing a PhD on just such a problem, right?
Perhaps I should have explicitly mentioned that I was being sarcastic ... but thanks, you filled in the answer nicely :)
Sorry Love [they made me say that] but I don't think your strategy works. When the average IT guy has no idea which products are good, he asks someone who does. Someone who does, is a "Linux enthusiast". You have now alienated (even further) just the folks that you need.
Hmmm... Let's say "irradicate all the GIFs from the web" == "99.99% of images on the web do not use GIF format". Wanna bet?
Huh?
Does this mean that Unisys has a legal responsibility to raise the price of all products and services they provide by $1 today, because they MUST PROTECT the interests of their shareholders?
Does that mean they have to raise the price tomorrow by $1 again because they MUST PROTECT the interests of their shareholders?
Hey - why stop there - let's demand that Unisys double the price of everything they sell - because they MUST PROTECT the interests of shareholders! That's shareholders value they're wasting - every day!
I'll see them in court.
Nothing I guess. But my counterquestion is: who cares? If there is no copyright behind a piece of software, it will likely be leaked or reverse engineered before very long. If encryption mechanisms like compilation were effective for copyright-like purposes, I doubt copyright would have been introduced.
More philosophically, I don't believe anyone should be compelled to tell a secret (e.g. divulge their own source)... but I don't believe anyone should be able to compel another to silence (e.g. licence software such that the buyer cannot examine, copy, redistribute what (s)he has "bought").
Free software is about freedom and not about cost. Your scenario will never happen. Corporations don't value freedom they only value profits.
I didn't read the original poster as meaning the software was cost free.
Also, just because corporations don't value freedom doesn't mean we won't end up with a largely free-software driven industry. What matters most is what the consumers value, and how well they can translate those values into contracts and agreements.
For instance, if most consumers of IT products were tomorrow to decide that the security risks of closed source software were too high to be borne any longer, then companies that could deliver open source software, support and consulting would make a bundle - whether their highest values were freedom, profits or the care and feeding of unwashed ponytails.
Java IMHO is a hardcore OO language.
Some have dared go further
Also, comparing this to a Comp Sci degree is a bit silly. This is essentially what all of the "computer colleges" out there are offering - intensive skill-based training. It may be free, and taught by some of the best people in the world, but that is still what it is.
If you look at the curriculum and then compare to a computer college I reckon you'd say it's more theoretical. It's basically a Graduate Diploma (to allow non-CS grads to move into IT). "Skills-based" learning often means MSCE-like "Windows 2000 Administration" courses...
About the burn-out... it'll be real tough to last the distance. They mention one catch-up day per week. I hope Saturday's catch-up day and you don't have to go if you were there all week.
No doubt a major reason for all this is to pull in bright young things for ArsDigita's business side.
I read (and wrote) MSCS for MCSE, sorry. But then... what is the original post talking about? I don't know too many MSc.'s who spent a lot of time worrying about what version of windows was popular while they were studying.
I mean, can you imagine what the people that have spent 6 years getting a MSCS have gone through?
MSCS and other "trade certificates" are not so useful, I agree. But virtually everything I learned about Comp. Sci. at university remains useful. MS will have a hard time obsoleting recursion...
It's difficult to argue that securing patents-that-should-never-have-been increases shareholder value in any case. A lot of Amazons' customers have strong feelings about what they did, maybe strong enough that they'd like to buy elsewhere now. Where's the shareholder value increase in that?
Chasing patent rights is a business strategy - with costs (both direct and indirect) and risks. I find it very unconvincing to say the business case is so riskless and obviously beneficial that Amazon would be expected and required to procure and sue on such a small piece of technology.
Management excuse #247: The shareholders made me do it.
I couldn't find a reference for this, but not long ago I heard the Australian Army was priming soldiers before battle exercises by pumping them full of extra blood. (Their own blood taken out awhile back). Made 'em faster and so on. Haven't some athletes been disqualified for this too?
You should bury this post in the backyard. In forty years your grandkids (or "the young generation" if you don't have 'em yourself) will probably have the same reaction to this, as most young people do today to the idea that "credit cards are too dangerous to use... the only real money is cash in the hand".
(Note: I'm not saying they're wrong about the money...)
Someone famous said intelligence was (something like) "patterned behaviour which we can't explain" didn't they? At least this account allows a little leeway for a creature to be intelligent without following our (possibly screwed up) patterns. Of course, it still has to be sufficiently familiar that we recognise it as a pattern...
It's hard to swallow the idea in the opening comment that Slashdoot was a high hoop for Asynchrony. The tone of the letter is advertising AND THERE IS EVEN USE OF CAPS. And the token "we love open source" sounded... token.
As I said it must be difficult to juggle the personal tone and represent the company as well. Maybe some training would be helpful.