I agree totally. I don't really care about the spelling, or the grammar even (although in this case it did make the article difficult to parse). But there has to be a point. I read all the way through this article twice, and I just don't get it.
I agree with what you say in general principle, but it does make me very uneasy. It's the first step down the proverbial slippery slope.
It's a very small step, IMHO, from poor design to bugs in a program. They're really just bugs of a different kind (a design bug vs. an implementation bug).
And if a software writer can be sued because of bugs in their program...?
It's one thing if they knowingly left bugs in their program that could be exploited without telling anyone, but quite another to prosecute/sue people simply for having bugs in a distributed program, whether they sell it or not. (If this isn't what you're getting at, I apologize; it just sound's like it to me).
I do not like micros~1, for many reasons. However, this is not one of them. It scares the $#!+ out of me to think that someone could bring a lawsuit against me simply because of some bugs that happened to be in my program that I didn't catch. After all, it is extremely difficult, even impossible, to be sure you have eliminated all bugs in a program. Of course, I will do my best to fix the ones I find, but some will remain, undetected by me. And some of those just may be serious; I won't deny that possibility.
No, I tend to think that limited liability for programmers is a good thing... -- - Sean
Exactly what the subject says. If MS was forced to auction off it's source code, WINE's work would be over.
Not because they would be able to rip the code. They wouldn't even need to look at the code. The real benefit from opening Windows' code would be that at last the "secret hooks" and undocumented API's would be out in the open. In other words, WINE's target would be properly defined.
Once thay have that; once they know *exactly* what they need to emulate, all the hard work is done, as far as I can figure. -- - Sean
Ummm... I have no idea where you live, dude, but around here, you do have to give your employees breaks and vacations -- it's required by law.
Besides, let's say you were allowed to not give your employees these prviledges. But in your jov advertisements you said you would give them anyway. Even in your interviews. Then, once the employees were on-board, and tied into a lengthy contract, you revealed to them that they werent... I'm sure even you would admit this is highly immoral, if not illegal. And this is exactly the sort of thing that Microsoft does.
Also, you are not allowed to discriminate against who you allow into your store -- again, it's the law.
Either you're trolling, or incredibly naive that's all I can say. -- - Sean
In my eyes, a (sic) OS monopoly is a *GOOD* thing.
Imagine where software would be today if we weren't trying to recreate the wheel on 100 different platforms.
No, I don't think breaking down monopolies just because they're monopolies is a good thing.
This is a good analogy -- a very good analogy, in fact, but like many analogies, it is misapplied.
First off, just because a wheel exists, don't assume that it is not worth reinventing. You bring your analogy from the "Real World," where we have a nice round wheel which works very well. But imagine if that wheel were in fact square? The wheel as we know it is not worth reinventing, which is the basis for that expression. However, if the wheel were substandard, ie: a square wheel, it would be very much worth reinventing. And that is how I believe the analogy would be more apliccable to our Operating System situation. We have a square wheel (Windows), which we are trying to re-invent, or rather supplement with a better, round version (ie: *nix).
Don't let the mantra of "Don't reinvent the wheel" prevent you from replacing a substandard wheel.
Second of all, A wheel may be worth improving. Not necessarily reinventing, but improving on the original design. Such as (for a real-world wheel) adding shocks, tyres, and so on. In terms of our software analogy, this can mean anything from Microsoft's attemtpts at "embrace and extend" (bad) to Linux's DE's such as KDE and GNOME (good). On their own, each individual strain of improvements may lead to OS clutter and usability setbacks (witenss the KDE/GNOME flame wars on Slashdot), but on the whole, especially once you combine them, and/or relegate each improvement to its specific purpose, they are a definite improvement.
Third, and finally, don't confuse standards with monopolies. This is where the wheel <--> Operating System analogy really shines, and also where you misapplied it. There is absolutely nothing wrong whatsoever with a standard -- standards are good. The wheel is a standard; everyone can use a wheel. But the wheel is not a monopoly. No one company or person has control over either the design or the implementation of the wheel.
Rather, the wheel's design is well-known. It is ubiquitous -- a standard. Anyone who wants to can take a wheel apart, find out how it works, and make another one. This is the reason why it is a standard -- also why no-one reinvents it. Simply because no-one has to. Unfortunately, Microsoft does not allow this with their Operating Systems. Not only is no-one allowed to take them apart and see how they work, but it is very difficult to emulate them. And this is exactly the problem that the "solution" propsed by the states is trying to solve!
By opening Windows up, and forcing Microsoft to release/auction off their source code, they would be effiectively freeing up the design for the wheel, and letting anyone build their own, so they don't have to reinvent it!
So don't confuse design and implementation. Ubiquity of design is great. A monopoly on the implementation (which Microsoft has) is not. We don't have to reinvent the wheel because the design is open to us. With the opening up of the "Windows" wheel, we won't have to reinvent that one, either, and maybe, just maybe, we can improve it, changing the wheel from a rolling log into a spoked set of discs on an axle, with tyres and shocks. Pardon the analogy, but I think you can see where I'm going.
Basically, you hit on the problem, but misapplied it, coming up with the wrong solution. Or rather, you had the right solution, but translated it the wrong way.
I agree, we shouldn't have to reinvent the wheel. That is what we are trying to put out. And A monopoly can be good. But a monopoly on design, belonging to the people, rather than a monopoly on implementation, belonging to a company.
If Dunlop had a monopoly on the wheel, and priced it/did quality control to match, I'm pretty sure there would be alternatives that people would use, leading to fragmentation. Thankfully, they don't. The design of the wheel is open, leading to standardization as people realize what features are and are not useful.
Don't let a fear of fragmentation lead you into accepting a square wheel. -- - Sean
Warning: This is probably going to be a bit long-winded. This is because I have a number of related things that I want to say, and they are all sort-of jumbled up. So.
Yes. Hackers are arrogant.
Why? They just are. It's the so-called Nature of the Beast. Asking whether or not a hacker is arrogant is akin to asking whether or not a fish can swim.
And the arrogant hacker, as with the swimming fish, has been brought about by necessity.
There are two main contributions to this arrogance. The first is the much-vaunted "gift culture" that is so predominant. In the absence of "real" money, reputation counts as currency. How many times reading a day's worth of Slashdot, do you come across the phrase "show me the code!"? This phrase is somewhat of a double entendre (excuse me for iterating through obvious concepts; I'm just trying to be thorough). The first meaning refers, obviously, to the notion of code/source being open to the community. The second, and more relevant meaning is that a person's reputation is built almost entirely on what they have done. This point is important; I will return to it later.
If a person has not written code, they have zero reputation. For this reason are people such as Linus, Alan Cox, RMS and so on so highly respected. They have contributed a lot of very useful code to the community. For the same reason are people such as ESR, Tim O'Reilly and so on -- people who have coded, but are known more for other things -- somewhat less respected. Are they respected? Yes, but less so, especially in the Linux/Slashdot/Open Source community.
This reputation is treated as currency; if there is no respect for your code, you are no-one. Alternatively, if your code is respected, people listen to you and will do things for you.
I have been derided on Slashdot for the fact that I am a "VB Code Monkey." This is not considered coding. I have no problem with that; agree with it for the most part.
Now, this contributes to arrogance. When you have coded something, you are respected (and rightly so) for it. You become arrogant. This is RMS to a tee.
I mentioned two contributions. The other is the constant tug-of-war that the hacker must play with their code. Hackers tend to be introverted -- we spend most of our time playing with code. Code that can become annoyingly frustrating at times, yet code which is our bread-and-butter, our oxygen even.
Always forefront in the hacker's mind is the realization that if their code doesn't work, there is one thing at fault: the person who wrote it. That's the great thing about computers -- if your program ain't working, you got no-one to blame but yourself. But it's this same relationship between the hacker and their code that brings out the dominant traits of the hacker. How many times have you stared at a piece of code that would not work and wished that you could rip it off the screen and give it a good smack?
Ok, I'm being a bit melodramatic -- or am I? Perhaps not in such a corporeal manner, but there have been many times when I have wanted to force the program to to what I wanted, not what I said. When I have sat, cursing, at the code. When I have all but ripped the monitor from my desk and heaved it out the window.
And yes, I am displaying arrogance toward my program.
And before you deride me for being silly and childish, actually think about the relation between you and your code. You are constantly forcing the code to your will. You are constantly making it do what you want it to. And is this not arrogance?
If not, I don't know what is.
Because the hacker is always in a dominant position over his/her code, that tends to translate to their dealings with other people. Because they have a similar relationship with other hackers (who are the people they tend to deal with), where one person always has a bigger reputation than the other, based on the code they have written, this permits them to display arrogance towards the other. Thus the behaviour pattern is further entrenched.
Ok. As I said, I'm being pedantic, and probably have said nothing that isn't common knowledge. Hackers are arrogant. Out of necessity. And it's a good thing. If we weren't, we would be far less efficient at what we did. Namely, write good code.
Unfortunately, the combination of this arrogance with the fact that non-coders are not afforded respect means that few "outsiders" are allowed entry into the ranks. Oh, sure, they are allowed to come in, but they are quickly turned off and leave.
This I'm not so sure is a good thing.
My mother uses computers; she is a chemistry college teacher, and certainly intelligent in her own right. However, she is by no means a hacker. She sent an email recently to a colleague, which got misdirected to the sysadmin at that college. Because she is not a "hacker," or even an "uberuser," she didn't know about the Reply-To line, or using a signature. So she just included within the text of the eMail where the person could reply to.
The sysadmin replied, informing my mother that the eMail was mis-addressed. When I next spoke to her, she showed me his (the sysadmin's) eMail, and I read it. He had also pointed out that it wasn't necessary to include the return address in the eMail text, and included instructions on how to set up a.sig file and how to set the Reply-To adress. I read it, and turned to my mother, who couldn't for the life of her understand what she had done to provoke this guy's ire. She was wondering why he was being so condescending and, yes, arrogant.
Stunned, I re-read the email. There was no condescension, anger or arrogance that I could see. It was just a letter, rather polite, I thought, explaining to her how to set up her preferences and stuff. Until I looked at it through her eyes -- those of a non power-user, who has no interest in computers beyond them being a tool to write her exams in Word, make her marking tables in Excel and send off email to her colleagues in Pine. All of a sudden, this email was written by some holier-than-thou guy who was demanding that she make a whole bunch of changes to some obscure file because he didn't want to see a return address embedded in the text, where he wasn't expecting it.
Now I re-iterate that I found it to be completely innocuous. It was not until I looked at it as someone who made minimal usage of computers that I saw this.
And I realized that this attitude was very pervasive in the hacker community. I wouldn't normally give it a second thought. Hell, I wouldn't even notice it for the most part. Because, as I showed above, it is not only natural, but necessary. It is a fact of life. The nature of the beast.
Imagine now that you are a regular user... a luser. Perhaps a reporter covering a Linux Expo event. Maybe even a chemistry teacher. Someone who uses computers minimally. Someone who definitely does not code. Someone who doesn't necessarily even want to know more about computers, because it is not relevant to you. Someone deserving the contempt of any hacker. Someone who... whoops! I think we see the problem.
Now we are at a dilemma. The hacker needs to be arrogant. But at the same time, the hacker's arrogance turns off other people. How do we solve this problem?
Before I continue, I recognize that there are many people who do not necessarily consider this a problem to be solved. That is your prerogative.
I argue, however, that it is. The reason why is the old world-domination thing. We can't get more users if we keep scaring them off. We can't have world domination if we don't get any new users. And I believe that world domination is necessary. At least a little bit. Does Linux need to be a Microsoft-killer? No. But it must reach "critical mass". I'll explain that in a second. I think that we are almost there. But not quite. We are approaching it, though, and it is a heady time. Linux are getting "official" recognition from Oracle, IBM, Dell, Corel, you name it. Linux is maturing. But it's also a dangerous time. If we turn people off now, we may not get this chance again. I remember OS/2 being in the same position about 5 years ago. I was strongly involved in the OS/2 community at that time. And one of the common complaints was that people involved with TeamOS/2 were needlessly arrogant toward outsiders. Exactly the same complaint as is being levelled at the Open Source community today. OS/2 almost reached critical mass, but blew it. I think Linux has a good chance, and don't want to see the same thing happen.
And I define critical mass is that point where it is self-supporting. Yes, the fact that Linux is open source is a tremendous advantage over OS/2, and all the proprietary OS'es that have either succeeded or failed up until now. But don't fool yourself into thinking that it is self-supporting. All the open source in the world won't build drivers for closed hardware interfaces. All the open source in the world won't build support for new types of hardware. Once the user base reaches a certain size, that in itself will be enough incentive for companies to either write drivers, or release the specs. That's the critical mass.
But it's not quite there yet.
It needs just a little bit of a push... if we don't stop that push before it even starts.
So what do we do? We are arrogant, because it's needed for what we do. We do what the "standard" software companies do. We get a spokesperson. Or spokespeople. It would be a bit hard to swallow, I'll wager, because such a person would not be a hacker. Such a person would not write code. Such a person would never truly "understand." Such a person would be very similar to Jon Katz, and we know what controversy he creates around here.
But he is precisely the sort of person we need. Someone who can talk to "the public" in terms that they can understand. Someone who doesn't talk code. Someone who doesn't necessarily get it all right, because there are a lot of people out there who simply don't care.
Your average user has absolutely no use for detailed knowledge of programming, or even computers in general, any more than a hacker has use for detailed knowledge of open-heart surgery. It just isn't applicable. But yet we will jump to condemn these people for not wanting to learn. Well, surprise! Some people don't want to. Does this make us better than them? Perhaps. But it's probably not a good idea to let on. Unfortunately, the hacker's natural arrogance tends to reveal these true feelings. We don't have to hide our feelings from the computer, therefore we aren't used to hiding them from people.
Which is why a hacker should not be representing hackers to the "rest of the world". Unfortunately, RMS is a hacker. He insists that other people follow his instructions, just as he insists that the computer follow his code. And he is possibly one of the worst representatives we could have.
A plea, then, for sane representation:-)
Or someone who can talk to the non-hackers in a way that won't threaten them. Yup, he (or she) will get a few things wrong. Sure, he won't totally understand. And he/she may even say a few things that a hacker would consider utterly moronic.
But better that than appealing to the hacker, and driving everyone else away.
-----
Well. I applaud you for getting this far, because I know my style can be rather pedantic. I also haven't been able to say quite what I wanted to say, but I think it's close enough. You're going to have to do a bit of interpretation there on some of the points I've made. Basically, if what I said sounded really strange, or a bit loony, it probably came out wrong:-)
Anyway, I'm just trying to give people something to think about. -- - Sean
Given that time is always represented on any graph along the x-axis, saying that something that is hard for a newbie to learn has a "steep learning curve" is factually incorrect -- it's a misnomer.
If anything, it should be considered a shallow learning curve -- for a given time period (delta-x), the gain in proficiency will be minimal (a small delta-y). Represented thusly (note: you may have to widen your browser to see the whole graph properly):
| | P | r | o | f | * i | * c | * * i | * * e | * * * n | * * * c | * * * * y | * * * * | * * * * * -------------------------------------------------- -- T i m e
In reality, however, even this is somewhat misrepresentative. That's not the way people tend to learn any given product.
A "real-world" learning curve would actually have two focii, splitting the curve into 3 sections:
A shallow one, where the user was just starting out, and didn't know where to start. This is true of any product.
A much steeper one, as they gained enough proficiency and familiarity with the product to learn on their own, very quickly (as opposed to having to rely on others).
Another shallow one, once they have learned the most useful features, and they don't need to learn much more, except for a few one-time uses. With some, less expansive products, the curve may end at this point, and go flat, indicating there is nothing more to be learned. With others (ie: Linux) with a huge feature set, this section may be delayed for quite a ways into the curve. But it will come.
To wit:
| * * * * * * * * | * * * * * | * * * P | * * r | * o | * f | * i | * c | * i | * e | * n | * c | * y | * | * * | * * * | * * * -------------------------------------------------- --------------- T i m e
Obviously, in this latter case, it is of little relevance to say whether or not the curve is "steep" unless you also mention what part of the curve you are referring to.
It is not alway correct to apply a generalization (the curve for such-and-such a product is steeper/shallower) to the curve as a whole. In the case of Linux vs Windows '9[58], for example, I belive that the initial part is much, much, much shallower (ie: harder to learn) for Linux, whereas once you get far enough along the curve, the second part becomes steeper (ie: easier to learn). Also, any comparaison is totally non-applicable to the third part, which is rather short and comes early in the case of Windows, but is much longer and comes much later for Linux.
Just my $0.02 worth (or by this point $2.00...?):-) -- - Sean
This is very strange... I'm running MSIE (at work; no choice) on a Windows box with MS' "embraced and extended" character set-thingie, and I'm still getting the ?'s instead of ''s...
I'm beginning to think it's a different problem this time, although I can't imagine what... -- - Sean
As with everything MS, every new version tends to introduce a fair amount of bloat.
If you use the original v1.0, you can get it down pretty small. Actually, versions 1 through 4 basically "compile" it to a form of Bytecode, then package it up with the interpreter embedded in the.exe (or as VBRun100.dll, VBRun200.dll, VBRun300.dll, VB40016.dll, VB40032.dll -- should be self explanatory).
The size of these things has steadily increased from about 100k (I think) for VBRun100.dll up to almost a meg for VB40032.dll.
But in any case, the program itself was not much more than what you have above. In fact, it was sometimes smaller than the actual VB code.
Versions 5 and 6, however, offer the ability to do a "true" compilation (into native code). Although the program will run faster (95 times out of 100), it all of a sudden requires huge globs of "things" to run.
Exactly what these "things" are tends to vary... for the most part, they are dll's and ActiveX components. Most of which are not required by your actual program (assuming you are using the Hello World example above).
However, not requiring them and not having them are 2 different things.
Microsoft, aiming VB at the lowest common denominator, tends to lump everything it thinks you may (ie: commonly) need into the package whether you ask for it or not.
The result is that the size of your executable + associated DLL's all of a sudden mushrooms to the vicinity of 2 1/2 megs.
Of which you only need (for your simple "Hello World" program) MSVBVM50.dll (the Virtual Machine stuff -- it actually is Java-like in that respect -- "native code" is a "bit" of a misnomer), which clocks in at 1.28 megs. But don't even think of trying to get rid of the rest. It has so many dependency tables that reference the other crap that deleting them will break your program irrecoverably (unless you put 'em back).
So is there a way to get rid of them? Yes, futz around with the options a whole bunch before you compile. You will also have to add and remove a few lines from your project (*.vbp) files by hand (the IDE doesn't offer the option), as well as a couple of other config files. But don't do it unless you know what you're doing. You won't succeed.
And me? I know because I have to work with this shit (and that's what it is) for a living.
I've thought about it myself, not so much in the context of Slashdot, but as a general principle.
Working out algorithms to maximize efficiency and utility, and so on...
The main drawback, I think, would be the hardware requirements, in terms of processing, storage space for the DB, server load, and so on. My experience is that those sorts of interrelated sets and the associated processing are a HUGE drain on the processor, require lots of memory, and also require enormous databases (at least, to do properly).
Even with Slashdot's spiffy new server, I don't think it meets the requirements. Yet.
But maybe someday... we can always hope... -- - Sean
Slashdot Mirror
He does have something important to say hidden deep within.
...which is? Sorry... I read it twice and didn't get it. I would be more than grateful if someone could elucidate it for me...
Thanks!
--
- Sean
Oh god no!
It was all I could do to keep from spewing...!
--
- Sean
Bottom line: The point is the point.
I agree totally. I don't really care about the spelling, or the grammar even (although in this case it did make the article difficult to parse). But there has to be a point. I read all the way through this article twice, and I just don't get it.
Rowan, what are you trying to say?
What is your point???
--
- Sean
Is the slang use of the word "bender" only relevant to English people or something?
I dunno... I've never heard the expression before...
--
- Sean
No, yet another proof of workplaces dictating what their employees use.
I use office and outlook (ok, I'll use lowercase). Not because they're good. Because I'm told to.
--
- Sean
Maybe someone will figure that if they make an example of one cracker, others will be "scared straight".
Sounds suspiciously like what they tried with Kevin Mitnick, to me...
--
- Sean
I agree with what you say in general principle, but it does make me very uneasy. It's the first step down the proverbial slippery slope.
It's a very small step, IMHO, from poor design to bugs in a program. They're really just bugs of a different kind (a design bug vs. an implementation bug).
And if a software writer can be sued because of bugs in their program...?
It's one thing if they knowingly left bugs in their program that could be exploited without telling anyone, but quite another to prosecute/sue people simply for having bugs in a distributed program, whether they sell it or not. (If this isn't what you're getting at, I apologize; it just sound's like it to me).
I do not like micros~1, for many reasons. However, this is not one of them. It scares the $#!+ out of me to think that someone could bring a lawsuit against me simply because of some bugs that happened to be in my program that I didn't catch. After all, it is extremely difficult, even impossible, to be sure you have eliminated all bugs in a program. Of course, I will do my best to fix the ones I find, but some will remain, undetected by me. And some of those just may be serious; I won't deny that possibility.
No, I tend to think that limited liability for programmers is a good thing...
--
- Sean
Exactly what the subject says. If MS was forced to auction off it's source code, WINE's work would be over.
Not because they would be able to rip the code. They wouldn't even need to look at the code. The real benefit from opening Windows' code would be that at last the "secret hooks" and undocumented API's would be out in the open. In other words, WINE's target would be properly defined.
Once thay have that; once they know *exactly* what they need to emulate, all the hard work is done, as far as I can figure.
--
- Sean
Ummm... I have no idea where you live, dude, but around here, you do have to give your employees breaks and vacations -- it's required by law.
Besides, let's say you were allowed to not give your employees these prviledges. But in your jov advertisements you said you would give them anyway. Even in your interviews. Then, once the employees were on-board, and tied into a lengthy contract, you revealed to them that they werent... I'm sure even you would admit this is highly immoral, if not illegal. And this is exactly the sort of thing that Microsoft does.
Also, you are not allowed to discriminate against who you allow into your store -- again, it's the law.
Either you're trolling, or incredibly naive that's all I can say.
--
- Sean
In my eyes, a (sic) OS monopoly is a *GOOD* thing.
Imagine where software would be today if we weren't trying to recreate the wheel on 100 different platforms.
No, I don't think breaking down monopolies just because they're monopolies is a good thing.
This is a good analogy -- a very good analogy, in fact, but like many analogies, it is misapplied.
First off, just because a wheel exists, don't assume that it is not worth reinventing. You bring your analogy from the "Real World," where we have a nice round wheel which works very well. But imagine if that wheel were in fact square? The wheel as we know it is not worth reinventing, which is the basis for that expression. However, if the wheel were substandard, ie: a square wheel, it would be very much worth reinventing. And that is how I believe the analogy would be more apliccable to our Operating System situation. We have a square wheel (Windows), which we are trying to re-invent, or rather supplement with a better, round version (ie: *nix).
Don't let the mantra of "Don't reinvent the wheel" prevent you from replacing a substandard wheel.
Second of all, A wheel may be worth improving. Not necessarily reinventing, but improving on the original design. Such as (for a real-world wheel) adding shocks, tyres, and so on. In terms of our software analogy, this can mean anything from Microsoft's attemtpts at "embrace and extend" (bad) to Linux's DE's such as KDE and GNOME (good). On their own, each individual strain of improvements may lead to OS clutter and usability setbacks (witenss the KDE/GNOME flame wars on Slashdot), but on the whole, especially once you combine them, and/or relegate each improvement to its specific purpose, they are a definite improvement.
Third , and finally, don't confuse standards with monopolies. This is where the wheel <--> Operating System analogy really shines, and also where you misapplied it. There is absolutely nothing wrong whatsoever with a standard -- standards are good. The wheel is a standard; everyone can use a wheel. But the wheel is not a monopoly. No one company or person has control over either the design or the implementation of the wheel.
Rather, the wheel's design is well-known. It is ubiquitous -- a standard. Anyone who wants to can take a wheel apart, find out how it works, and make another one. This is the reason why it is a standard -- also why no-one reinvents it. Simply because no-one has to. Unfortunately, Microsoft does not allow this with their Operating Systems. Not only is no-one allowed to take them apart and see how they work, but it is very difficult to emulate them. And this is exactly the problem that the "solution" propsed by the states is trying to solve!
By opening Windows up, and forcing Microsoft to release/auction off their source code, they would be effiectively freeing up the design for the wheel, and letting anyone build their own, so they don't have to reinvent it!
So don't confuse design and implementation. Ubiquity of design is great. A monopoly on the implementation (which Microsoft has) is not. We don't have to reinvent the wheel because the design is open to us. With the opening up of the "Windows" wheel, we won't have to reinvent that one, either, and maybe, just maybe, we can improve it, changing the wheel from a rolling log into a spoked set of discs on an axle, with tyres and shocks. Pardon the analogy, but I think you can see where I'm going.
Basically, you hit on the problem, but misapplied it, coming up with the wrong solution. Or rather, you had the right solution, but translated it the wrong way.
I agree, we shouldn't have to reinvent the wheel. That is what we are trying to put out. And A monopoly can be good. But a monopoly on design, belonging to the people, rather than a monopoly on implementation, belonging to a company.
If Dunlop had a monopoly on the wheel, and priced it/did quality control to match, I'm pretty sure there would be alternatives that people would use, leading to fragmentation. Thankfully, they don't. The design of the wheel is open, leading to standardization as people realize what features are and are not useful.
Don't let a fear of fragmentation lead you into accepting a square wheel.
--
- Sean
fortran is evil
Perhaps... but better than the "competition"... (aka: COBOL).
--
- Sean
...and click the little button at the bottom that turns all the blinky, flashy things off!
--
- Sean
Hi! I'm al Gore and I invented the web. I mean the internet. Nono... I mean ethernet... no, um... oh hell. I used BackOrifice once.
--
- Sean
And more, I bet I can write a Perl script that will generate similar texts which, with very little editing, would be of comparable quality.
That's it! Thieme is a Perl Script!
hehehehehehehe...............
--
- Sean
this article actually hurt my head
Oh good! I'm glad it's not only me... Man, that was a tough read!
I struggled through a lot of it, re-read most, and still am not entirely sure of the author's point...
I'm sure it was a good one, but the style is so confrontational that I didn't get it.
--
- Sean
Warning: This is probably going to be a bit long-winded. This is because I have a number of related things that I want to say, and they are all sort-of jumbled up. So.
.sig file and how to set the Reply-To adress. I read it, and turned to my mother, who couldn't for the life of her understand what she had done to provoke this guy's ire. She was wondering why he was being so condescending and, yes, arrogant.
:-)
:-)
Yes. Hackers are arrogant.
Why? They just are. It's the so-called Nature of the Beast. Asking whether or not a hacker is arrogant is akin to asking whether or not a fish can swim.
And the arrogant hacker, as with the swimming fish, has been brought about by necessity.
There are two main contributions to this arrogance. The first is the much-vaunted "gift culture" that is so predominant. In the absence of "real" money, reputation counts as currency. How many times reading a day's worth of Slashdot, do you come across the phrase "show me the code!"? This phrase is somewhat of a double entendre (excuse me for iterating through obvious concepts; I'm just trying to be thorough). The first meaning refers, obviously, to the notion of code/source being open to the community. The second, and more relevant meaning is that a person's reputation is built almost entirely on what they have done. This point is important; I will return to it later.
If a person has not written code, they have zero reputation. For this reason are people such as Linus, Alan Cox, RMS and so on so highly respected. They have contributed a lot of very useful code to the community. For the same reason are people such as ESR, Tim O'Reilly and so on -- people who have coded, but are known more for other things -- somewhat less respected. Are they respected? Yes, but less so, especially in the Linux/Slashdot/Open Source community.
This reputation is treated as currency; if there is no respect for your code, you are no-one. Alternatively, if your code is respected, people listen to you and will do things for you.
I have been derided on Slashdot for the fact that I am a "VB Code Monkey." This is not considered coding. I have no problem with that; agree with it for the most part.
Now, this contributes to arrogance. When you have coded something, you are respected (and rightly so) for it. You become arrogant. This is RMS to a tee.
I mentioned two contributions. The other is the constant tug-of-war that the hacker must play with their code. Hackers tend to be introverted -- we spend most of our time playing with code. Code that can become annoyingly frustrating at times, yet code which is our bread-and-butter, our oxygen even.
Always forefront in the hacker's mind is the realization that if their code doesn't work, there is one thing at fault: the person who wrote it. That's the great thing about computers -- if your program ain't working, you got no-one to blame but yourself. But it's this same relationship between the hacker and their code that brings out the dominant traits of the hacker. How many times have you stared at a piece of code that would not work and wished that you could rip it off the screen and give it a good smack?
Ok, I'm being a bit melodramatic -- or am I? Perhaps not in such a corporeal manner, but there have been many times when I have wanted to force the program to to what I wanted, not what I said. When I have sat, cursing, at the code. When I have all but ripped the monitor from my desk and heaved it out the window.
And yes, I am displaying arrogance toward my program.
And before you deride me for being silly and childish, actually think about the relation between you and your code. You are constantly forcing the code to your will. You are constantly making it do what you want it to. And is this not arrogance?
If not, I don't know what is.
Because the hacker is always in a dominant position over his/her code, that tends to translate to their dealings with other people. Because they have a similar relationship with other hackers (who are the people they tend to deal with), where one person always has a bigger reputation than the other, based on the code they have written, this permits them to display arrogance towards the other. Thus the behaviour pattern is further entrenched.
Ok. As I said, I'm being pedantic, and probably have said nothing that isn't common knowledge. Hackers are arrogant. Out of necessity. And it's a good thing. If we weren't, we would be far less efficient at what we did. Namely, write good code.
Unfortunately, the combination of this arrogance with the fact that non-coders are not afforded respect means that few "outsiders" are allowed entry into the ranks. Oh, sure, they are allowed to come in, but they are quickly turned off and leave.
This I'm not so sure is a good thing.
My mother uses computers; she is a chemistry college teacher, and certainly intelligent in her own right. However, she is by no means a hacker. She sent an email recently to a colleague, which got misdirected to the sysadmin at that college. Because she is not a "hacker," or even an "uberuser," she didn't know about the Reply-To line, or using a signature. So she just included within the text of the eMail where the person could reply to.
The sysadmin replied, informing my mother that the eMail was mis-addressed. When I next spoke to her, she showed me his (the sysadmin's) eMail, and I read it. He had also pointed out that it wasn't necessary to include the return address in the eMail text, and included instructions on how to set up a
Stunned, I re-read the email. There was no condescension, anger or arrogance that I could see. It was just a letter, rather polite, I thought, explaining to her how to set up her preferences and stuff. Until I looked at it through her eyes -- those of a non power-user, who has no interest in computers beyond them being a tool to write her exams in Word, make her marking tables in Excel and send off email to her colleagues in Pine. All of a sudden, this email was written by some holier-than-thou guy who was demanding that she make a whole bunch of changes to some obscure file because he didn't want to see a return address embedded in the text, where he wasn't expecting it.
Now I re-iterate that I found it to be completely innocuous. It was not until I looked at it as someone who made minimal usage of computers that I saw this.
And I realized that this attitude was very pervasive in the hacker community. I wouldn't normally give it a second thought. Hell, I wouldn't even notice it for the most part. Because, as I showed above, it is not only natural, but necessary. It is a fact of life. The nature of the beast.
Imagine now that you are a regular user... a luser. Perhaps a reporter covering a Linux Expo event. Maybe even a chemistry teacher. Someone who uses computers minimally. Someone who definitely does not code. Someone who doesn't necessarily even want to know more about computers, because it is not relevant to you. Someone deserving the contempt of any hacker. Someone who... whoops! I think we see the problem.
Now we are at a dilemma. The hacker needs to be arrogant. But at the same time, the hacker's arrogance turns off other people. How do we solve this problem?
Before I continue, I recognize that there are many people who do not necessarily consider this a problem to be solved. That is your prerogative.
I argue, however, that it is. The reason why is the old world-domination thing. We can't get more users if we keep scaring them off. We can't have world domination if we don't get any new users. And I believe that world domination is necessary. At least a little bit. Does Linux need to be a Microsoft-killer? No. But it must reach "critical mass". I'll explain that in a second. I think that we are almost there. But not quite. We are approaching it, though, and it is a heady time. Linux are getting "official" recognition from Oracle, IBM, Dell, Corel, you name it. Linux is maturing. But it's also a dangerous time. If we turn people off now, we may not get this chance again. I remember OS/2 being in the same position about 5 years ago. I was strongly involved in the OS/2 community at that time. And one of the common complaints was that people involved with TeamOS/2 were needlessly arrogant toward outsiders. Exactly the same complaint as is being levelled at the Open Source community today. OS/2 almost reached critical mass, but blew it. I think Linux has a good chance, and don't want to see the same thing happen.
And I define critical mass is that point where it is self-supporting. Yes, the fact that Linux is open source is a tremendous advantage over OS/2, and all the proprietary OS'es that have either succeeded or failed up until now. But don't fool yourself into thinking that it is self-supporting. All the open source in the world won't build drivers for closed hardware interfaces. All the open source in the world won't build support for new types of hardware. Once the user base reaches a certain size, that in itself will be enough incentive for companies to either write drivers, or release the specs. That's the critical mass.
But it's not quite there yet.
It needs just a little bit of a push... if we don't stop that push before it even starts.
So what do we do? We are arrogant, because it's needed for what we do. We do what the "standard" software companies do. We get a spokesperson. Or spokespeople. It would be a bit hard to swallow, I'll wager, because such a person would not be a hacker. Such a person would not write code. Such a person would never truly "understand." Such a person would be very similar to Jon Katz, and we know what controversy he creates around here.
But he is precisely the sort of person we need. Someone who can talk to "the public" in terms that they can understand. Someone who doesn't talk code. Someone who doesn't necessarily get it all right, because there are a lot of people out there who simply don't care.
Your average user has absolutely no use for detailed knowledge of programming, or even computers in general, any more than a hacker has use for detailed knowledge of open-heart surgery. It just isn't applicable. But yet we will jump to condemn these people for not wanting to learn. Well, surprise! Some people don't want to. Does this make us better than them? Perhaps. But it's probably not a good idea to let on. Unfortunately, the hacker's natural arrogance tends to reveal these true feelings. We don't have to hide our feelings from the computer, therefore we aren't used to hiding them from people.
Which is why a hacker should not be representing hackers to the "rest of the world". Unfortunately, RMS is a hacker. He insists that other people follow his instructions, just as he insists that the computer follow his code. And he is possibly one of the worst representatives we could have.
A plea, then, for sane representation
Or someone who can talk to the non-hackers in a way that won't threaten them. Yup, he (or she) will get a few things wrong. Sure, he won't totally understand. And he/she may even say a few things that a hacker would consider utterly moronic.
But better that than appealing to the hacker, and driving everyone else away.
-----
Well. I applaud you for getting this far, because I know my style can be rather pedantic. I also haven't been able to say quite what I wanted to say, but I think it's close enough. You're going to have to do a bit of interpretation there on some of the points I've made. Basically, if what I said sounded really strange, or a bit loony, it probably came out wrong
Anyway, I'm just trying to give people something to think about.
--
- Sean
--
- Sean
If anything, it should be considered a shallow learning curve -- for a given time period (delta- x ), the gain in proficiency will be minimal (a small delta- y ). Represented thusly (note: you may have to widen your browser to see the whole graph properly):
In reality, however, even this is somewhat misrepresentative. That's not the way people tend to learn any given product.
A "real-world" learning curve would actually have two focii, splitting the curve into 3 sections:
To wit:
Obviously, in this latter case, it is of little relevance to say whether or not the curve is "steep" unless you also mention what part of the curve you are referring to.
It is not alway correct to apply a generalization (the curve for such-and-such a product is steeper/shallower) to the curve as a whole. In the case of Linux vs Windows '9[58], for example, I belive that the initial part is much, much, much shallower (ie: harder to learn) for Linux, whereas once you get far enough along the curve, the second part becomes steeper (ie: easier to learn). Also, any comparaison is totally non-applicable to the third part, which is rather short and comes early in the case of Windows, but is much longer and comes much later for Linux.
Just my $0.02 worth (or by this point $2.00...?)
--
- Sean
One for every post the user has ever posted.
:-)
Another for the average over the last month (or something).
That way, among other things, you could see what posters were getting better (and which were getting worse).
...hmm... methinks this be beginning to sound too much like the dreaded grade-school report card...
--
- Sean
--
- Sean
This is very strange... I'm running MSIE (at work; no choice) on a Windows box with MS' "embraced and extended" character set-thingie, and I'm still getting the ?'s instead of ''s...
I'm beginning to think it's a different problem this time, although I can't imagine what...
--
- Sean
It really depends on the version of VB you use.
.exe (or as VBRun100.dll, VBRun200.dll, VBRun300.dll, VB40016.dll, VB40032.dll -- should be self explanatory).
As with everything MS, every new version tends to introduce a fair amount of bloat.
If you use the original v1.0, you can get it down pretty small. Actually, versions 1 through 4 basically "compile" it to a form of Bytecode, then package it up with the interpreter embedded in the
The size of these things has steadily increased from about 100k (I think) for VBRun100.dll up to almost a meg for VB40032.dll.
But in any case, the program itself was not much more than what you have above. In fact, it was sometimes smaller than the actual VB code.
Versions 5 and 6, however, offer the ability to do a "true" compilation (into native code). Although the program will run faster (95 times out of 100), it all of a sudden requires huge globs of "things" to run.
Exactly what these "things" are tends to vary... for the most part, they are dll's and ActiveX components. Most of which are not required by your actual program (assuming you are using the Hello World example above).
However, not requiring them and not having them are 2 different things.
Microsoft, aiming VB at the lowest common denominator, tends to lump everything it thinks you may (ie: commonly) need into the package whether you ask for it or not.
The result is that the size of your executable + associated DLL's all of a sudden mushrooms to the vicinity of 2 1/2 megs.
Of which you only need (for your simple "Hello World" program) MSVBVM50.dll (the Virtual Machine stuff -- it actually is Java-like in that respect -- "native code" is a "bit" of a misnomer), which clocks in at 1.28 megs. But don't even think of trying to get rid of the rest. It has so many dependency tables that reference the other crap that deleting them will break your program irrecoverably (unless you put 'em back).
So is there a way to get rid of them? Yes, futz around with the options a whole bunch before you compile. You will also have to add and remove a few lines from your project (*.vbp) files by hand (the IDE doesn't offer the option), as well as a couple of other config files. But don't do it unless you know what you're doing. You won't succeed.
And me? I know because I have to work with this shit (and that's what it is) for a living.
YUCK! LEMME OUTTA HERE!!!
--
- Sean
Lots...
There are lots of things that I can say that would require space in the 'hundreds of pages' range.
Are you trying to say that all books are unnecessary, that all books can be whittled down to a coupla paragraphs...?
Somehow, I doubt it...
--
- Sean
LOL!
But howabout the other way around?
I mean... how do you know if you are a moderator? Does it show up in your user prefs, or something?
--
- Sean
Yes... I really like that idea...
I've thought about it myself, not so much in the context of Slashdot, but as a general principle.
Working out algorithms to maximize efficiency and utility, and so on...
The main drawback, I think, would be the hardware requirements, in terms of processing, storage space for the DB, server load, and so on. My experience is that those sorts of interrelated sets and the associated processing are a HUGE drain on the processor, require lots of memory, and also require enormous databases (at least, to do properly).
Even with Slashdot's spiffy new server, I don't think it meets the requirements. Yet.
But maybe someday... we can always hope...
--
- Sean