Maybe it will come back as a series of signed files on Freenet. After all, Freenet was designed to resist censorship. Of course that would put an end to the speedy de-listings many ISP's got after booting their spammers. I guess it would go to a daily release of the zone file.
I understand how unpleasant this is for you, but I also understand the ISP's viewpoint. Mail from a cable modem address is very likely to be spam. As someone operating a legitimate server on a residential address, you're too tiny a minority for these giants to notice.
Your colo price is way too high. You can get a dedicated server for $100/month at Rackshack and many companies on webhostingtalk. Or, if you rebuild your server into a 1ru case, you can colo it for $40-$50 / month. But if you know someone who rents a full rack at a colo, they might accept your box for $20 or $30. Ask on the isp-colo list.
If you know a small business owner with DSL, you might arrange the same deal. They usually get 5 IP addresses and only use 1. Also try small local ISPs. They may already have "obsolete" boxes in their racks that they'd like to rent. Emphasize that you're a low-bandwidth, low maintenance customer.
The economy is down, and many folks overinvested in network infrastructure. It's a buyer's market.
That reward is just a way of publicly backing up DJB's belief in his code quality. I would think it's obvious that the motivation of anyone researching a qmail vulnerability is the advancement of knowledge and the recognition. By publicizing this award, Bernstein has increased the attention on this issue, guarranteeing more recognition for the person who claims it.
The award is significant, not for its amount, but because it's a very rare public declaration by a software author/vendor that his code is secure.
Who told you that the abstraction was going to solve every problem for you? Did you think the C++ Pixies were going to arrive and write your code for you?;-)
You make it sound silly, but I've heard roughly that sentiment from C++ advocates. I describe what I'm working on in C, and they say, "sounds like a job for C++." Why? "Because the problem space you describe has objects and inheritance."
What they miss, which grandparent was trying to express, is that the "objects and inheritance" part of the problem is the easy part. There's still a ton of hard work at the heart of the problem. When I'm putting in that ton of work, it really doesn't matter how I handle method dispatch, for example.
It's actually kind of insulting to the depth of the work to say it could be better written in C++. At least it shows that the advisor has no idea where the real meat of the problem lies.
I'm not against C++, but at its current state of maturation it is more of a distraction than C is. C is a humble tool which gets out of the way and leaves the programmer free to build the extended paradigms and programming methods needed for a particular app.
You're probably right about GUIs. I'm not sure you're right about embedded though - I know of one embedded C++ app, and the author certainly had the background to choose C if he had wanted to.
No, I think you missed the point. C was a sweet spot in the evolution of computer languages. Nothing since then has delivered that much improvement over what went before, although I think Perl comes close.
The entire journey into OO with java and C++ will probably be regarded as a bizarre and pointless fad by future programmers.
Not being an expert on the topic, I find DNSSEC a little worrying, as it seems to be a consolidation of the centralized power of Verisign or whatever. Ideally we should be planning how to move away from traditional DNS altogether, as the single-rooted namespace has led to much political abuse. But that is a really hard problem to solve.
I have to disagree with your math. The $350k is not the budget for the whole department, just for the expertise to replace RH support. You don't need your final guru in shifts - let him work in the day and wake him with a pager if an emergency occurs. The kind of problem you need him for usually isn't a middle-of-the night emergency anyway. It's more like "application foo keeps segfaulting, we tried recompiling, tried the latest version, Google is no help and nobody else has this problem." It probably takes the sysadmin staff a couple of days just to refine the problem to guru level.
That sounds good, but how far out of date is stable? Do your lab users get annoyed with running old software? I'm especially thinking of browsers, where there's a strong incentive to run the newest. Can you use apt-get to install a package that's not in stable? Or would it break due to dependencies?
Realistically, how do you run an up-to-date Mozilla/Galeon/whatever on Debian?
I am not attacking Debian; I'm genuinely curious. I currently use Red Hat and am evaluating Gentoo and Debian since it's clear RH wants to ditch users like me.
No, the letter clearly states on whose behalf they are acting. This part is true, and it's the only part sworn under penalty of perjury. The DMCA is carefully written to protect the complainant in a case like this.
The major problem here is that while the DMCA requires the complainant to swear under penalty of perjury that he is an authorized representative of the copyright holder, it deliberately does not require the complainant to swear that an infringement has occurred. That loophole enabled the automated abuse we are now seeing.
One only has to look at GNOME and KDE to see how much the Unix landscape has been infiltrated by people that just don't get it.
While I understand where you're coming from, that isn't quite fair. KDE and Gnome appear to be attempts to produce a Windows-style GUI atop Linux. This doesn't mean the developers don't get it - it could mean that they get it and yet prefer a Windows-style GUI for some or all tasks. Or that they see a need for such a GUI to let "normal" people use Linux.
Subjectively, I feel the same alienation you are expressing from KDE and Gnome.
This shouldn't be confusing. How about foo[ bar ][ foo2 ]? Indexing into a 2-dimensional array in C. Your example is the same except that we index associative arrays.
Actually, even if SCO is right in their claims against IBM, they could still be liable to Red Hat for damages arising from the way they've conducted themselves during the suit. It does look like they've gone out of their way to harm Red Hat's business.
To take an obvious counterpoint, when Unisys pursued their GIF patent claims, they didn't issue press releases saying they would destroy Amazon and Ebay.
Reed tried to educate us about how the system works. You are now discussing how the system should work - that is, you're educating us about a utopian or ideal world. While that has its place, Slashdotters desperately need to learn how the real world works.
Your moral indignation at campaign contributions is duly noted. But it will not help you win in the legislative arena, where our futures are being decided.
If I gave you $100,000 to argue for cracking eggs on the little end, and gave Reed $50,000 to argue for cracking eggs on the big end, I'm sure Reed would win. Despite having a smaller war chest, he's far more realistic and perceptive about the people he's trying to influence. You might approach legislators with thinly disguised contempt and a complete lack of understanding of what their lives and work consist of. That makes you equivalent to a boss that despises programmers for being nerds, and condescendingly flings mandates at them.
If you reply, please tell me whether you agree. Would your larger war chest enable you to conquer Reed? Or does he actually have talents that outweigh a 2:1 financial advantage?
I think you missed the point. Even granting your premise that more expensive lawyers are more effective, you bought the results from your lawyer, not from the court. My point is that paying court fees does not guarrantee a desired outcome - it's merely the price of admission. Likewise, you cannot simply write a check to a legislator and buy votes. Campaign contributions simply buy you a hearing. You still must have convincing arguments.
Yes, this is ridiculous. It's as if these head-in-the-clouds corporate types are on a completely different planet. Kind of like Microsoft with ".NET". And the sad thing is that they're running over something that actually works to make space for something that may never work.
Slashdot Mirror
Maybe it will come back as a series of signed files on Freenet. After all, Freenet was designed to resist censorship. Of course that would put an end to the speedy de-listings many ISP's got after booting their spammers. I guess it would go to a daily release of the zone file.
Can you cite some examples of such manipulation?
Please tell me more about these ISP-critical machines that don't affect innocent users. But then why are they critical?
As for narrowly listing spammers, it's been tried. Sleazy ISPs move the spammers around to evade such blocks.
I understand how unpleasant this is for you, but I also understand the ISP's viewpoint. Mail from a cable modem address is very likely to be spam. As someone operating a legitimate server on a residential address, you're too tiny a minority for these giants to notice.
Your colo price is way too high. You can get a dedicated server for $100/month at Rackshack and many companies on webhostingtalk. Or, if you rebuild your server into a 1ru case, you can colo it for $40-$50 / month. But if you know someone who rents a full rack at a colo, they might accept your box for $20 or $30. Ask on the isp-colo list.
If you know a small business owner with DSL, you might arrange the same deal. They usually get 5 IP addresses and only use 1. Also try small local ISPs. They may already have "obsolete" boxes in their racks that they'd like to rent. Emphasize that you're a low-bandwidth, low maintenance customer.
The economy is down, and many folks overinvested in network infrastructure. It's a buyer's market.
Never having heard of them, I looked on Google Groups.
Maybe AOL users don't want to become Millionaire Traders, join the Big Money Downline Club, or read a free $3600 Money Game Report.
That reward is just a way of publicly backing up DJB's belief in his code quality. I would think it's obvious that the motivation of anyone researching a qmail vulnerability is the advancement of knowledge and the recognition. By publicizing this award, Bernstein has increased the attention on this issue, guarranteeing more recognition for the person who claims it.
The award is significant, not for its amount, but because it's a very rare public declaration by a software author/vendor that his code is secure.
You make it sound silly, but I've heard roughly that sentiment from C++ advocates. I describe what I'm working on in C, and they say, "sounds like a job for C++." Why? "Because the problem space you describe has objects and inheritance."
What they miss, which grandparent was trying to express, is that the "objects and inheritance" part of the problem is the easy part. There's still a ton of hard work at the heart of the problem. When I'm putting in that ton of work, it really doesn't matter how I handle method dispatch, for example.
It's actually kind of insulting to the depth of the work to say it could be better written in C++. At least it shows that the advisor has no idea where the real meat of the problem lies.
I'm not against C++, but at its current state of maturation it is more of a distraction than C is. C is a humble tool which gets out of the way and leaves the programmer free to build the extended paradigms and programming methods needed for a particular app.
You're probably right about GUIs. I'm not sure you're right about embedded though - I know of one embedded C++ app, and the author certainly had the background to choose C if he had wanted to.
No, I think you missed the point. C was a sweet spot in the evolution of computer languages. Nothing since then has delivered that much improvement over what went before, although I think Perl comes close.
The entire journey into OO with java and C++ will probably be regarded as a bizarre and pointless fad by future programmers.
That's pretty cool. Did the junk peak in 1997? Are we heading back to quality?
Of course, no discussion of DNSSEC would be complete without Bernstein's comments. And here are the slides from his talk in pdf.
Not being an expert on the topic, I find DNSSEC a little worrying, as it seems to be a consolidation of the centralized power of Verisign or whatever. Ideally we should be planning how to move away from traditional DNS altogether, as the single-rooted namespace has led to much political abuse. But that is a really hard problem to solve.
I have to disagree with your math. The $350k is not the budget for the whole department, just for the expertise to replace RH support. You don't need your final guru in shifts - let him work in the day and wake him with a pager if an emergency occurs. The kind of problem you need him for usually isn't a middle-of-the night emergency anyway. It's more like "application foo keeps segfaulting, we tried recompiling, tried the latest version, Google is no help and nobody else has this problem." It probably takes the sysadmin staff a couple of days just to refine the problem to guru level.
That sounds good, but how far out of date is stable? Do your lab users get annoyed with running old software? I'm especially thinking of browsers, where there's a strong incentive to run the newest. Can you use apt-get to install a package that's not in stable? Or would it break due to dependencies?
Realistically, how do you run an up-to-date Mozilla/Galeon/whatever on Debian?
I am not attacking Debian; I'm genuinely curious. I currently use Red Hat and am evaluating Gentoo and Debian since it's clear RH wants to ditch users like me.
That sounds like a way more interesting story than yet another SCO dropping. Why don't you write it up and submit it here or on kuro5hin?
No, the letter clearly states on whose behalf they are acting. This part is true, and it's the only part sworn under penalty of perjury. The DMCA is carefully written to protect the complainant in a case like this.
The major problem here is that while the DMCA requires the complainant to swear under penalty of perjury that he is an authorized representative of the copyright holder, it deliberately does not require the complainant to swear that an infringement has occurred. That loophole enabled the automated abuse we are now seeing.
While I understand where you're coming from, that isn't quite fair. KDE and Gnome appear to be attempts to produce a Windows-style GUI atop Linux. This doesn't mean the developers don't get it - it could mean that they get it and yet prefer a Windows-style GUI for some or all tasks. Or that they see a need for such a GUI to let "normal" people use Linux.
Subjectively, I feel the same alienation you are expressing from KDE and Gnome.
This shouldn't be confusing. How about foo[ bar ][ foo2 ]? Indexing into a 2-dimensional array in C. Your example is the same except that we index associative arrays.
Actually, even if SCO is right in their claims against IBM, they could still be liable to Red Hat for damages arising from the way they've conducted themselves during the suit. It does look like they've gone out of their way to harm Red Hat's business.
To take an obvious counterpoint, when Unisys pursued their GIF patent claims, they didn't issue press releases saying they would destroy Amazon and Ebay.
It's pretty easy, actually. If @x is a list of hashes of objects of class foobar, which is what I think you mean, you can:
Far from being a nightmare, the ease of such manipulations is one of the nicest things about Perl.
Actually, you could make your point more strongly by omitting the parentheses after the function calls:
$time = time;
etc.
And doesn't that line kind of warrant the comment:
#Dread at control, mon, an' I an' I come a Freeside when I an' I come.
Reed tried to educate us about how the system works. You are now discussing how the system should work - that is, you're educating us about a utopian or ideal world. While that has its place, Slashdotters desperately need to learn how the real world works.
Your moral indignation at campaign contributions is duly noted. But it will not help you win in the legislative arena, where our futures are being decided.
If I gave you $100,000 to argue for cracking eggs on the little end, and gave Reed $50,000 to argue for cracking eggs on the big end, I'm sure Reed would win. Despite having a smaller war chest, he's far more realistic and perceptive about the people he's trying to influence. You might approach legislators with thinly disguised contempt and a complete lack of understanding of what their lives and work consist of. That makes you equivalent to a boss that despises programmers for being nerds, and condescendingly flings mandates at them.
If you reply, please tell me whether you agree. Would your larger war chest enable you to conquer Reed? Or does he actually have talents that outweigh a 2:1 financial advantage?
I think you missed the point. Even granting your premise that more expensive lawyers are more effective, you bought the results from your lawyer, not from the court. My point is that paying court fees does not guarrantee a desired outcome - it's merely the price of admission. Likewise, you cannot simply write a check to a legislator and buy votes. Campaign contributions simply buy you a hearing. You still must have convincing arguments.
Lawyer::Court Fees::Court = Lobbyist::Campaign Contribution::Legislature
Thanks.
I can't seem to find the demo download. Where is it?
Yes, this is ridiculous. It's as if these head-in-the-clouds corporate types are on a completely different planet. Kind of like Microsoft with ".NET". And the sad thing is that they're running over something that actually works to make space for something that may never work.