On a related note, has anyone looked at spamassassin and spamcop integration? I would like to pass all the caught spam through spamcop (manual review of each before sending is fine) and hopefully shut some more spammers down. This is the one thing I miss since moving to spamassassin -- my previous spamcop efforts helped shut monsterhut down:-)
They've fixed them, or at least some of them. The F3 now has an odometer check length. But "speedometer" check length actually works better for their purposes, because the reason the check lengths are there is to give you something to do every so often so you don't fall asleep and die. And it certainly seemed to get *you* thinking, even if you didn't count off the clicks:-)
Re:Of course, they are all turning compatable
on
The Problem Of Developing
·
· Score: 2, Informative
I'm sorry? What's the connection between the amount of storage on the device and the ability to read that storage from a long distance? And why is six lines a limiting factor when eight bytes is more than enough to store your shiny new serial number in?
While setting up CUPS last night I found this tool which, according to the page, provides the one last thing not in CUPS: a GUI to adjust printer settings and schedule jobs. These things aren't covered properly/at all by the CUPS web interface. For myself, I'm right with lpoptions for my configuration needs and lp for scheduling, but if you miss Print and Print Setup in Windows you might want to check out XPP.
replace the connector (which is basically a bundle of wires in shrinkwrap) with slightly longer cabling
LCD signal cable suffers horribly from attenuation, so the cable is slightly special and has a very limited maximum length. Last time I built an embedded system and needed a longer display cable than what was supplied, it cost about AU$330 for modulator boards and a 3m cable. Of course, I think AU$330 is about 37 US cents these days, so maybe that's not so bad.
Why is an OS with a "built-in" browser more useful than an OS on which you can install a browser with the same features?
Other programs can then embed the functionality of that browser using OLE or whatever, with the guarantee that it will be present. HTML editors, graphics viewers, documentation systems and other things that would like to render HTML can simply rely on there being a browser available for embedding, rather than rolling their own solution, packaging someone else's in as libraries, or throwing a copy of a browser onto the CD and having a "you look deprived! here, let me install 80MB of browser and then you can use LeetEdit v0.37" message during installation.
What makes you say this? You're inside the system; you can't possibly tell what's going on outside it. In any case, until you define your terms better I'd say that "unconscious" and "undirected" apply to the operation of a GA.
how do you tell if a certain modification introduced by the system, is actually an improvement?
GAs are graded using a fitness function, which represents your concept of how well an instance meets whatever purpose you had in mind. Defining the fitness function is hard work for complex problems, but once you have it you can tell straight off which modifications are improvements and which are not.
The problem is not being unable to test fitness during self-improvement, but being unable to define the test before you start. To take the extreme case, how do you determine a particular human's fitness? What is the end-goal of human evolution anyway?
What if the modifications makes the program slightly faster
Using program execution speed as a fitness function doesn't seem to represent any particular problem you'd like to solve. In general I can't imagine it being used as one.
Well, that's a nice oversimplistic explanation which is totally wrong.
It's not an explanation, it's something to think about. I'm aware of the standards mandated for domain names, and if you think the fact that you can register any 2 to 63-character name is a solution, I invite you to name your next enterprise ua7gnc6209hhaycn.com.
Your solution of creating many, micro-managed TLDs is unworkable.
I beg to differ; Australia makes a very good fist of it..com.au is entirely composed of registered companies..org.au is entirely composed of legitimate non-profit organisations..net.au is reserved for internetworking companies..edu.au contains national learning institutions and is also subdivided by state. Each of these domains is controlled by a different organisation, and it all works. I could go on, we have several more useful 2lds.
I'll also add its unusable -- how do I know which TLD to find something?
Use a search engine. The DNS is not a directory service. How do you phone redhat when their phone number isn't 1-800-REDHAT?
The "problem" won't be settled until there's no TLDs to fight over.
The "problem" is an obvious scarcity of domain names. Increasing that scarcity does not solve the problem.
This is the big reason we need a hierarchial namespace, and you seem to be conveniently ignoring it. Why do you think a company can usually only trademark a given name in one of many categories of product and service? (True in Australia, probably true in most places.) It's because there is a finite number of possible identifying words for products and companies, and people have to share. We don't need fewer TLDs, or none, we need more, one for each category of trademark, ones for all kinds of organisations, ones for individuals. And they should be subdomains of the country in which the entity exists, since there are few world organisations in the legal sense, and few global citizens. You don't stop the fighting over a limited resource by constricting the availability of that resource, you stop it by finding a way through which most people can share. This is what the hierarchy is for.
You know, some of us have been around long enough to recognise text copied from the (1990 vintage) Jolly Roger's Cookbook. Just because the author of the Cookbook took large amounts of the text from other people without attribution doesn't make it good manners for you to similarly attempt to palm it off as your own work.
I believe mandatory voting, as required in some nations, is ridicuoous.
We have mandatory voting in Australia. What's good about it? Well, it means that every single adult in the country got out of bed and trundled down to a polling booth. Every adult is partially responsible for the current set of people governing them. I believe that the USA has about 30% voter turnout, and it's my personal opinion that more than 30% of the population of both countries have some idea of whom they'd like in power. In Australia, all the people who think "This seat is going to the Liberals no matter what, why should I bother voting Labor?" get out and vote anyway.
Also, having 100% voter turnout is a huge stabilising influence. The preference of the person who supports a party, but isn't exactly marching through the streets over it, is just as important as that of the person who has absolutely rabid support for someone. It's great to hear from them. In the last elections we had a party called One Nation which may have even made the news slightly in the USA. They had very strong policies on immigration, welfare and race relations which would probably have been rather damaging if they'd won power, but because we have 100% turnout, any candidate needs a majority of the actual populace in his seat to support him, not a majority of the people who can be bothered to vote. This keeps our government stable.
Of course, if you really think all parties suck equally, and you can't find even one to support (there tend to be about 6 lower house candidates for each seat, and about 100 upper house candidates for each state) then you can put a big line through your ballots, or leave them blank, or write a charming message for the scrutineers. But you've consciously said "you all suck!" Your opinion will be counted and noted. Unlike not voting at all, this really means something.
Replace the phrase "open source" with the word "commercial software," and the assertions are just as true.
They may be just as true, but they're not desirable, and free software means that things don't have to be this way.
There seems to be a really sharp line between the kind of quality, robust software that a lot of us made the switch away from Windows for, and the bits and pieces of junk that we have to use for other tasks. I have a stable OS, a stable windowing system and window manager, a stable text editor, all full of the features I need to get stuff done. But I have an ICQ client and a Napster client which are maybe 50 to 75 per cent of the way along, and these are the best of breed. I've tried them all. The feeling of "It Just Works" that I once had whenever I needed software to do some new thing is all gone! Who does it serve for coders to release buggy, crashy, badly-written, undocumented code? If you're writing for public release, write well, make it stable, and accept patches. If you're writing YATE as a way of learning to code, keep that code in your pants! Freshmeat doesn't need you to submit console editor number 180. The world at large doesn't benefit from your "My First C" style coding job. So decide what your aims are. If you want to learn to code, certainly start a little project, but don't clutter the world with it unless it's actually release-quality. If you want to add to the pool of programs that people use, get involved with an existing project. That's how you add the most. It doesn't matter how brilliant your new editing feature is if it's been added to YATE that only you and some guy from Sweden are ever going to use.
FreeBSD users will be pleased to learn that the linux-redhat client works fine under emulation. brandelf -t Linux client and run as per the instructions.
50,000 years from now, human beings may be
creating our own universes. Perhaps we will be able to spawn life as well. 50,000 years from now, can we realistically expect evolved human beings to
be able to read our "data"? I doubt it.
I don't see how one follows from the other. Surely the same evolution and development that has given humans the power to create universes has also finally taught them to RTFM (one of which is enclosed in the capsule).
Turning the data on the discs back into the original characters should be extremely easy for the super-humans. Understanding it will be more interesting, just as understanding writings from our own ancient history is interesting and challenging, but ultimately largely achievable and well worth the effort.
It's hard to take any benchmark, or any study in general, seriously unless it sets out the methodology well enough that someone else could run the same tests himself. I use postgresql for several websites and I would love to get the kind of performance that was reported in the benchmark. Something along the lines of "our tests report that postgresql stomps on other databases, and here's how you can make your own postgresql installation do the same" would be so much more useful and credible than "postgresql is better, we've got a couple of numbers which prove it."
Slashdot Mirror
On a related note, has anyone looked at spamassassin and spamcop integration? I would like to pass all the caught spam through spamcop (manual review of each before sending is fine) and hopefully shut some more spammers down. This is the one thing I miss since moving to spamassassin -- my previous spamcop efforts helped shut monsterhut down :-)
Please, the phrase is "copy prevention." You aren't being protected from anything. You're being prevented from exercising your rights.
They've fixed them, or at least some of them. The F3 now has an odometer check length. But "speedometer" check length actually works better for their purposes, because the reason the check lengths are there is to give you something to do every so often so you don't fall asleep and die. And it certainly seemed to get *you* thinking, even if you didn't count off the clicks :-)
Mate, Turing complete. As in, Alan Turing. As in the father of computing.
I'm sorry? What's the connection between the amount of storage on the device and the ability to read that storage from a long distance? And why is six lines a limiting factor when eight bytes is more than enough to store your shiny new serial number in?
While setting up CUPS last night I found this tool which, according to the page, provides the one last thing not in CUPS: a GUI to adjust printer settings and schedule jobs. These things aren't covered properly/at all by the CUPS web interface. For myself, I'm right with lpoptions for my configuration needs and lp for scheduling, but if you miss Print and Print Setup in Windows you might want to check out XPP.
LCD signal cable suffers horribly from attenuation, so the cable is slightly special and has a very limited maximum length. Last time I built an embedded system and needed a longer display cable than what was supplied, it cost about AU$330 for modulator boards and a 3m cable. Of course, I think AU$330 is about 37 US cents these days, so maybe that's not so bad.
If only you had spelled the search term correctly. A winnar is not you!
Other programs can then embed the functionality of that browser using OLE or whatever, with the guarantee that it will be present. HTML editors, graphics viewers, documentation systems and other things that would like to render HTML can simply rely on there being a browser available for embedding, rather than rolling their own solution, packaging someone else's in as libraries, or throwing a copy of a browser onto the CD and having a "you look deprived! here, let me install 80MB of browser and then you can use LeetEdit v0.37" message during installation.
What makes you say this? You're inside the system; you can't possibly tell what's going on outside it. In any case, until you define your terms better I'd say that "unconscious" and "undirected" apply to the operation of a GA.
GAs are graded using a fitness function, which represents your concept of how well an instance meets whatever purpose you had in mind. Defining the fitness function is hard work for complex problems, but once you have it you can tell straight off which modifications are improvements and which are not.
The problem is not being unable to test fitness during self-improvement, but being unable to define the test before you start. To take the extreme case, how do you determine a particular human's fitness? What is the end-goal of human evolution anyway?
What if the modifications makes the program slightly faster
Using program execution speed as a fitness function doesn't seem to represent any particular problem you'd like to solve. In general I can't imagine it being used as one.
Reminds me of the Ministry of Truth...
See here. In essence it allows you to create virtual machines programmatically.
Care to share some of your suggestions then?
"Click here to find lowest price for AthlonMP 1.2 GHz."
I'll take two!
Except that the definition of "normal" was covered in the article, which surely everybody read before post--
Oh, really? I see.
It's not an explanation, it's something to think about. I'm aware of the standards mandated for domain names, and if you think the fact that you can register any 2 to 63-character name is a solution, I invite you to name your next enterprise ua7gnc6209hhaycn.com.
Your solution of creating many, micro-managed TLDs is unworkable.
I beg to differ; Australia makes a very good fist of it. .com.au is entirely composed of registered companies. .org.au is entirely composed of legitimate non-profit organisations. .net.au is reserved for internetworking companies. .edu.au contains national learning institutions and is also subdivided by state. Each of these domains is controlled by a different organisation, and it all works. I could go on, we have several more useful 2lds.
I'll also add its unusable -- how do I know which TLD to find something?
Use a search engine. The DNS is not a directory service. How do you phone redhat when their phone number isn't 1-800-REDHAT?
The "problem" won't be settled until there's no TLDs to fight over.
The "problem" is an obvious scarcity of domain names. Increasing that scarcity does not solve the problem.
cat /usr/share/dict/words |wc -l
235881
This is the big reason we need a hierarchial namespace, and you seem to be conveniently ignoring it. Why do you think a company can usually only trademark a given name in one of many categories of product and service? (True in Australia, probably true in most places.) It's because there is a finite number of possible identifying words for products and companies, and people have to share. We don't need fewer TLDs, or none, we need more, one for each category of trademark, ones for all kinds of organisations, ones for individuals. And they should be subdomains of the country in which the entity exists, since there are few world organisations in the legal sense, and few global citizens. You don't stop the fighting over a limited resource by constricting the availability of that resource, you stop it by finding a way through which most people can share. This is what the hierarchy is for.
A sample reference for comparison: http://www.textfiles.com/ana rch y/JOLLYROGER/010.jrc
We have mandatory voting in Australia. What's good about it? Well, it means that every single adult in the country got out of bed and trundled down to a polling booth. Every adult is partially responsible for the current set of people governing them. I believe that the USA has about 30% voter turnout, and it's my personal opinion that more than 30% of the population of both countries have some idea of whom they'd like in power. In Australia, all the people who think "This seat is going to the Liberals no matter what, why should I bother voting Labor?" get out and vote anyway.
Also, having 100% voter turnout is a huge stabilising influence. The preference of the person who supports a party, but isn't exactly marching through the streets over it, is just as important as that of the person who has absolutely rabid support for someone. It's great to hear from them. In the last elections we had a party called One Nation which may have even made the news slightly in the USA. They had very strong policies on immigration, welfare and race relations which would probably have been rather damaging if they'd won power, but because we have 100% turnout, any candidate needs a majority of the actual populace in his seat to support him, not a majority of the people who can be bothered to vote. This keeps our government stable.
Of course, if you really think all parties suck equally, and you can't find even one to support (there tend to be about 6 lower house candidates for each seat, and about 100 upper house candidates for each state) then you can put a big line through your ballots, or leave them blank, or write a charming message for the scrutineers. But you've consciously said "you all suck!" Your opinion will be counted and noted. Unlike not voting at all, this really means something.
They may be just as true, but they're not desirable, and free software means that things don't have to be this way.
There seems to be a really sharp line between the kind of quality, robust software that a lot of us made the switch away from Windows for, and the bits and pieces of junk that we have to use for other tasks. I have a stable OS, a stable windowing system and window manager, a stable text editor, all full of the features I need to get stuff done. But I have an ICQ client and a Napster client which are maybe 50 to 75 per cent of the way along, and these are the best of breed. I've tried them all. The feeling of "It Just Works" that I once had whenever I needed software to do some new thing is all gone! Who does it serve for coders to release buggy, crashy, badly-written, undocumented code? If you're writing for public release, write well, make it stable, and accept patches. If you're writing YATE as a way of learning to code, keep that code in your pants! Freshmeat doesn't need you to submit console editor number 180. The world at large doesn't benefit from your "My First C" style coding job. So decide what your aims are. If you want to learn to code, certainly start a little project, but don't clutter the world with it unless it's actually release-quality. If you want to add to the pool of programs that people use, get involved with an existing project. That's how you add the most. It doesn't matter how brilliant your new editing feature is if it's been added to YATE that only you and some guy from Sweden are ever going to use.
FreeBSD users will be pleased to learn that the linux-redhat client works fine under emulation. brandelf -t Linux client and run as per the instructions.
I don't see how one follows from the other. Surely the same evolution and development that has given humans the power to create universes has also finally taught them to RTFM (one of which is enclosed in the capsule).
Turning the data on the discs back into the original characters should be extremely easy for the super-humans. Understanding it will be more interesting, just as understanding writings from our own ancient history is interesting and challenging, but ultimately largely achievable and well worth the effort.
It's hard to take any benchmark, or any study in general, seriously unless it sets out the methodology well enough that someone else could run the same tests himself. I use postgresql for several websites and I would love to get the kind of performance that was reported in the benchmark. Something along the lines of "our tests report that postgresql stomps on other databases, and here's how you can make your own postgresql installation do the same" would be so much more useful and credible than "postgresql is better, we've got a couple of numbers which prove it."