Here's the greater point, why do people even go INTO channels if they're not going to chat? There were 50+ people in the channel I was in, and only one of them typed *anything* in 5 entire minutes.
Welcome, you must be new here!
Seriously, IRC is not IM. A lot of people are in multiple channels or are merely idling while they are actually doing useful stuff. You can't jump into an IRC channel and expect support on-the-spot. IRC doesn't work that way. Join, lurk a bit, if you notice some activity launch a question and don't expect an answer immediately.
I use IRC as a secondary support method (next to a mailinglist) for a project with a small following. The people who get IRC are relaxed and polite, even if they have to wait half an hour for an answer and I go out of my way to help them out. The people who don't get IRC frequently leave the channel just seconds before I help them out. C'est la vie.
But then Google wouldn't be able to show you all their Adwords on those websites (and the ones they link to).
Why do you think Google isn't doing anything against link-farming? Because they merely have to act ignorant and rake in the cash. Vote with your feet and use a different search engine (or meta-searchengine like clusty), diversity is good.
I know a number of businesses and private people who use Open Office every day exchanging documents with others without a hitch, whereas I have never heard of anyone who gave it up because it was huge, buggy, or had difficulty using other formats.
Funny thing is, last week I paid a visit to a client. They have Firefox and OOo installed on every machine and have so for a number of years effectively, yet one of the directors still mentioned being glad to use MS Office back home.
I couldn't blame with him. Being the unix hippie I am, personally I prefer LaTeX over any word processor. When I have to I use OOo, but I never enjoy the experience (I'm using Abiword more and more lately). The occasions that I am behind an XP machine MS Office works like a charm.
Lets be honest: OOo is a slow bloated monster that really isn't compatible with MS's proprietary formats. OOo is improving, but slowly. Only when the open source product is better than the proprietary product (ie. Firefox) will you get people to switch, and OOo is not yet better than Office. Thinking so is fooling yourself.
This is exactly identical to you SSHing into my console server, running a terminal app, and me handing back its output.
Similar, but not exactly. In this case if the terminal app would be licensed under the GPL I would have the right, as a user, to request a copy of the source code of that terminal app.
The identical case would be for me to SSH into your console server, me giving some input file to you while you actually run the application and hand me a copy of the application's output.
I understand what the FSF is trying to do, however it is extremely impractical and has a number of loopholes in my point of view. What if the communication was over a local interface (socket or other IPC)? What if I use virtualized OS instances and communicate between these? What if I supply the webservice but only handle requests via pigeon-mail? None of these are done over a computer network, thus I wouldn't have to distribute the service's source code.
From a practical perspective, how would I know a remote web service is using my AGPL'ed source code? You can only compare the output of the service, while with binaries you can at least investigate the machine code or use a decompiler, which gives you much more certainty.
Again, I understand what the FSF is trying to do with this license. But in my book output == output, code == code, and restricting one using a license on the other leads to a very healthy paycheck for a lawyer.
The danger with C is not that you might forget to free some object, but rather, that you do free some object and forget to remove a reference to it.
Euh, the danger with C is _both_ that you might forget to free an object _and_ that you are able to use free'ed references.
The former leads to memory leaks, the latter to a lot more nastier problems. But both are dangers if you don't code defensively, it is only that memory leaks take longer to become a problem.
But OLPC is STUPID to compete in America with the low-end power of Wal-Mart.
Let's see, OLPC: One _Laptop_ Per Child? This article is about a PC. Without a monitor. Designed to offer Americans an alternative to PC's with the Microsoft Tax. OLPC is a laptop designed for kids in third world countries.
Maybe you'd like to rant against Oranges for not being an Apple next?
In fact, some professors *still* use overhead projectors.
About a third of the CS professors and all of the mathematics profs at our faculty still do. Works fine.
Just yesterday I held my first presentation in years _without slides_, not even a projector. Damn I was nervous, but after you get over it you can concentrate on telling your story to the audience instead of everyone being distracted by a projection. It was probably my best presentation this year.
Use the best tool for the job. That includes no tool at all.
Well, over here in the EU software patents still aren't legal. If "invitation-only" access were given to any interested open source software developers, I guess there wouldn't be any legal hurdles to take. Depending on which entity (not-for-profit foundation) would have access to the specs, this would be a perfectly legal means of having access to these specs.
They do mention open source developers explicitly, so yes, I do think the point is to make us happy. If it actually would be worth the effort, that's naturally a completely different story.
So, how about the EU uses a very small portion of those fines (say, a nominal one-off payment of E 10000) to obtain these documents once and publish them for all us open source developers via a new sourceforge.net project?
I'm all for interoperability, but it's not like I'm going to pay $10000 for half-undocumented Microsoft protocols. At least a small portion of all those millions would be put to good use, instead of it all disappearing into the black hole that is the EU budget.
Typically, new releases of software tend to have more features - these added features are what cause the bigger-slower-bloatier effect, as the take space to store, time to execute, and not everyone wants them. I don't know of one piece of software that has managed to avoid this fate.
How about Netscape -> Mozilla -> Phoenix?
I'll give you that Mozilla Firefox currently is comparable in bloat (and features, in many cases) to Netscape, however Phoenix and the earlier Firefox releases were a dream compared to the earlier Gecko-based browsers. Who needs a feed-reader, I just want to browse the web ma'am!
In general software does tend to get bigger and slower over new releases, but when said software goes open source anyone can evolve it further in the direction they want. Openoffice.org is a good counter-example though, but I'm hoping IBM will pick up the slack.
If you can define a state space and a fitness function, you're almost certainly better off using (non-)linear programming, constraint programming, or a local search method like tabu search.
Linear programming and hill-climbing algorithms have their place when it comes to simple search spaces, however when you have many dimensions and a vast non-linear search space none of the algorithms you named will get very far.
In the past I had similar discussions with a prof on search spaces with dozens of dimensions. His answer to GAs was to reset the local search/hill-climbing algorithm after a lack of increase in the found solution. A quick calculation showed that on average the amount of trials required would be much more inefficient compared to the amount of trials a GA, or evolutionary algorithm in general, would require.
But as with all search problems, your search space should define what kind of algorithm you use, not the other way around.
Linux, Apache and all the server-side scripting languages normally aren't the problem. Many hosts I have audited have old installations of (mostly) PHP-based software, and these automated attacks tend to target them leading to (sometimes multiple) botnet infections.
Many administrators didn't even know what was running on their servers. It only takes a couple of minutes to install packages like *coughthesecurityholecalled* phpBB, however if you are doing this independently from your package management system you will lose track of the installs. Even worse, the installs won't be automatically upgraded, which is a major reason for sticking with stock Debian/RHEL/SuSE package repositories.
If you choose to install software outside your distribution's package management system, subscribe to the announcement-lists of the software used. Document on which servers you installed what software. And if you leave the company, make sure your replacement can hop right in and will know what you know.
Common sense, but far too often forgotten or ignored.
according to insiders, it's actually a problem with blu-ray's java, and the players that are having problems just need a firmware update.
The fact that these players need an update at all is enough cause for outrage. How often do you have to update the firmware on your DVD-player?
People might accept this kind of crap on their PC or console, you're connected to the Internet anyway, however consumer electronics should "Just Work(tm)" without a connection. Most non-geeks I know would initially blame such problems on the BR-disk itself, but if they were to experience these types of problems on a regular basis they would sooner chuck out the BR-player than upgrade their firmware.
What would John Doe choose: a movie that works in a 20-buck player or his PC, or a slightly-prettier movie that costs more, needs a 300-buck player and has the risk of becoming useless for no apparent reason? Sure, a technophile with cash to burn might choose the latter, but until these problems are sorted out producers seriously run the risk of alienating their early-adopters.
Apple itself used open source with OS X (kernel, web browser) mainly because they knew it would irritate Microsoft.
I was under the impression that Apple wanted to dump their aging code base and get a tried-and-proven *nix kernel + HTML/JS engine for free.
Flame me all you want, but I haven't noticed a lot of open-source love (or user-love in general) from Apple, and I'm sure they didn't use Darwin because they wanted to annoy Microsoft. If they wanted to annoy Microsoft, they would have joined the Linux/OpenOffice/Firefox-camp.
No matter how Apple fanboys twist reality, bricking a phone is yet another way how Apple rapes their user base. It goes to show that no matter how you abuse your customers, great PR fixes everything.
This whole issue wouldn't have existed without thorough research and lots of persistence of the group at "We Don't Trust Voting Computers". These men and women have dived into the voting computers used for decades in the Netherlands, found numerous serious flaws and made them public. They forced our government to install this commission, which has lead to the best possible outcome: no more electronic voting.
What kind of country are we living in, where wearing the wrong thing to the airport has a decent chance of getting us killed?
Considering the killing of a Brazilian commuter in the London subway two years ago by the police because they thought he was a terrorist, it's clear that this has little to do with which country you are in. Whatever the terrorists were trying to succeed, they did succeed in getting authorities paranoid.
Having said that, wearing a fake-bomb in an airport (no matter how close to the real thing it is) wouldn't have been a very smart move prior to 9/11 either.
Suppose you start with 1000 lines of GPL code and build something new and add 100,000 lines of your won. You are working for the original author and you have no freedom or choice about it, independent of your contribution.
Not true. If you choose to use GPL'ed code, then the resulting product would indeed be covered by the GPL. However the original author can't relicense the resulting product any more than you can. It will remain under the GPL as long as both authors don't agree to relicense the product under a single new license.
And you always have the choice of replacing the GPL'ed code with your own, or with code licensed differently. If you add 100,000 lines, an extra 1000 isn't that much of a deal, is it?
Ignoring the confusion between PHP5 and the web frameworks named, in your situation I wouldn't go with any of them. If all you want is simply an easily-maintainable website then MODx is a much better choice. It's flexible enough to wrap around most websites without writing a single line of code, most of the work goes into the actual HTML-templates. It's more of a CMS/CMF-combination, with a large amount of plugins and an active community. The tree-based folder/document structures, publication-features, multiple users for different website sections, meta-tag editing... my users love the built-in functionalities and I don't see the point in implementing them all myself (anymore).
Don't get me wrong, I'm a Django-fan and have used it and other web frameworks for quite a few commercial projects (specialized web applications, mostly intranet stuff). Django/RoR/Cake are great for such projects, however for maintainers of websites a fully-featured CMS (like MODx has) knocks the socks off of even the auto-generated backends. If your website falls into the multiple-level/multiple-document structure then it is pointless IMHO to build your own CMS from scratch.
And yes, MODx is buzzword-compliant (AJAX/XHTML/Web2.0/SEO), so it's an easy sell to management.
The obvious next step here, and one I've been warning our user base (some of whom are up front about their ad blocking policies, despite the fact they're more than happy to accept our free services) about for quite some time is a new generation of ad code from advertisers that runs directly from the host webserver - making it more difficult to block.
You mean stuff like openads? Packages like these have been blocked for quite some time through the default Adblock Plus subscriptions. It's a no-brainer.
I'm with the 80%: I have the right to block what I want and allow what I want. I'm not 'taking your content' (the same excuse that the RIAA makes for declining CD sales), I have the right to use whatever is connected to the internet. I'm sick of all the sites that claim bandwidth-costs as an excuse to stuff their pages full with ads and rake in $k's.
Make your site good enough, and we'll whitelist you. If you try to actively block us, we will leave. Lots of good that will do, won't it?
Slashdot Mirror
Welcome, you must be new here!
Seriously, IRC is not IM. A lot of people are in multiple channels or are merely idling while they are actually doing useful stuff. You can't jump into an IRC channel and expect support on-the-spot. IRC doesn't work that way. Join, lurk a bit, if you notice some activity launch a question and don't expect an answer immediately.
I use IRC as a secondary support method (next to a mailinglist) for a project with a small following. The people who get IRC are relaxed and polite, even if they have to wait half an hour for an answer and I go out of my way to help them out. The people who don't get IRC frequently leave the channel just seconds before I help them out. C'est la vie.
But then Google wouldn't be able to show you all their Adwords on those websites (and the ones they link to).
Why do you think Google isn't doing anything against link-farming? Because they merely have to act ignorant and rake in the cash. Vote with your feet and use a different search engine (or meta-searchengine like clusty), diversity is good.
Linking a 40MB quicktime movie from Slashdot somehow doesn't seem like a very smart idea...
For the wasp this looks like a very useful move. Why haul your food when it can walk for you?
Does anyone have a .torrent for the iso?
What would you expect from a book called "Head First SQL"?
And a minor nitpick: a seasoned DBA should know it is MySQL, not mySQL. That will be all.
Funny thing is, last week I paid a visit to a client. They have Firefox and OOo installed on every machine and have so for a number of years effectively, yet one of the directors still mentioned being glad to use MS Office back home.
I couldn't blame with him. Being the unix hippie I am, personally I prefer LaTeX over any word processor. When I have to I use OOo, but I never enjoy the experience (I'm using Abiword more and more lately). The occasions that I am behind an XP machine MS Office works like a charm.
Lets be honest: OOo is a slow bloated monster that really isn't compatible with MS's proprietary formats. OOo is improving, but slowly. Only when the open source product is better than the proprietary product (ie. Firefox) will you get people to switch, and OOo is not yet better than Office. Thinking so is fooling yourself.
Similar, but not exactly. In this case if the terminal app would be licensed under the GPL I would have the right, as a user, to request a copy of the source code of that terminal app.
The identical case would be for me to SSH into your console server, me giving some input file to you while you actually run the application and hand me a copy of the application's output.
I understand what the FSF is trying to do, however it is extremely impractical and has a number of loopholes in my point of view. What if the communication was over a local interface (socket or other IPC)? What if I use virtualized OS instances and communicate between these? What if I supply the webservice but only handle requests via pigeon-mail? None of these are done over a computer network, thus I wouldn't have to distribute the service's source code.
From a practical perspective, how would I know a remote web service is using my AGPL'ed source code? You can only compare the output of the service, while with binaries you can at least investigate the machine code or use a decompiler, which gives you much more certainty.
Again, I understand what the FSF is trying to do with this license. But in my book output == output, code == code, and restricting one using a license on the other leads to a very healthy paycheck for a lawyer.
Euh, the danger with C is _both_ that you might forget to free an object _and_ that you are able to use free'ed references.
The former leads to memory leaks, the latter to a lot more nastier problems. But both are dangers if you don't code defensively, it is only that memory leaks take longer to become a problem.
Let's see, OLPC: One _Laptop_ Per Child? This article is about a PC. Without a monitor. Designed to offer Americans an alternative to PC's with the Microsoft Tax. OLPC is a laptop designed for kids in third world countries.
Maybe you'd like to rant against Oranges for not being an Apple next?
About a third of the CS professors and all of the mathematics profs at our faculty still do. Works fine.
Just yesterday I held my first presentation in years _without slides_, not even a projector. Damn I was nervous, but after you get over it you can concentrate on telling your story to the audience instead of everyone being distracted by a projection. It was probably my best presentation this year.
Use the best tool for the job. That includes no tool at all.
Just wanted to say kudos to Mako for this book and his work for both Debian and Ubuntu. You're doing a fine job!
Well, over here in the EU software patents still aren't legal. If "invitation-only" access were given to any interested open source software developers, I guess there wouldn't be any legal hurdles to take. Depending on which entity (not-for-profit foundation) would have access to the specs, this would be a perfectly legal means of having access to these specs.
They do mention open source developers explicitly, so yes, I do think the point is to make us happy. If it actually would be worth the effort, that's naturally a completely different story.
So, how about the EU uses a very small portion of those fines (say, a nominal one-off payment of E 10000) to obtain these documents once and publish them for all us open source developers via a new sourceforge.net project?
I'm all for interoperability, but it's not like I'm going to pay $10000 for half-undocumented Microsoft protocols. At least a small portion of all those millions would be put to good use, instead of it all disappearing into the black hole that is the EU budget.
How about Netscape -> Mozilla -> Phoenix?
I'll give you that Mozilla Firefox currently is comparable in bloat (and features, in many cases) to Netscape, however Phoenix and the earlier Firefox releases were a dream compared to the earlier Gecko-based browsers. Who needs a feed-reader, I just want to browse the web ma'am!
In general software does tend to get bigger and slower over new releases, but when said software goes open source anyone can evolve it further in the direction they want. Openoffice.org is a good counter-example though, but I'm hoping IBM will pick up the slack.
Linear programming and hill-climbing algorithms have their place when it comes to simple search spaces, however when you have many dimensions and a vast non-linear search space none of the algorithms you named will get very far.
In the past I had similar discussions with a prof on search spaces with dozens of dimensions. His answer to GAs was to reset the local search/hill-climbing algorithm after a lack of increase in the found solution. A quick calculation showed that on average the amount of trials required would be much more inefficient compared to the amount of trials a GA, or evolutionary algorithm in general, would require.
But as with all search problems, your search space should define what kind of algorithm you use, not the other way around.
I have noticed this as well.
Linux, Apache and all the server-side scripting languages normally aren't the problem. Many hosts I have audited have old installations of (mostly) PHP-based software, and these automated attacks tend to target them leading to (sometimes multiple) botnet infections.
Many administrators didn't even know what was running on their servers. It only takes a couple of minutes to install packages like *coughthesecurityholecalled* phpBB, however if you are doing this independently from your package management system you will lose track of the installs. Even worse, the installs won't be automatically upgraded, which is a major reason for sticking with stock Debian/RHEL/SuSE package repositories.
If you choose to install software outside your distribution's package management system, subscribe to the announcement-lists of the software used. Document on which servers you installed what software. And if you leave the company, make sure your replacement can hop right in and will know what you know.
Common sense, but far too often forgotten or ignored.
The fact that these players need an update at all is enough cause for outrage. How often do you have to update the firmware on your DVD-player?
People might accept this kind of crap on their PC or console, you're connected to the Internet anyway, however consumer electronics should "Just Work(tm)" without a connection. Most non-geeks I know would initially blame such problems on the BR-disk itself, but if they were to experience these types of problems on a regular basis they would sooner chuck out the BR-player than upgrade their firmware.
What would John Doe choose: a movie that works in a 20-buck player or his PC, or a slightly-prettier movie that costs more, needs a 300-buck player and has the risk of becoming useless for no apparent reason? Sure, a technophile with cash to burn might choose the latter, but until these problems are sorted out producers seriously run the risk of alienating their early-adopters.
I was under the impression that Apple wanted to dump their aging code base and get a tried-and-proven *nix kernel + HTML/JS engine for free.
Flame me all you want, but I haven't noticed a lot of open-source love (or user-love in general) from Apple, and I'm sure they didn't use Darwin because they wanted to annoy Microsoft. If they wanted to annoy Microsoft, they would have joined the Linux/OpenOffice/Firefox-camp.
No matter how Apple fanboys twist reality, bricking a phone is yet another way how Apple rapes their user base. It goes to show that no matter how you abuse your customers, great PR fixes everything.
This whole issue wouldn't have existed without thorough research and lots of persistence of the group at "We Don't Trust Voting Computers". These men and women have dived into the voting computers used for decades in the Netherlands, found numerous serious flaws and made them public. They forced our government to install this commission, which has lead to the best possible outcome: no more electronic voting.
Thanks guys, you rock.
Considering the killing of a Brazilian commuter in the London subway two years ago by the police because they thought he was a terrorist, it's clear that this has little to do with which country you are in. Whatever the terrorists were trying to succeed, they did succeed in getting authorities paranoid.
Having said that, wearing a fake-bomb in an airport (no matter how close to the real thing it is) wouldn't have been a very smart move prior to 9/11 either.
Not true. If you choose to use GPL'ed code, then the resulting product would indeed be covered by the GPL. However the original author can't relicense the resulting product any more than you can. It will remain under the GPL as long as both authors don't agree to relicense the product under a single new license.
And you always have the choice of replacing the GPL'ed code with your own, or with code licensed differently. If you add 100,000 lines, an extra 1000 isn't that much of a deal, is it?
everything looks like a nail.
Ignoring the confusion between PHP5 and the web frameworks named, in your situation I wouldn't go with any of them. If all you want is simply an easily-maintainable website then MODx is a much better choice. It's flexible enough to wrap around most websites without writing a single line of code, most of the work goes into the actual HTML-templates. It's more of a CMS/CMF-combination, with a large amount of plugins and an active community. The tree-based folder/document structures, publication-features, multiple users for different website sections, meta-tag editing... my users love the built-in functionalities and I don't see the point in implementing them all myself (anymore).
Don't get me wrong, I'm a Django-fan and have used it and other web frameworks for quite a few commercial projects (specialized web applications, mostly intranet stuff). Django/RoR/Cake are great for such projects, however for maintainers of websites a fully-featured CMS (like MODx has) knocks the socks off of even the auto-generated backends. If your website falls into the multiple-level/multiple-document structure then it is pointless IMHO to build your own CMS from scratch.
And yes, MODx is buzzword-compliant (AJAX/XHTML/Web2.0/SEO), so it's an easy sell to management.
1. Go short on company stock
2. Go public on fact that company is lawyer-bait
3. Profit!
Who said you couldn't make money with the GPL?
You mean stuff like openads? Packages like these have been blocked for quite some time through the default Adblock Plus subscriptions. It's a no-brainer.
I'm with the 80%: I have the right to block what I want and allow what I want. I'm not 'taking your content' (the same excuse that the RIAA makes for declining CD sales), I have the right to use whatever is connected to the internet. I'm sick of all the sites that claim bandwidth-costs as an excuse to stuff their pages full with ads and rake in $k's.
Make your site good enough, and we'll whitelist you. If you try to actively block us, we will leave. Lots of good that will do, won't it?
They will pry my precious bodily fluids out of my cold dead... oh wait, never mind.