5 years ago is nothing. HP really went downhill recently. If you bought their stuff 10, 15 years ago, it probably still works, and you can still get carts for it. Their business class printers are still pretty good, but there are better choices in that space now as well.
The guy was specifically referring removing apps by deleting their directory under/usr/local, meaning he's trying to show how easy it is to remove apps in Linux without need for an uninstaller. It was HE (or she) who wasn't comparing apples to apples. I was just pointing out the absurdity of it, having just had to recompile a bunch of libraries I long forgotten about after a fresh system install because no RPMs were available.
This rarely works, unless you compile-install everything.
Most apps will install in/usr/local/bin/xxx and/usr/local/etc/xxx and/var/xxx and/usr/local/lib/xxx and possibly/include and god knows what else. Add to this any separately installed support libraries, rc.d init scripts, and you can have a much more difficult time 'uninstalling' than under windows, where the uninstaller takes care of all that (theoretically anyways).
It's not just help links (though what about when you need help and troubleshooting info and your app won't start?). What about Revert to Defaults, when settings get messed up beyond easy user repair? Games have that option often. Sure, it's just a command-line switch, but deserves its own Start Menu entry. There could be various post-install readmes, registration links, survey/merketing links, etc.
Also, I hate having icons in my root start menu, they should be in folders. Most icons are so butt ugly (and are GUARANTEED to clash with other icons) that it just doesn't look good. Instead, organize your start menu folders into groups, like Apps, Internet, Graphics, Video, Sound, P2P, etc. After this, you might as well break em out, and put them in the quick start area.
Yeah, yeah, whatever, I didn't say it was foolproof. You can also use a hash of a combination of client-provided headers that are very likely to remain constant during the session: agent string, screen resolution, etc. It's not meant to be foolproof. The scheme also breaks when people are behind large multi-ip proxies, or when http is proxied but https is not and your app mixes both.
if you're worried about someone snooping your session id (either in transit or from disk cache), then a cookie-based or url based schemes are equally at risk.
Also, have fun hijacking an md5() generated session id. And there's nothing stopping you from changing the session id on every request. Yes, I've done that once before, and it works pretty damn good, under some circumstances.
And there's no data stored on the client machine, just a session key. Again, see up two paragraphis if you're worried about this being stolen.
Something has to give when you're sticking to standard web protocols; there's only so much you can do. But if you run your session over https all the way, cookie-tied-to-ip scheme works just fine.
Well? The challenge still stands, Mr I'm-a-hot-web-developer. Tell us your secret method of keeping track of sessions that are better then cookies. Or do you have a patent on it?
Sorry, but you can't beat a cookie. All major dynamic web page schemes have easy cookie handling. ASP, PHP, Perl, etc. Most have built-in session capabilities. ASP and PHP have options for both cookie and url based sessions, and ASP will even do the autodetection for you. But URL rewriting will break when you have complex JavaScript generating URLs on the fly, or Shockwave menus, or Java applets. As long as it's the browser sending the request, the cookie is guaranteed to be sent.
You say major browsers have broken cookie support. Well, please, do tell us more, we're all waiting with baited breath. Just one example please. Personally I've never had a problem with cookies in all my years of web development. You set a cookie, you get it back on the next request. The reason people don't trust cookies, and turn them off completely, is because of a) very early security issues, and b) idiots like you spouting off bullshit.
If you're worried about cookies being hijacked, you have some very simple things you can do server side: - Tie session to IP. If you receive a session id that does not match the IP that set it up initially, either redirect to a login page, or ignore the request. - Time outs. If you get a session id, and last time you saw it was 30 minutes or whatever ago, time out the session and redirect to a login page.
These are just the extremely obvious ones, and I regularly use both in my web apps. There must be other methods, some more some less secure, out there.
How do you keep state between browser sessions? With cookie! That's what they were MADE for! This is as obvious as using a butter knife to spread mud on a slice of bread. Sure, the application may be somewhat novel, but the solution is obvious to anyone engaged in the field. How would you keep state between browser sessions? With cookies perhaps?
The reason IBM lost in the hardware space is that it became irrelevant what hardware the OS and software was running on. It became a commodity, and after that IBM didn't have the competitive advantage of being the makers of the best compatible hardware. They were what everyone had to be compatible with and once that happened, IBM had nothing but a brand.
The same thing will happen with software. We will get to a point where it doesn't matter what OS you are running, or even what applications; they will all work with open format data, doing open format actions and transforms on it. It's the promise of XML, a world of services which pass around chunks of data, and it won't matter who does the work, because they will all do it more or less equally as good. Sure MS and other large software makers will initially take the lead, but eventually the 'clone makers' will gain credibility, like they did in the hardware market.
May not happen soon, it may not be XML or services, but something will come along and wipe out the current software landscape.
OTOH, Windows could follow Apple's lead, and use Linux or BSD as a starting point for their next-generation OS. The problem with that idea is that it doesn't really match MS's current goals of DRM, software leases, and increasing MS's revenues.
Why would you say that? By taking over an existing, working OS, MS could concentrate on DRM, as opposed to putting out fires in their OS code base.
asp2php apparently did handle some db stuff, as far back as 2 years ago when I used it.
It didn't do the nicely wrapped stuff using Commerce Server components, which did nice db and lists and dictionaries/hashes, which was a pita, but simple enough to replicate in a php class. It also broke when you used one-line if-then statements, with no 'end'; these are legal in VB. I reported that bug after I noticed my resulting php code was getting very heavily indented. COM objects would be very difficult to handle... it's compiled code!
asp2php handled my ton of convoluted, structured and very nested (via includes) mess of an asp site. At the very least it will save you lots of time converting aspisms to phpisms.
No it's not art. You don't write code to create art, but to create functional applications. If at the end of the day some of your code can be positively appreciated by a very select few, bonus. It's, say, like an appliance, or a TV. It's designed for a certain task, and may look pretty, should even, belong in a museum of modern art, but sorry, it ain't art.
I agree with the view of programming as a craft. It'll probably never be an exact science, not until computers can program themselves (but then you still have the underlying code). It will probably never be like engineering either. Sure, the processes and approaches and stuff at the project level will be more structured and defined, but ultimately the individual programmer will at best follow a function spec, beat out some code, and return what amounts to a (hopefully) well documented black box. Code that works around library and system bugs, is optimized in certain places, and follows the programmers thought pattern. It's problem solving, even though the problem has already been solved, at a higher level though.
I would argue programming is much more like a craft than anything else.
He made it quite clear he was downloading songs from old albums he already owned.
Owned, in a different format. If I have a VHS of some movie, am I entitiled to a free DVD of same? You'd probably say I'd be entitled to a DivX of it? It's no different. Just because it's an mp3, which we're accustomed to getting for free doesn't mean they should be free.
If the guy's record player broke, and record players were no longer available, tough break. If no one makes parts for my car anymore, I'm not entitled to driving a new one off the car lot for free. You're free to try to reverse engineer the technology, and build your own replacement.
You want an mp3 of something you own? Make it yourself. Can't? Find a way, or though break. Or pay for it. Just because you reportedly already own the track doesn't allow you to download it. More importantly, the guy you're downloading it from probably didn't pay any licence fees to engage in the activity.
You go to a club, and the DJ plays a track that you already own, have on your person, in your portable CD player, even if you're the only person there, the club still has to pay for the right to play that track, and have you listen to it.
Different format, different context. It's not an interchangable substitute good. They're different products, and you pay for each.
The question at hand is, when you buy music, are you purchasing a physical CD or the rights to intellectual property?
They're backpedaling furiously and claiming you're licencing the music, purchasing the right to listen to it. It's a specious claim, as there's not even so much as a token click through on a CD case.
anyway, thanks for reading and responding to my post.
I've been telling that to Honda for the past two years. I already paid for my 1970 Civic, it's obsolete now, and all scratched up, and doesn't even run half the time. Why should I have to pay for a 2003 version all over again? It's the same stuff, just an engine on wheels.
The main reason is that producing a quality special edition DVD would take away the brothers' time away from the sequels and the ultimate (as in 'inevitable') trilogy special edition release. From that point of view, not having a Matrix SE makes perfect sense.
Also, now that the last two movies are almost out, most people know a 3-movie single release edition is just around the corner, and will wait a while and spend their money on THAT. In the mean-time, the cheaper regular edition will do to catch them up on the plot.
Why would I pay $10 per month to watch the Sopranos any time I want when $14 per month to Tivo allows me to record any show I want to watch and view it at any time?
That's only because the current economic model of cable TV allows you (and the relatively small band of other PVR owners) piggyback on the service; ie. you're being subsidized by regular subscribers. Once the numbers of PVR owners grow large enough for advertisers to really notice (which may already be happening, hence MystroTV), TiVo will need to get with the times and pay royalties or fees to content producers, likle networks do now, and these will be of course passed on to the subscribers.
It's only a matter of time, and your days of nearlt free personalized TV watching are very numbered.
Well, I would say actors 'back then' were as good as they are now. Sure, they look much better today, what with editing and effects and polished scripts. But, while they were acting more several decades ago, the overall effect was much weaker. Watch some of the old time movies, and you can't help but cringe, not at the cheesiness or the set, but the acting style. After a 30 second long take, actual acting starts to show through, and frankly, most actors aren't that good.
It's the long takes that really take good actors, cameramen and directors. The close-to-opening scene in Pulp Fiction for example (where Jules and Vincent walk to the appartment) and a stupid long shot, but you don't really notice it, it's so good. Lesser people would absolutely ruin it.
Heh, that's SO true. The various copy protection and activation schemes have long ago all but eliminated the so called casual copying. Tightening the screwes any more is stepping on legitimate customer's toes, and you'll never stop the warez groups from cracking commodity software protection.
How do you define a pirate? What if a downloader is simply streaming a batch of files to the speakers, emulating a radio? Would you call that a pirate?
If the streamer pays the required fees and acquires licenses as needed, no, I wouldn't call that a pirate. The downloader has nothing to do with your example, it is the sender's responsibility to comply with the law. That's any all the major busts were WRT distributing, not downloading. Downloading is still somewhat a grey area.
No, patents DO cover implementations. You can't patent an idea, you have to build a device that implements all the points/claims in your patent.
The easiest way to get around a patent is to do the same thing, but slightly differently. Then patent THAT, so no one else can use the same hole to compete with you.
Slashdot Mirror
5 years ago is nothing. HP really went downhill recently. If you bought their stuff 10, 15 years ago, it probably still works, and you can still get carts for it. Their business class printers are still pretty good, but there are better choices in that space now as well.
The guy was specifically referring removing apps by deleting their directory under /usr/local, meaning he's trying to show how easy it is to remove apps in Linux without need for an uninstaller. It was HE (or she) who wasn't comparing apples to apples. I was just pointing out the absurdity of it, having just had to recompile a bunch of libraries I long forgotten about after a fresh system install because no RPMs were available.
This rarely works, unless you compile-install everything.
/usr/local/bin/xxx and /usr/local/etc/xxx and /var/xxx and /usr/local/lib/xxx and possibly /include and god knows what else. Add to this any separately installed support libraries, rc.d init scripts, and you can have a much more difficult time 'uninstalling' than under windows, where the uninstaller takes care of all that (theoretically anyways).
Most apps will install in
It's not just help links (though what about when you need help and troubleshooting info and your app won't start?). What about Revert to Defaults, when settings get messed up beyond easy user repair? Games have that option often. Sure, it's just a command-line switch, but deserves its own Start Menu entry. There could be various post-install readmes, registration links, survey/merketing links, etc.
Also, I hate having icons in my root start menu, they should be in folders. Most icons are so butt ugly (and are GUARANTEED to clash with other icons) that it just doesn't look good. Instead, organize your start menu folders into groups, like Apps, Internet, Graphics, Video, Sound, P2P, etc. After this, you might as well break em out, and put them in the quick start area.
Yeah, yeah, whatever, I didn't say it was foolproof. You can also use a hash of a combination of client-provided headers that are very likely to remain constant during the session: agent string, screen resolution, etc. It's not meant to be foolproof. The scheme also breaks when people are behind large multi-ip proxies, or when http is proxied but https is not and your app mixes both.
if you're worried about someone snooping your session id (either in transit or from disk cache), then a cookie-based or url based schemes are equally at risk.
Also, have fun hijacking an md5() generated session id. And there's nothing stopping you from changing the session id on every request. Yes, I've done that once before, and it works pretty damn good, under some circumstances.
And there's no data stored on the client machine, just a session key. Again, see up two paragraphis if you're worried about this being stolen.
Something has to give when you're sticking to standard web protocols; there's only so much you can do. But if you run your session over https all the way, cookie-tied-to-ip scheme works just fine.
Well? The challenge still stands, Mr I'm-a-hot-web-developer. Tell us your secret method of keeping track of sessions that are better then cookies. Or do you have a patent on it?
Sorry, but you can't beat a cookie. All major dynamic web page schemes have easy cookie handling. ASP, PHP, Perl, etc. Most have built-in session capabilities. ASP and PHP have options for both cookie and url based sessions, and ASP will even do the autodetection for you. But URL rewriting will break when you have complex JavaScript generating URLs on the fly, or Shockwave menus, or Java applets. As long as it's the browser sending the request, the cookie is guaranteed to be sent.
You say major browsers have broken cookie support. Well, please, do tell us more, we're all waiting with baited breath. Just one example please. Personally I've never had a problem with cookies in all my years of web development. You set a cookie, you get it back on the next request. The reason people don't trust cookies, and turn them off completely, is because of a) very early security issues, and b) idiots like you spouting off bullshit.
If you're worried about cookies being hijacked, you have some very simple things you can do server side:
- Tie session to IP. If you receive a session id that does not match the IP that set it up initially, either redirect to a login page, or ignore the request.
- Time outs. If you get a session id, and last time you saw it was 30 minutes or whatever ago, time out the session and redirect to a login page.
These are just the extremely obvious ones, and I regularly use both in my web apps. There must be other methods, some more some less secure, out there.
No! It IS obvious!
How do you keep state between browser sessions? With cookie! That's what they were MADE for! This is as obvious as using a butter knife to spread mud on a slice of bread. Sure, the application may be somewhat novel, but the solution is obvious to anyone engaged in the field. How would you keep state between browser sessions? With cookies perhaps?
The stock price isn't because of hardware, even PPC... what's your point anyways?
The reason IBM lost in the hardware space is that it became irrelevant what hardware the OS and software was running on. It became a commodity, and after that IBM didn't have the competitive advantage of being the makers of the best compatible hardware. They were what everyone had to be compatible with and once that happened, IBM had nothing but a brand.
The same thing will happen with software. We will get to a point where it doesn't matter what OS you are running, or even what applications; they will all work with open format data, doing open format actions and transforms on it. It's the promise of XML, a world of services which pass around chunks of data, and it won't matter who does the work, because they will all do it more or less equally as good. Sure MS and other large software makers will initially take the lead, but eventually the 'clone makers' will gain credibility, like they did in the hardware market.
May not happen soon, it may not be XML or services, but something will come along and wipe out the current software landscape.
Why would you say that? By taking over an existing, working OS, MS could concentrate on DRM, as opposed to putting out fires in their OS code base.
Filesharing is not a violent offence... last time I did it anyways.
With DRM you NEVER hold the keys; the content producers/distributors do. Otherwise, uhm, what's the point?
asp2php apparently did handle some db stuff, as far back as 2 years ago when I used it.
It didn't do the nicely wrapped stuff using Commerce Server components, which did nice db and lists and dictionaries/hashes, which was a pita, but simple enough to replicate in a php class. It also broke when you used one-line if-then statements, with no 'end'; these are legal in VB. I reported that bug after I noticed my resulting php code was getting very heavily indented. COM objects would be very difficult to handle... it's compiled code!
asp2php handled my ton of convoluted, structured and very nested (via includes) mess of an asp site. At the very least it will save you lots of time converting aspisms to phpisms.
No it's not art. You don't write code to create art, but to create functional applications. If at the end of the day some of your code can be positively appreciated by a very select few, bonus. It's, say, like an appliance, or a TV. It's designed for a certain task, and may look pretty, should even, belong in a museum of modern art, but sorry, it ain't art.
I agree with the view of programming as a craft. It'll probably never be an exact science, not until computers can program themselves (but then you still have the underlying code). It will probably never be like engineering either. Sure, the processes and approaches and stuff at the project level will be more structured and defined, but ultimately the individual programmer will at best follow a function spec, beat out some code, and return what amounts to a (hopefully) well documented black box. Code that works around library and system bugs, is optimized in certain places, and follows the programmers thought pattern. It's problem solving, even though the problem has already been solved, at a higher level though.
I would argue programming is much more like a craft than anything else.
Owned, in a different format. If I have a VHS of some movie, am I entitiled to a free DVD of same? You'd probably say I'd be entitled to a DivX of it? It's no different. Just because it's an mp3, which we're accustomed to getting for free doesn't mean they should be free.
If the guy's record player broke, and record players were no longer available, tough break. If no one makes parts for my car anymore, I'm not entitled to driving a new one off the car lot for free. You're free to try to reverse engineer the technology, and build your own replacement.
You want an mp3 of something you own? Make it yourself. Can't? Find a way, or though break. Or pay for it. Just because you reportedly already own the track doesn't allow you to download it. More importantly, the guy you're downloading it from probably didn't pay any licence fees to engage in the activity.
You go to a club, and the DJ plays a track that you already own, have on your person, in your portable CD player, even if you're the only person there, the club still has to pay for the right to play that track, and have you listen to it.
Different format, different context. It's not an interchangable substitute good. They're different products, and you pay for each.
They're backpedaling furiously and claiming you're licencing the music, purchasing the right to listen to it. It's a specious claim, as there's not even so much as a token click through on a CD case.
anyway, thanks for reading and responding to my post.
Don't do that, I hate it when people do that...
I've been telling that to Honda for the past two years. I already paid for my 1970 Civic, it's obsolete now, and all scratched up, and doesn't even run half the time. Why should I have to pay for a 2003 version all over again? It's the same stuff, just an engine on wheels.
The main reason is that producing a quality special edition DVD would take away the brothers' time away from the sequels and the ultimate (as in 'inevitable') trilogy special edition release. From that point of view, not having a Matrix SE makes perfect sense.
Also, now that the last two movies are almost out, most people know a 3-movie single release edition is just around the corner, and will wait a while and spend their money on THAT. In the mean-time, the cheaper regular edition will do to catch them up on the plot.
That's only because the current economic model of cable TV allows you (and the relatively small band of other PVR owners) piggyback on the service; ie. you're being subsidized by regular subscribers. Once the numbers of PVR owners grow large enough for advertisers to really notice (which may already be happening, hence MystroTV), TiVo will need to get with the times and pay royalties or fees to content producers, likle networks do now, and these will be of course passed on to the subscribers.
It's only a matter of time, and your days of nearlt free personalized TV watching are very numbered.
Well, I would say actors 'back then' were as good as they are now. Sure, they look much better today, what with editing and effects and polished scripts. But, while they were acting more several decades ago, the overall effect was much weaker. Watch some of the old time movies, and you can't help but cringe, not at the cheesiness or the set, but the acting style. After a 30 second long take, actual acting starts to show through, and frankly, most actors aren't that good.
It's the long takes that really take good actors, cameramen and directors. The close-to-opening scene in Pulp Fiction for example (where Jules and Vincent walk to the appartment) and a stupid long shot, but you don't really notice it, it's so good. Lesser people would absolutely ruin it.
Heh, that's SO true. The various copy protection and activation schemes have long ago all but eliminated the so called casual copying. Tightening the screwes any more is stepping on legitimate customer's toes, and you'll never stop the warez groups from cracking commodity software protection.
If the streamer pays the required fees and acquires licenses as needed, no, I wouldn't call that a pirate. The downloader has nothing to do with your example, it is the sender's responsibility to comply with the law. That's any all the major busts were WRT distributing, not downloading. Downloading is still somewhat a grey area.
No, patents DO cover implementations. You can't patent an idea, you have to build a device that implements all the points/claims in your patent.
The easiest way to get around a patent is to do the same thing, but slightly differently. Then patent THAT, so no one else can use the same hole to compete with you.
Nah, classic rock stations are turning formulaic as well these days. They have their playlists down to a science.
This service is only for their broadband customers.