Trying to claim neither side is good or evil just waters down the game and makes it more bland and boring.
I couldn't disagree more. What is boring are simplistic labels like Good and Evil. Good is never good as seems on the surface and Evil always has its side to the story. I guess there is still room for the superficial battles between Good and Evil, but truely rich, engaging, insightful stories have a depth that is beyond good and evil. Sometimes it is more like "light" and "dark," but I think that is more a matter of personalities and not necessarily based on some religious sense of absolute good and evil. For example, someone might like morbid music and maybe prefer to spend their waking hours at nighttime. This person might have a dark personality. But it doesn't make them evil. Evil is as evil does, and Good is just as likely to do evil... in the real world, at least.
Well, that is the other extreme argument, I suppose. I was merely countering the idea that Linux (running on IBM hardware, I presume) has no paid support and that your only recourse to problems is posting a bug report on "Freshmeat." It just isn't true. No, paid support isn't perfect. But it is another tool in the chest. When all the normal channels fail, it is sometimes nice to have a paid, professional engineer on your problem. And if you're Quantas, you'll probably get more than one regardless of whether you're running Linux or AIX.
The point is that the GP came the conclusion that the system design was the problem with absolutely no information. I certainly didn't read anything in TFA that suggested they were trying to do anything unusual or untested with Linux. Of course, I don't know that they weren't either.
Anyway, If this were Windows we were talking about, there would be no question that it was the OS.
As customer's data and number of clients grow they will start experiencing scalability problems.
How many airlines these days are actually growing?
Propose much more scalable, reliable, dependable (and much more expensive) solution on AIX, AS/400, Mainframe.
What makes you think they weren't running LInux on a mainframe? IBM has has Linux running under a hypervisor on their big iron machines for some time now.
Posting a bug report at freshmeat doesn't quite cut it when you have planes grounded...
Show your ignorance much?
AFAIK, IBM fully supports running Linux on their hardware. I'm sure whatever contract Quantas has with IBM covers Linux so they don't have to be posting bug reports on "Freshmeat" (still can't believe you said that) when the system goes down. That said, even Linux support from IBM may not as good as running AIX which (presumably) wouldn't have had the given problems in the first place.
Oh come on, could you be any more biased towards Linux? If this was just about any other OS outside of maybe OpenBSD you'd be all like "the OS is the problem."
Seriously. I like Linux just as much as the next Slashdot zealot, but lets be real. There is a decent chance that Linux running on big IBM hardware really wasn't/isn't as stable as the standard IBM alternative. Yeah, I know IBM supports Linux on those machines, but they don't keep AIX around for nothin'.
Like pretty much all MacOS software, you mean? Seriously, have a look in ~/Library some time.
That isn't spewing files all over the system. That is personal data related to the application all put in one well known place. Even Linux package managers dont' clean up any of that stuff. Do Windows unintallers? The only Windows programs I've seen actually clean up user information is when that user information is (wrongly) stored in the C:\Program Files\ structure. I suppose SOME uninstallers will ask you about saved games and such, but then you start complicating things. Each program has its own uninstaller and you get a lot of inconsistency for users. I think the important thing in making a system user friendly is keeping it simple. Uninstall = Trash. Easy.
You'd be amazed how much crap is left behind there and never, ever removed, because you deleted the application that created it but there's no facility whatsoever available to clean up after it.
The Finder works for me. Part of periodic house cleaning. No big deal. Everything is usually very clearly labeled.
You know, I think Apple might be onto something with their complete lack of any sort of package management built in (you can get a third party tool). I think the ideal situation is one where you don't need special tools to remove software. You just dump it in the trash. Though some software will inevitably spew files all over your system no matter what OS you run.
With that in mind, I think that the Linux style of having EVERYTHING as packages probably detracts from user-friendliness. I mean, it is nice for power users, but most people just want to find that Yahoo game that was installed, and remove it. Then again, maybe they don't. Sometimes I wonder if users ever delete ANYTHING.:P
lash Video is tied to the Flash Desktop Player. More and more people are starting to take their video watching beyond the desktop - portable media players, PSP's, PDA's, mobile phones, set-top-boxes on your large TV screen are all candidates for video watching, especially short videos such as found on YouTube,
YouTube style video is exactly NOT the type of video that people want to watch on a set-top-box, mobile, etc. YouTube is something you view when you're bored at home surfing the web. Who'd want to stretch that overcompressed video to a large screen TV? That would just look stupid. And besides, the videos are like 4 minutes long. That isn't something you sit down on your couch to watch. YouTube is interactive. It is not a service where you queue up a bunch of videos and just sit back and watch.
Well, if it really matters, they'll GET support for the flash video format. But for now, YouTube, Google Video, etc are aimed at the browser.
Linux has Flash Player. I believe it is even up to the same version as Windows and OS X versions. That's one of the reasons why Flash is such a great way of doing video on the web. Despite the proprietary file format, it is, overall, the most portable method of delivering video in a browser.
What's wrong with Flash video? I find that it works the most consistently and reliably of all the ways of doing video that I've seen so far. I think it is great. So what if it is a proprietary format? You're not meant to download the video files anyway.
I am working a site that accepts video upload (not a YouTube knock-off) and I just started implementing a transcoding system to convert all videos to FLV so we can quit our dependence on Quicktime which not everyone has. Flash is much more portable. Remember, it isn't just about the file format. It is also about embedding the video in the web page. Fact is that Flash is currently the most portable and reliable way of doing it. Plus, we get to control the player.
Way to condescend. I know that a PRI is just a T1. That was the point. T1 is the same kind of circuit that many businesses relying for phone service... and it is extremely reliable.
Next time maybe you could just say "you're right" or not say anything at all.
But a lot of that amounts to the level of service/support and not necessarily something inherent in T1 service. I am sure a T1 gets you a little bit closer to the Internet before sharing than, say, DSL, but it is still going to be shared inside you're ISP's network at some point. And they're most definately overselling their total bandwidth. So you don't really have anything guaranteed. It is just that they will presumably be more willing to work with you when you don't get the service you paid for.
The nice thing about T1 service is that you usually have more options ISP-wise. Since the T1 is point-to-point and doesn't require the ISP to have any physical presence in your local CO, you can connect to basically any ISP in your general area that is equiped to accept a virtual circuit (via T3/OC12/etc).
I remember a story from a guy who ended up getting a really cheap T1 to his rural location by ordering 3 (or maybe it was 4?) POTS lines. Because each POTS line uses its own pair up to the CO and copper pairs are in limited supply, it is policy to just run a DS1 circuit for more than a certain number of POTS lines. Somehow he worked the system and ended up getting access to the whole 24 channels and got T1 internet service (minus the 3 channels for the POTS) for the cost of 3 POTS lines. Well, he had to pay the ISP for internet service, but the line itself was cheap.
Hmm, I wish I could remember all the details. It was an awesome story demonstrating just how messed up telephone billing is.
I expect it's better now that we don't have the same sort of churning and consolidation in the industry. But my experience with T1 lines both at my ISP, and at other jobs, where we had them brought in, has been a lot rockier than anything I've ever experienced on either DSL or cable lines at home.
Was it the ISP (data)? Or the actual T1 circuits that were unrealiable? I ask because there is no reason why your data T1 circuit should be any less reliable than your average corporate PRI phone circuit (which is pretty darn reliable). It was most likely ISP problems (routing and such), which is not necessarily related to the specific type of circuit you used.
Indeed, we've had T1 service at work even during a neighborhood power failure. It is as realiable as phone service. Then again, I can't remember the last time my home DSL went out other than me flaking and forgetting to pay the bill.;-)
You do not, i repeat: DO NOT, have 1.5MB guaranteed up to the "peering point" with a T1. If your phone company told you that, they lied. Once your line (probably a virtual circuit) hits a data router, all bets are off. There is no guarantee that that the router your line ends up at will be an edge router. In fact, you probably dont' want it to be an edge/peering rouer. You want the opportunity to go over multiple peering points... presumably the best for your route.
No, but getting it installed in a userfriendly manner on multiple LInux distributions can be a pain. Not to mention library dependencies and version mismatches. Commercial vendors often resort to building static binaries and write their own home grown installers that just don't integrate well. And when they do decide to go with the distribution native packages/shared libs, they end up having to maintain like 5 different versions. Each with their own quirks. So really this gets back to the main point that sometimes choice is bad.
furthermore, the mysql query browser screws up when you paste a query containing tabs. this particular bug was reported to mysql over a year ago. it was somewhat fixed, but not really.
I've also had some serious problems with the ODBC client for Windows. There was one bug when accessing views on a MySQL > 5.0.10 server through ODBC that, as far as I know, has yet to be fixed. It has been open for over a year. And it is pretty serious. Basically you can't query a view from MS Access (didn't try from other programs). I had to get a user supplied patch, figure out how to recreate their ancient VC6 build environment, and recompile the ODBC client.
But I guess I should be grateful that I could recompile with a user supplied patch. There's commercial software that I use right now that leaves me helpless until the company gets around to releasing a binary patch. At least with open source there is some chance that I can get in there and fix a problem myself (i've done it).
It's not just how far north you are; weather patterns play as much of a role as latitude. Looking at an insolation map, where I am in Iowa shows up as ~3.5 kWh/m^2/day. The Seattle area shows up as 2 kWh/m^2.day. No part of Washington shows up as better than ~2.8 or so. Meanwhile, the desert southwest approaches 6 kWh/m^2/day.
That is suprising because eastern/central Washington is more or less a desert.. high desert. It's certainly nothing at all like the Seattle area. I wonder what causes the insolation.
Slashdot Mirror
I couldn't disagree more. What is boring are simplistic labels like Good and Evil. Good is never good as seems on the surface and Evil always has its side to the story. I guess there is still room for the superficial battles between Good and Evil, but truely rich, engaging, insightful stories have a depth that is beyond good and evil. Sometimes it is more like "light" and "dark," but I think that is more a matter of personalities and not necessarily based on some religious sense of absolute good and evil. For example, someone might like morbid music and maybe prefer to spend their waking hours at nighttime. This person might have a dark personality. But it doesn't make them evil. Evil is as evil does, and Good is just as likely to do evil... in the real world, at least.
-matthew
Well, that is the other extreme argument, I suppose. I was merely countering the idea that Linux (running on IBM hardware, I presume) has no paid support and that your only recourse to problems is posting a bug report on "Freshmeat." It just isn't true.
No, paid support isn't perfect. But it is another tool in the chest. When all the normal channels fail, it is sometimes nice to have a paid, professional engineer on your problem. And if you're Quantas, you'll probably get more than one regardless of whether you're running Linux or AIX.
-matthew
The point is that the GP came the conclusion that the system design was the problem with absolutely no information. I certainly didn't read anything in TFA that suggested they were trying to do anything unusual or untested with Linux. Of course, I don't know that they weren't either.
Anyway, If this were Windows we were talking about, there would be no question that it was the OS.
-matthew
There's nothing cheap about any IBM solution.. particularly the kind of solution where AIX is an option. As for the MySQL comment, they are runnign Oracle according to this article: http://www.zdnet.com.au/news/software/soa/Qantas-
How many airlines these days are actually growing?
What makes you think they weren't running LInux on a mainframe? IBM has has Linux running under a hypervisor on their big iron machines for some time now.
-matthew
Show your ignorance much?
AFAIK, IBM fully supports running Linux on their hardware. I'm sure whatever contract Quantas has with IBM covers Linux so they don't have to be posting bug reports on "Freshmeat" (still can't believe you said that) when the system goes down. That said, even Linux support from IBM may not as good as running AIX which (presumably) wouldn't have had the given problems in the first place.
-matthew
Oh come on, could you be any more biased towards Linux? If this was just about any other OS outside of maybe OpenBSD you'd be all like "the OS is the problem."
Seriously. I like Linux just as much as the next Slashdot zealot, but lets be real. There is a decent chance that Linux running on big IBM hardware really wasn't/isn't as stable as the standard IBM alternative. Yeah, I know IBM supports Linux on those machines, but they don't keep AIX around for nothin'.
-matthew
Even with enough oxygen, you'd need to worry about global warming with all those greenhouse gases such as plasmafied helium...
-matthew
Well, I dunno about a mile away... your scope would have to be pretty powerful. But I suppose it isn't impossible...
-matthew
That isn't spewing files all over the system. That is personal data related to the application all put in one well known place. Even Linux package managers dont' clean up any of that stuff. Do Windows unintallers? The only Windows programs I've seen actually clean up user information is when that user information is (wrongly) stored in the C:\Program Files\ structure. I suppose SOME uninstallers will ask you about saved games and such, but then you start complicating things. Each program has its own uninstaller and you get a lot of inconsistency for users. I think the important thing in making a system user friendly is keeping it simple. Uninstall = Trash. Easy.
The Finder works for me. Part of periodic house cleaning. No big deal. Everything is usually very clearly labeled.
-matthew
You know, I think Apple might be onto something with their complete lack of any sort of package management built in (you can get a third party tool). I think the ideal situation is one where you don't need special tools to remove software. You just dump it in the trash. Though some software will inevitably spew files all over your system no matter what OS you run.
:P
With that in mind, I think that the Linux style of having EVERYTHING as packages probably detracts from user-friendliness. I mean, it is nice for power users, but most people just want to find that Yahoo game that was installed, and remove it. Then again, maybe they don't. Sometimes I wonder if users ever delete ANYTHING.
-matthew
Oops, that last bit was supposed to be a response to your last sentence. Preview preview preview...
The only thing you're going to be measuring in ass piles is physical discomfort.
-matthew
YouTube style video is exactly NOT the type of video that people want to watch on a set-top-box, mobile, etc. YouTube is something you view when you're bored at home surfing the web. Who'd want to stretch that overcompressed video to a large screen TV? That would just look stupid. And besides, the videos are like 4 minutes long. That isn't something you sit down on your couch to watch. YouTube is interactive. It is not a service where you queue up a bunch of videos and just sit back and watch.
Linux has Flash Player. I believe it is even up to the same version as Windows and OS X versions. That's one of the reasons why Flash is such a great way of doing video on the web. Despite the proprietary file format, it is, overall, the most portable method of delivering video in a browser.
-matthew
What's wrong with Flash video? I find that it works the most consistently and reliably of all the ways of doing video that I've seen so far. I think it is great. So what if it is a proprietary format? You're not meant to download the video files anyway.
I am working a site that accepts video upload (not a YouTube knock-off) and I just started implementing a transcoding system to convert all videos to FLV so we can quit our dependence on Quicktime which not everyone has. Flash is much more portable. Remember, it isn't just about the file format. It is also about embedding the video in the web page. Fact is that Flash is currently the most portable and reliable way of doing it. Plus, we get to control the player.
-matthew
Way to condescend. I know that a PRI is just a T1. That was the point. T1 is the same kind of circuit that many businesses relying for phone service... and it is extremely reliable.
Next time maybe you could just say "you're right" or not say anything at all.
But a lot of that amounts to the level of service/support and not necessarily something inherent in T1 service. I am sure a T1 gets you a little bit closer to the Internet before sharing than, say, DSL, but it is still going to be shared inside you're ISP's network at some point. And they're most definately overselling their total bandwidth. So you don't really have anything guaranteed. It is just that they will presumably be more willing to work with you when you don't get the service you paid for.
The nice thing about T1 service is that you usually have more options ISP-wise. Since the T1 is point-to-point and doesn't require the ISP to have any physical presence in your local CO, you can connect to basically any ISP in your general area that is equiped to accept a virtual circuit (via T3/OC12/etc).
-matthew
I remember a story from a guy who ended up getting a really cheap T1 to his rural location by ordering 3 (or maybe it was 4?) POTS lines. Because each POTS line uses its own pair up to the CO and copper pairs are in limited supply, it is policy to just run a DS1 circuit for more than a certain number of POTS lines. Somehow he worked the system and ended up getting access to the whole 24 channels and got T1 internet service (minus the 3 channels for the POTS) for the cost of 3 POTS lines. Well, he had to pay the ISP for internet service, but the line itself was cheap.
Hmm, I wish I could remember all the details. It was an awesome story demonstrating just how messed up telephone billing is.
-matthew
Was it the ISP (data)? Or the actual T1 circuits that were unrealiable? I ask because there is no reason why your data T1 circuit should be any less reliable than your average corporate PRI phone circuit (which is pretty darn reliable). It was most likely ISP problems (routing and such), which is not necessarily related to the specific type of circuit you used.
-matthew
Indeed, we've had T1 service at work even during a neighborhood power failure. It is as realiable as phone service. Then again, I can't remember the last time my home DSL went out other than me flaking and forgetting to pay the bill. ;-)
-matthew
You do not, i repeat: DO NOT, have 1.5MB guaranteed up to the "peering point" with a T1. If your phone company told you that, they lied. Once your line (probably a virtual circuit) hits a data router, all bets are off. There is no guarantee that that the router your line ends up at will be an edge router. In fact, you probably dont' want it to be an edge/peering rouer. You want the opportunity to go over multiple peering points... presumably the best for your route.
-matthew
No, but getting it installed in a userfriendly manner on multiple LInux distributions can be a pain. Not to mention library dependencies and version mismatches. Commercial vendors often resort to building static binaries and write their own home grown installers that just don't integrate well. And when they do decide to go with the distribution native packages/shared libs, they end up having to maintain like 5 different versions. Each with their own quirks. So really this gets back to the main point that sometimes choice is bad.
-matthew
I've also had some serious problems with the ODBC client for Windows. There was one bug when accessing views on a MySQL > 5.0.10 server through ODBC that, as far as I know, has yet to be fixed. It has been open for over a year. And it is pretty serious. Basically you can't query a view from MS Access (didn't try from other programs). I had to get a user supplied patch, figure out how to recreate their ancient VC6 build environment, and recompile the ODBC client.
But I guess I should be grateful that I could recompile with a user supplied patch. There's commercial software that I use right now that leaves me helpless until the company gets around to releasing a binary patch. At least with open source there is some chance that I can get in there and fix a problem myself (i've done it).
-matthew
That is suprising because eastern/central Washington is more or less a desert.. high desert. It's certainly nothing at all like the Seattle area. I wonder what causes the insolation.
-matthew
Just FYI, Washington isn't that much farther north than Iowa and eastern Washington gets plenty of sunshine.