GOOG is one of the best trading stocks around in the last couple months. Ride it up, ride it down, up, down, over and over. Don't change a thing! Who cares about long term stock ownership anymore, anyway? Didn't we learn our lesson about that in 2000?
Would you buy an operating system for your computer from Ford Motor Company? If you could, would you expect it to be produced efficiently and be a quality product? (perhaps I selected a bad example:-) )
Well, you kind of did, because OSes and cars have nothing to do with each other. However if you look at GM's OnStar service, there you have a good example of a service that a car maker has branched into for added revenue (not that it's helped much in their case)
I agree that telcos are horrible at providing services and media. I never claimed any different (except Comcast, but they were already in the content delivery business). I'm just explaining why they keep trying to do it. It's kind of like Sony continually trying to make the world adopt one of their proprietary media formats. They've tried over and over for 30 years, it's never happened, but they're still hoping they can get just once chance at it. Same with any of these telcos and the content they're trying to deliver. They're just hoping for one thing to stick and become a massive influx of new cash.
Re:Is anyone actually using tv phone functionality
on
No 3G for HP Until 2007
·
· Score: 3, Insightful
Telcos don't get it. Provide bandwidth, and let people build services that run on that bandwidth.
No, actually they do get it. A one way ticket to obsolescence is just being a bandwidth provider. Ask any mid-level dialup ISP from the 90s how they're doing today now that Comcast and SBC have deployed broadband. Where do you think wireless providers will be if someone actually deploys wi-fi throughout a city, or a new technology comes along that provides 5mbit download speeds anywhere?
That's why every wireless provider... scratch that... every media provider of any kind... is trying to bring added value content to the consumer. Ringtones, portable video, SMS: these are all services desperately trying to avoid being just another bandwidth provider. What do you think all of those DSL deals with Yahoo, or Comcast ON-Demand is all about? Bandwidth always gets cheaper until it becomes commoditized. Telcos have to stay ahead of that curve and becomes content providers. So when someone else says they can get a faster line than Comcast to your house, you'll say "Aww, but it doesn't have On-Demand?"
THAT is the point of VCast and every other lame attempt to avoid becoming just pure bandwidth providers.
Investing in VZW certainly must've worked out better than investing in any other US cellular company would have at the time, Verizon was the #1 carrier until the Cingular-AT&T merger.
If memory serves, VZW will be regaining that spot later this year. So many people were already bailing from AWE that the buyout only put Verizon out of their spot for a short time.
Vodafone also have a stake in a CDMA network in the US. How long before they divest that as well? When they are backing 3GSM in the rest of the world it seems odd to have a stake in CDMA in the US.
I think Vodafone bought Airtouch because they were looking for any entry point possible into the US, and the decision to go with CDMA was made before that acquisition. To rip it up and deploy GSM for pure technology compatibility would have been too expensive and frankly not worth it. Most US users don't think about or care about GSM or CDMA.
To answer your "how long until they divest that" question... the answer is no time soon. Vodafone got outbid for the AT&T Wireless network. At $41b, that would have been an enormous price to pay to acquire a network that (still) needs a lot of help to get to 3GSM. Meanwhile, they're deploying EV-DO like gangbusters on VZW. Frankly, I think the Vodafone bid for AWE was a token bid, they had no intention of outbidding SBC and Bellsouth.
So today, either they would have to acquire T-Mobile, a hopelessly small network with very little spectrum and one that DT probably doesn't want to sell, or they should stick with Verizon Wireless. Unlike Japan, they would never get rid of their VZW piece unless they could buy another network in the US. The US is the most lucrative mobile phone market in the world, the only way they'd leave the VZW deal is if they could find a way to get their own brand deployed into the US. Its been a while since I looked, but their VZW ownership had paid them upwards of $1b dividends in the past.
I'd probably use C, but to use it for a standard web app would be the very definition of premature optimization in my eyes.
At what point is it not premature to optimize a website? Should we start optimizing when the hockey stick takes over, and our site goes down? Or when it gets/.ed?
Just like any other programmer, web programmers need to design to performance requirements and optimize along the way. If the required performance requires using C, do that from the start. If you plan to deploy to a million people, why would you start by writing a website that only deploys to 500, then spend the next two years trying to dig yourself out of that hole? I'm sure the Orkut and Friendster developers have thoughts on how that's gone for them, or several thousand other web designers whose websites have died under load in the name of easier programming.
I'm not saying it's impossible meet high performance requirements with Ruby or Python. I'm just saying that it's insane not to design to some level of performance in a market where your server can get crushed overnight.
I'm overwhelmed by your citation of sources and real-world examples proving your claim of Ruby's slowness!
Rails, not counting rendering or database query, takes anywhere from a few to several millseconds to respond to any web query on a modern Xeon processor.
I'm not sure I understand the argument of "real world examples" when it comes to performance. Either someone cares about performance or they don't. Over the span of the entire life of Rails, the argument has been "the database is the bottleneck anyway" and "well, it's not slow enough for my blog app!" Ok, so when is it slow enough? We live in a world where people still have to bust out ASM to get desired performance for some applications. For web applications, is performance so unimportant compared to development time that we're willing to make any sacrifice necessary? I don't agree with that sentiment, and neither does the developer of Ruby, who's working on a new, performance-oriented interpreter http://www.rubygarden.org/ruby?Rite
No - it is a killer app for getting mentions on Slashdot. Having a development system with the relative lack of performance of Ruby, and the very close tie-in to the database and schema of Rails is more of a website killer than an killer app, I am afraid to report.
I'm with Decaff on this one. I drank the RoR kool-aid after one of the earlier posts to slashdot. The first few weeks was awesome. Then essentially what happened was I ended up trying to rip out every aspect of RoR until I was just left with Ruby... which had terrible performance.
If you're going to go with RoR, make sure you take the long view. While everyone says "it scales, because it has FastCGI", I'd really like to hear more about extreme high volume sites that are using it. How's it going for them?
Before you invest a lot of time, effort and money crafting a GUI front-end for your application, you should really stop and consider that you may not need one.... Especially now that AJAX is on the scene... modern AJAX tools and a Java backend can put together some very powerful applications
First point is that if you can't accomplish what you want to accomplish in HTML with post-backs, you aren't going to do it with AJAX. AJAX does little more than load panes in the background.
Second point is it's a lot harder to develop good UIs with AJAX. It takes a remarkable amount of work to get simple things to work in Javascript. There are starting points like Prototype, but when you're finished, it still doesn't work as well as an hour with.NET, SWT or QT. And users have no familiarity with how these arcane UIs are supposed to work. There's no standard.
AJAX, IMO, is mostly good for one type of organization: internet companies who deploy to the public. Deploying AJAX-y intranet applications to the enterprise sounds like that group needs to revisit their requirements. Most organizations won't want to go there when a harsh light is placed on the idea.
I've heard for at least a year that MS is going for 100% XHTML 1.0 strict compliance on their MSN sites. And the main site is also 100% 4.0 compliant. Not that it's particularly "Web 2.0".
An amazing framework will be easy to use, well documented, intuitive and will make you feel smart & productive all within 10 minutes.
This is quite possibly the most ludicrous "rule" I've ever heard. Granted, I've spent a lot of time trying to wrap my head around crazy Java frameworks and have wondered why they couldn't just make some tutorial that would make it easy to get into the software. That said, the world isn't built in 10 minutes.
It's easy for any framework to give you a 10 minute experience. The problem comes once you've gotten past that 10 minutes. I was blown away with my early experiences with Ruby on Rails. However, as soon as I started getting deep into it and wanted to do more complicated things... sure enough, that up-front simplicity reared its ugly head by adding more complexity later. Try subclassing and extending ActiveRecord sometime. Some of the code they've created to allow for the simplicity of the Rails experience is mightily arcane once you want to make some simple extensions.
So, I think the 10 minute rule is great for hype, horrible in the real world. 10 minutes is what it takes to do a good demo, nothing more... and we all know what demos are worth. There are plenty of extremely powerful tools and frameworks out there in all disciplines (the web isn't the only thing people do, although you wouldn't know it from looking at slashdot) that take significantly more time to feel productive in than 10 minutes. But that up front learning curve pays off later.
Go ahead - try to send video of your newborn baby to your relatives across the country using 56k.
Uhhh... try to send it with ADSL. The upload speeds on some of these broadband packages is so crippled that it's still painful to try to upload anything. Sure, they'll give you 8mbit downstream for ad delivery, but you have to pay out the nose to get any more than 300kbit up.
You guys don't see one major problem with OSS: the potential for copyright and patent infringement.
Microsoft is never going to come after its users and claim some kind of patent infringement. If someone goes after Microsoft claiming that, it becomes their problem, not the end user. Now imagine you had used an OSS product that gets shut down for patent infringement. No guarantees are made to protect your company from exposure to that lawsuit.
Or just imagine the OSS had been lifted from other GPLed software and put under a license you believed to be GPL-free. Oops, suddenly you have GPL guys coming after your company for extending GPLed code.
Undoubtedly Intel is paying for this. eBay has to figure out every way possible they can make money with Skype, which could be the most overpriced acquisition since Time Warner and AOL.
Is it really worth the hassle to use Windows, especially when it comes to business users?
I'd say so.
Better RAD tools available, like.NET.
Better/standard office application available.
About a 99.9% chance employees have used Windows and have experience with it
Better selection of general applications
No concerns with a fragile glibc and other fun linux configuration problems
Honestly, I think you'd have the kind of problem in TFA if OpenBSD or Solaris were deployed to 95% of the desktop world like Windows. It's the bane of being the #1 OS, you'll have application battles/incompatiblities/whatever.
Just curious, what kind of businesses do you deploy to? I haven't worked anywhere where I disagree with having the desktop be on Windows. I've worked many places where it would have paid to have more linux/bsd servers, though.
I like the different languages the CLR can run, but it doesn't help the problem the OP describes.
When you have a dozen, or dozens, of developers having to go into, understand and modify each other's code, the.NET language agnostic VM won't help you. What if the one guy writing in Boo up and quits one day. The guys all writing in C#, C++, VB.NET, whatever language(s) you've allowed, have to take it over. That costs time and money. Up front, it also indicates how much of a risk allowing random languages can be.
I think I disagree with most of the other posters here, it is a good idea to standardize on a language. While good programmers can jump into just about any language and understand, it takes years to be an expert (C++ anyone?). That said, it is sometimes necessary to introduce multiple languages. Many Java and.NET developers still bridge the gap to C or C++ very often. Even so, I just don't agree that introducing a language like Python or Ruby into that environment necessarily helps productivity. It might help the one guy who loves to use that language, but overall it just hurts the project.
SQL ended up in the same category as Fortran and Assembly? I'd guess Sql's probably the most used language in the world behind C, C++ or Java -- all of these exotic languages people are mentioning (Ruby, Haskell) still go through SQL to utilize any relational database.
Slashdot Mirror
GOOG is one of the best trading stocks around in the last couple months. Ride it up, ride it down, up, down, over and over. Don't change a thing! Who cares about long term stock ownership anymore, anyway? Didn't we learn our lesson about that in 2000?
If you're going to publish something for hundreds of thousands of people to read
./ comments. :)
Yes, because for the really important news and insight, people turn to the
Would you buy an operating system for your computer from Ford Motor Company? If you could, would you expect it to be produced efficiently and be a quality product? (perhaps I selected a bad example :-) )
Well, you kind of did, because OSes and cars have nothing to do with each other. However if you look at GM's OnStar service, there you have a good example of a service that a car maker has branched into for added revenue (not that it's helped much in their case)
I agree that telcos are horrible at providing services and media. I never claimed any different (except Comcast, but they were already in the content delivery business). I'm just explaining why they keep trying to do it. It's kind of like Sony continually trying to make the world adopt one of their proprietary media formats. They've tried over and over for 30 years, it's never happened, but they're still hoping they can get just once chance at it. Same with any of these telcos and the content they're trying to deliver. They're just hoping for one thing to stick and become a massive influx of new cash.
Telcos don't get it. Provide bandwidth, and let people build services that run on that bandwidth.
No, actually they do get it. A one way ticket to obsolescence is just being a bandwidth provider. Ask any mid-level dialup ISP from the 90s how they're doing today now that Comcast and SBC have deployed broadband. Where do you think wireless providers will be if someone actually deploys wi-fi throughout a city, or a new technology comes along that provides 5mbit download speeds anywhere?
That's why every wireless provider... scratch that... every media provider of any kind... is trying to bring added value content to the consumer. Ringtones, portable video, SMS: these are all services desperately trying to avoid being just another bandwidth provider. What do you think all of those DSL deals with Yahoo, or Comcast ON-Demand is all about? Bandwidth always gets cheaper until it becomes commoditized. Telcos have to stay ahead of that curve and becomes content providers. So when someone else says they can get a faster line than Comcast to your house, you'll say "Aww, but it doesn't have On-Demand?"
THAT is the point of VCast and every other lame attempt to avoid becoming just pure bandwidth providers.
Investing in VZW certainly must've worked out better than investing in any other US cellular company would have at the time, Verizon was the #1 carrier until the Cingular-AT&T merger.
If memory serves, VZW will be regaining that spot later this year. So many people were already bailing from AWE that the buyout only put Verizon out of their spot for a short time.
Vodafone also have a stake in a CDMA network in the US. How long before they divest that as well? When they are backing 3GSM in the rest of the world it seems odd to have a stake in CDMA in the US.
I think Vodafone bought Airtouch because they were looking for any entry point possible into the US, and the decision to go with CDMA was made before that acquisition. To rip it up and deploy GSM for pure technology compatibility would have been too expensive and frankly not worth it. Most US users don't think about or care about GSM or CDMA.
To answer your "how long until they divest that" question... the answer is no time soon. Vodafone got outbid for the AT&T Wireless network. At $41b, that would have been an enormous price to pay to acquire a network that (still) needs a lot of help to get to 3GSM. Meanwhile, they're deploying EV-DO like gangbusters on VZW. Frankly, I think the Vodafone bid for AWE was a token bid, they had no intention of outbidding SBC and Bellsouth.
So today, either they would have to acquire T-Mobile, a hopelessly small network with very little spectrum and one that DT probably doesn't want to sell, or they should stick with Verizon Wireless. Unlike Japan, they would never get rid of their VZW piece unless they could buy another network in the US. The US is the most lucrative mobile phone market in the world, the only way they'd leave the VZW deal is if they could find a way to get their own brand deployed into the US. Its been a while since I looked, but their VZW ownership had paid them upwards of $1b dividends in the past.
Isn't this exactly what corporate law has been doing for several decades?
I'd probably use C, but to use it for a standard web app would be the very definition of premature optimization in my eyes.
/.ed?
At what point is it not premature to optimize a website? Should we start optimizing when the hockey stick takes over, and our site goes down? Or when it gets
Just like any other programmer, web programmers need to design to performance requirements and optimize along the way. If the required performance requires using C, do that from the start. If you plan to deploy to a million people, why would you start by writing a website that only deploys to 500, then spend the next two years trying to dig yourself out of that hole? I'm sure the Orkut and Friendster developers have thoughts on how that's gone for them, or several thousand other web designers whose websites have died under load in the name of easier programming.
I'm not saying it's impossible meet high performance requirements with Ruby or Python. I'm just saying that it's insane not to design to some level of performance in a market where your server can get crushed overnight.
I'm overwhelmed by your citation of sources and real-world examples proving your claim of Ruby's slowness!
Rails, not counting rendering or database query, takes anywhere from a few to several millseconds to respond to any web query on a modern Xeon processor.
I'm not sure I understand the argument of "real world examples" when it comes to performance. Either someone cares about performance or they don't. Over the span of the entire life of Rails, the argument has been "the database is the bottleneck anyway" and "well, it's not slow enough for my blog app!" Ok, so when is it slow enough? We live in a world where people still have to bust out ASM to get desired performance for some applications. For web applications, is performance so unimportant compared to development time that we're willing to make any sacrifice necessary? I don't agree with that sentiment, and neither does the developer of Ruby, who's working on a new, performance-oriented interpreter http://www.rubygarden.org/ruby?Rite
.NET's ORM system
.NET, check out Wilson's ORM, or check out NHibernate.
Do you mean ADO.NET, which is not ORM? Or do you mean ObjectSpaces, which is not yet released?
If you're serious about ORM in
No - it is a killer app for getting mentions on Slashdot. Having a development system with the relative lack of performance of Ruby, and the very close tie-in to the database and schema of Rails is more of a website killer than an killer app, I am afraid to report.
I'm with Decaff on this one. I drank the RoR kool-aid after one of the earlier posts to slashdot. The first few weeks was awesome. Then essentially what happened was I ended up trying to rip out every aspect of RoR until I was just left with Ruby... which had terrible performance.
If you're going to go with RoR, make sure you take the long view. While everyone says "it scales, because it has FastCGI", I'd really like to hear more about extreme high volume sites that are using it. How's it going for them?
Hey Ruby/Rails users...giving you a chance to evangelize
Don't worry, they've taken that chance for about a year now... trolling on every Java board and blog around.
The correct answer is, "None of the Above."
.NET, SWT or QT. And users have no familiarity with how these arcane UIs are supposed to work. There's no standard.
Before you invest a lot of time, effort and money crafting a GUI front-end for your application, you should really stop and consider that you may not need one.... Especially now that AJAX is on the scene... modern AJAX tools and a Java backend can put together some very powerful applications
First point is that if you can't accomplish what you want to accomplish in HTML with post-backs, you aren't going to do it with AJAX. AJAX does little more than load panes in the background.
Second point is it's a lot harder to develop good UIs with AJAX. It takes a remarkable amount of work to get simple things to work in Javascript. There are starting points like Prototype, but when you're finished, it still doesn't work as well as an hour with
AJAX, IMO, is mostly good for one type of organization: internet companies who deploy to the public. Deploying AJAX-y intranet applications to the enterprise sounds like that group needs to revisit their requirements. Most organizations won't want to go there when a harsh light is placed on the idea.
And guess who has just two errors:
msn.com
I've heard for at least a year that MS is going for 100% XHTML 1.0 strict compliance on their MSN sites. And the main site is also 100% 4.0 compliant. Not that it's particularly "Web 2.0".
An amazing framework will be easy to use, well documented, intuitive and will make you feel smart & productive all within 10 minutes.
This is quite possibly the most ludicrous "rule" I've ever heard. Granted, I've spent a lot of time trying to wrap my head around crazy Java frameworks and have wondered why they couldn't just make some tutorial that would make it easy to get into the software. That said, the world isn't built in 10 minutes.
It's easy for any framework to give you a 10 minute experience. The problem comes once you've gotten past that 10 minutes. I was blown away with my early experiences with Ruby on Rails. However, as soon as I started getting deep into it and wanted to do more complicated things... sure enough, that up-front simplicity reared its ugly head by adding more complexity later. Try subclassing and extending ActiveRecord sometime. Some of the code they've created to allow for the simplicity of the Rails experience is mightily arcane once you want to make some simple extensions.
So, I think the 10 minute rule is great for hype, horrible in the real world. 10 minutes is what it takes to do a good demo, nothing more... and we all know what demos are worth. There are plenty of extremely powerful tools and frameworks out there in all disciplines (the web isn't the only thing people do, although you wouldn't know it from looking at slashdot) that take significantly more time to feel productive in than 10 minutes. But that up front learning curve pays off later.
Go ahead - try to send video of your newborn baby to your relatives across the country using 56k.
Uhhh... try to send it with ADSL. The upload speeds on some of these broadband packages is so crippled that it's still painful to try to upload anything. Sure, they'll give you 8mbit downstream for ad delivery, but you have to pay out the nose to get any more than 300kbit up.
You guys don't see one major problem with OSS: the potential for copyright and patent infringement.
Microsoft is never going to come after its users and claim some kind of patent infringement. If someone goes after Microsoft claiming that, it becomes their problem, not the end user. Now imagine you had used an OSS product that gets shut down for patent infringement. No guarantees are made to protect your company from exposure to that lawsuit.
Or just imagine the OSS had been lifted from other GPLed software and put under a license you believed to be GPL-free. Oops, suddenly you have GPL guys coming after your company for extending GPLed code.
Also, Apple marketshare, .... [is] at their highest ever
Surely you're not considering the early 80s in that assessment. Apple dominated marketshare then.
Undoubtedly Intel is paying for this. eBay has to figure out every way possible they can make money with Skype, which could be the most overpriced acquisition since Time Warner and AOL.
but even with access on planes, there are always times and places you'll be cut off.
Like when there's snakes on a plane?
I'd say so.
Honestly, I think you'd have the kind of problem in TFA if OpenBSD or Solaris were deployed to 95% of the desktop world like Windows. It's the bane of being the #1 OS, you'll have application battles/incompatiblities/whatever.
Just curious, what kind of businesses do you deploy to? I haven't worked anywhere where I disagree with having the desktop be on Windows. I've worked many places where it would have paid to have more linux/bsd servers, though.
I'd rather keep al my e-mail to my self, as a company...
Having email handled off-site by an independent third party is a great way to have S-OX compliance, especially if it never gets deleted.
I like the different languages the CLR can run, but it doesn't help the problem the OP describes.
.NET language agnostic VM won't help you. What if the one guy writing in Boo up and quits one day. The guys all writing in C#, C++, VB.NET, whatever language(s) you've allowed, have to take it over. That costs time and money. Up front, it also indicates how much of a risk allowing random languages can be.
.NET developers still bridge the gap to C or C++ very often. Even so, I just don't agree that introducing a language like Python or Ruby into that environment necessarily helps productivity. It might help the one guy who loves to use that language, but overall it just hurts the project.
When you have a dozen, or dozens, of developers having to go into, understand and modify each other's code, the
I think I disagree with most of the other posters here, it is a good idea to standardize on a language. While good programmers can jump into just about any language and understand, it takes years to be an expert (C++ anyone?). That said, it is sometimes necessary to introduce multiple languages. Many Java and
Special: Sql, Fortran, ASM
SQL ended up in the same category as Fortran and Assembly? I'd guess Sql's probably the most used language in the world behind C, C++ or Java -- all of these exotic languages people are mentioning (Ruby, Haskell) still go through SQL to utilize any relational database.
Well, I hope Novell is ready for a legal fight with Apple.
Apple has a patent on the Exposé behavior.