Everyone's fixated on music, but the best thing for us road warriors is it's a mobile access point. I can bring it around and use the hotel's wired access wirelessly from my room. And it's tiny and light, so it won't load me down -or- take up a lot of space in the laptop bag.
If you've never been on the road, you don't know how much of a pain it is to be stuck to the crappy desks most hotels have. Go wireless!
You can also (with a y-cable) attach it to the in-room TV, so you can ditch those annoyingly-heavy travel speakers. Yahoo!
Need to enable wireless in a conference room really quick? Plug in one of these puppies, and bang, you're ready to go. You can even configure the drop in the conference room as not connected to the inside net, allowing instant ad-hoc outside access.
It's also something else: a security nightmare for IT. Imagine the problem IT had with unauthorized modems. Now you can have rogue access points the size of a pack of cards hiding out somewhere in your organization. You'd never find the freaking thing.
The problem with that line of argument is that the project given to the PhD will be late, too complicated, irrelevant to the project requirements, and subject to the perverse whims of the PhD.
Not that it would be any different for a non-PhD candidate, but at least a normal engineer would be cheaper.
It just struck me how amazing it is that a Windows user is complaining about the lack of interface consistency in Linux.
Only a few years ago Windows advocates denigrated interface consistency as a feature for the small-minded. Applications needed different key combinations for common actions because ever app was different, and anyway enforcing a standard was a totalitarian thing to do. Choice, choice, choice.
Before that, of course, DOS users mocked GUIs, mice, large memory spaces, etc.
It's unbellievable how time really does change how people look at things.
"We figure that the best way to evaluate an idea is to code it up and see how it works in practice. A working feature now is better than a perfect implementation that still isn't done."
What they're saying is "we'd rather introduce features that don't quite work and fix them later instead of making sure the feature works and make sense before we add it to the app."
This is why open source gives some people the heebie-jeebies. With Photoshop, Illustrator, etc the vendor waits until a feature works, then releases it. It may not have all the bells & whistles that people want, and it does get refined as time goes on, but they do wait until its perfect (in its current state) before releasing it."
The user is not a guinea pig that you can f*ck around with.
One neat thing with Vonage is you can get a phone # with a non-local area code. So if you're in CA and have relatives in NY, you can get a 212/718/914 number. This, of course, means their calls you you are local calls for them. It's $4.95/mo for you, tho.
Coolness!
I just got my service activated today, so I've been poking around and seeing what's what. So far, it seems pretty good.
Dell is a distribution channel, plain and simple. Its costs are minimized, there's no R&D, etc. Unlike Wal-Mart, Dell rebrands many of the items that it sells. Wal-Mart does too, under private labels, but for some reason Dell has been able to create "Dell" as a brand, unlike most white-box builders.
Dell now sells a ridiculously large amount of computer equipment: $11 Billion last quarter, $44 Billion on an annualized basis. They sold as much stuff as Microsoft last quarter, and they made 50% less. They've cut the monopoly premium to 50%, with margins of about 23%.
Plus, there's no reason to think they're going to stop anytime soon. They are the low-cost provider, period.
New technology? Probably not. But they sure are a cheap place to get boxes.
I wonder if that data is accessible by 3rd parties. You could make "mail maps" that let you visualize the clustering of your incoming messages, and you could actually see the spam...by looking at the outliers and noise.
In fact, you could do this with any large data set. How about the feds looking for anomalous chunks of data in the bitstream? Anomalous stuff would just pop out, literally. This would make the TSA's job much, much easier. How about that?
Dude, don't you know anything about the stock market? Fundamentals make no difference. See:
http://www.blogmaverick.com/
for a real-world view of the market. He says it better than I could. Scroll down to the "Why do people fall in love with stocks" entry. Here's the direct link:
You know, he's doing a bayesian survey
on
Google IPO Swami
·
· Score: 5, Insightful
I just realized - he's doing a bayesian average!
http://www.research.att.com/~volinsky/bma.html There was a possibly apochryphal story about this. A plane went down in a large area, and they needed to find it. Nobody really knew what happened to it. The leader of the search team went and asked a bunch of pilots where they thought the plane was, after giving them the course, heading, speed, and whatever data was available.
Well, they took the answers, narrowed the search area, then found the plane pretty much where the consensus said it would be.
A bit of thought will give you the reason this might have worked...
The range will be small, and downwards
on
Google IPO Swami
·
· Score: 3, Interesting
One thing I've noticed about the posters is they're not really sure why the dutch auction format is different than the normal IPO process.
In a normal IPO, the investment bank (or brokerage) sets the price, say, $20, then buys the shares off of the issuing company for $20. Any spread off of that goes directly to the brokerage (or investment bank).
This is a Bad Deal for the issuing company, because the spread should have gone to the company, not to the brokerage. Think about it: if MSDW priced a stock at $20, but were able to sell at $200, then the issuing company just lost $180! The brokerage, of course, made $180.
With the dutch auction, there will (or should be) little or no upside to google's stock because the price will be determined by how much people are willing to pay up front. In fact, it will probably drop on the first day, since people confused by this concept will try and flip, leading to a sell imbalance.
Depending on the shares that are to be issued, I'd say that conservatively google's stock should open at anywhere from $175 +- 8%, and drop about 15%. Why $175? Because -everyone- knows google, and companies that are well-known to the public trade at a premium.
The other interesting thing about google is that a savvy investor/hacker could manipulate the earnings by writing a virus that sneakily clicked on all those google ads that appear when you do a google search. Voila, instant earnings! With enough spread, you could do millions of hits a day, which translates to millions of dollars a day (thank you, adwords!)
When people talk about support, they always say "hey, you can always call MS."
But have you? Do you? When a problem occurs, the go-to guy is the IT guy in the company. And that guy (or gal) either searches the net or asks a friend.
Have you, and IT person, ever called the MS helpline? If so, were you able to get an answer?
This is the problem with science in the context of public policy, and why that statement about the Bush administration's science policy is a bit out of whack.
Almost by definition, anything that recommends a solution is bad science. Science isn't very good at outcomes, but that's what politicians need.
In the case of global warming, it's difficult because the costs are imposed now, and the outcome is always in doubt. If we do X, there's no guarantee that X will happen. So are you willing to spend hundreds of billions or trillions of dollars and affect every industry for possibly no gain? Nope.
Science doesn't determine goals, direction, and priorities - politicians (and the public) do. And that's how it should be. Scientists don't pay a price if they're wrong.
Hey everyone, it's easy to make sure you don't get outsourced: provide value to your business.
No part of the business that is deemed "essential" would be outsourced. This means that IT is not considered essential, and it's not contributing to the business as a whole.
This is true of most IT - how much value-add do you bring? Do you actually help make the business better, or do you sit around talking about Lusers and how dumb they are? How those business people are morons? How they're so stupid they can't even turn on their PCs?
Are you a BOFH? Then you'll become an unemployed BOFH, and a happy worker someone else will do a better job than you.
Are you actively involved with your business groups, and understand how you help them make money? Then you won't get outsourced. Period.
Outsourcing is the business striking back at the geeks. The geeks have held sway for too long, basically removing value from businesses by being totally unresponsive to business needs. And if you can't get what you want at home, you go somewhere else. Its' that simple.
This has been bandied about so much that it's taken the place of real, critical thought.
Think about this: when Apple attempted to license its OS, it sales were immediately cannibalized by its licensees.
When Palm did the same thing, the same thing happened.
Name one company besides Microsoft that's succeeded financially in licensing its stuff to others. I can think of one, but they were already in a dominant market position and were/are licensing to stave off anti-trust issues.
Now everyone knows why Microsoft was sitting on all that cash: they're gonna buy their way outta trouble. Why not? It's easier for them to buy their way out of trouble.
Andd after all that, they'll still have billions and billions lying around to cross-subsidize their money-losing ventures. Those money-losing ventures, of course, include almost everything Microsoft does except Windows and Office.
Microsoft's new slogan should be "Innovating financial solutions to legal problems."
Funny, I was just thinking today about all those people complaining about Apple's 1-button mouse usually are the same ones that use the 102-button mouse on a daily basis. They'll never be happy with less.
As a metric goes, this one has held up pretty well for me over the years. Back in the day, if a feature took around one day to code it took around 4 days of UI work to get a UI that was decent. I'd never claim my UIs were good, but they got the job done and were clear and straight-forward.
That includes testing the UI, BTW. Not user testing, just making sure the UI actually worked.
That five-to-one ratio is an average. For simple CLIs the ratio is less (there's no UI, but there's a usage statement) and for complicated features the ratio is more.
"I am reminded at this point of a fellow I used to know whose name was Henry, only to give you an idea of what a individualist he was, he spelled it H-E-N-3-R-Y. The three was silent, you see."
- Tom Lehrer, from "An Evening (Wasted) with Tom Lehrer, 1959
Slashdot Mirror
Everyone's fixated on music, but the best thing for us road warriors is it's a mobile access point. I can bring it around and use the hotel's wired access wirelessly from my room. And it's tiny and light, so it won't load me down -or- take up a lot of space in the laptop bag.
If you've never been on the road, you don't know how much of a pain it is to be stuck to the crappy desks most hotels have. Go wireless!
You can also (with a y-cable) attach it to the in-room TV, so you can ditch those annoyingly-heavy travel speakers. Yahoo!
Need to enable wireless in a conference room really quick? Plug in one of these puppies, and bang, you're ready to go. You can even configure the drop in the conference room as not connected to the inside net, allowing instant ad-hoc outside access.
It's also something else: a security nightmare for IT. Imagine the problem IT had with unauthorized modems. Now you can have rogue access points the size of a pack of cards hiding out somewhere in your organization. You'd never find the freaking thing.
What a neato gadget!
The problem with that line of argument is that the project given to the PhD will be late, too complicated, irrelevant to the project requirements, and subject to the perverse whims of the PhD.
Not that it would be any different for a non-PhD candidate, but at least a normal engineer would be cheaper.
It just struck me how amazing it is that a Windows user is complaining about the lack of interface consistency in Linux.
Only a few years ago Windows advocates denigrated interface consistency as a feature for the small-minded. Applications needed different key combinations for common actions because ever app was different, and anyway enforcing a standard was a totalitarian thing to do. Choice, choice, choice.
Before that, of course, DOS users mocked GUIs, mice, large memory spaces, etc.
It's unbellievable how time really does change how people look at things.
"We figure that the best way to evaluate an idea is to code it up and see how it works in practice. A working feature now is better than a perfect implementation that still isn't done."
What they're saying is "we'd rather introduce features that don't quite work and fix them later instead of making sure the feature works and make sense before we add it to the app."
This is why open source gives some people the heebie-jeebies. With Photoshop, Illustrator, etc the vendor waits until a feature works, then releases it. It may not have all the bells & whistles that people want, and it does get refined as time goes on, but they do wait until its perfect (in its current state) before releasing it."
The user is not a guinea pig that you can f*ck around with.
One neat thing with Vonage is you can get a phone # with a non-local area code. So if you're in CA and have relatives in NY, you can get a 212/718/914 number. This, of course, means their calls you you are local calls for them. It's $4.95/mo for you, tho.
Coolness!
I just got my service activated today, so I've been poking around and seeing what's what. So far, it seems pretty good.
If you look at the picture, it looks as if there's a bit of a dip from where the camera is to the next span.
Did they screw up the alignment a bit then hack it to get it to work? How do they do that anyway?
Well, think about it: they can, with the amount of money they have in the bank, sell an iPod at a massive loss forever.
Current liquid assets: $59b
iPod cost: $250
free mPods: 236 million
That's more than the total number of mp3 players sold. How do you compete with that?
There should be a way to stomp on Microsoft by using anti-dumping laws or something like that.
Dell is a distribution channel, plain and simple. Its costs are minimized, there's no R&D, etc. Unlike Wal-Mart, Dell rebrands many of the items that it sells. Wal-Mart does too, under private labels, but for some reason Dell has been able to create "Dell" as a brand, unlike most white-box builders.
Dell now sells a ridiculously large amount of computer equipment: $11 Billion last quarter, $44 Billion on an annualized basis. They sold as much stuff as Microsoft last quarter, and they made 50% less. They've cut the monopoly premium to 50%, with margins of about 23%.
Plus, there's no reason to think they're going to stop anytime soon. They are the low-cost provider, period.
New technology? Probably not. But they sure are a cheap place to get boxes.
I wonder if that data is accessible by 3rd parties. You could make "mail maps" that let you visualize the clustering of your incoming messages, and you could actually see the spam...by looking at the outliers and noise.
In fact, you could do this with any large data set. How about the feds looking for anomalous chunks of data in the bitstream? Anomalous stuff would just pop out, literally. This would make the TSA's job much, much easier. How about that?
Dude, don't you know anything about the stock market? Fundamentals make no difference. See:
3 36 /
http://www.blogmaverick.com/
for a real-world view of the market. He says it better than I could. Scroll down to the "Why do people fall in love with stocks" entry. Here's the direct link:
http://www.blogmaverick.com/entry/9817136470302
Enjoy and learn!
I just realized - he's doing a bayesian average!
http://www.research.att.com/~volinsky/bma.html
There was a possibly apochryphal story about this. A plane went down in a large area, and they needed to find it. Nobody really knew what happened to it. The leader of the search team went and asked a bunch of pilots where they thought the plane was, after giving them the course, heading, speed, and whatever data was available.
Well, they took the answers, narrowed the search area, then found the plane pretty much where the consensus said it would be.
A bit of thought will give you the reason this might have worked...
One thing I've noticed about the posters is they're not really sure why the dutch auction format is different than the normal IPO process.
In a normal IPO, the investment bank (or brokerage) sets the price, say, $20, then buys the shares off of the issuing company for $20. Any spread off of that goes directly to the brokerage (or investment bank).
This is a Bad Deal for the issuing company, because the spread should have gone to the company, not to the brokerage. Think about it: if MSDW priced a stock at $20, but were able to sell at $200, then the issuing company just lost $180! The brokerage, of course, made $180.
With the dutch auction, there will (or should be) little or no upside to google's stock because the price will be determined by how much people are willing to pay up front. In fact, it will probably drop on the first day, since people confused by this concept will try and flip, leading to a sell imbalance.
Depending on the shares that are to be issued, I'd say that conservatively google's stock should open at anywhere from $175 +- 8%, and drop about 15%. Why $175? Because -everyone- knows google, and companies that are well-known to the public trade at a premium.
The other interesting thing about google is that a savvy investor/hacker could manipulate the earnings by writing a virus that sneakily clicked on all those google ads that appear when you do a google search. Voila, instant earnings! With enough spread, you could do millions of hits a day, which translates to millions of dollars a day (thank you, adwords!)
We'll see.
There's an old quote that says it better:
"It's not that the dog talks well, it's that it talks at all."
This is the problem with a lot of open source: people are happy that it talks at all. Maybe someday they'll get around to talking well?
Unix -> gcc, lynx, ncftpd, ncftp, mkisofs, gzip, tcpwrappers, sshd, apache, gnu make, top
A lot of these are there by default now, at least on Solaris and AIX. But I still prefer to build my own, just so I know what the opetions are.
win32 -> Winzip, ntp client, pfe
That's about it for windows. I don't use it for day-to-day stuff, so Office etc isn't an issue.
When people talk about support, they always say "hey, you can always call MS."
But have you? Do you? When a problem occurs, the go-to guy is the IT guy in the company. And that guy (or gal) either searches the net or asks a friend.
Have you, and IT person, ever called the MS helpline? If so, were you able to get an answer?
Hey, remember when the iPod mini was announced?
"It's too expensive, nobody'll buy it."
"Only 1000 songs? How useless! Nobody'll buy it."
It just goes to show, never listen to slashdot when it comes to learning what the market wants.
This is the problem with science in the context of public policy, and why that statement about the Bush administration's science policy is a bit out of whack.
Almost by definition, anything that recommends a solution is bad science. Science isn't very good at outcomes, but that's what politicians need.
In the case of global warming, it's difficult because the costs are imposed now, and the outcome is always in doubt. If we do X, there's no guarantee that X will happen. So are you willing to spend hundreds of billions or trillions of dollars and affect every industry for possibly no gain? Nope.
Science doesn't determine goals, direction, and priorities - politicians (and the public) do. And that's how it should be. Scientists don't pay a price if they're wrong.
Hey everyone, it's easy to make sure you don't get outsourced: provide value to your business.
No part of the business that is deemed "essential" would be outsourced. This means that IT is not considered essential, and it's not contributing to the business as a whole.
This is true of most IT - how much value-add do you bring? Do you actually help make the business better, or do you sit around talking about Lusers and how dumb they are? How those business people are morons? How they're so stupid they can't even turn on their PCs?
Are you a BOFH? Then you'll become an unemployed BOFH, and a happy worker someone else will do a better job than you.
Are you actively involved with your business groups, and understand how you help them make money? Then you won't get outsourced. Period.
Outsourcing is the business striking back at the geeks. The geeks have held sway for too long, basically removing value from businesses by being totally unresponsive to business needs. And if you can't get what you want at home, you go somewhere else. Its' that simple.
From what I've heard, it's a great sales automation tool. For CRM, though, it "needs a lot of maturing."
More FUD, more FUD, more FUD.
This has been bandied about so much that it's taken the place of real, critical thought.
Think about this: when Apple attempted to license its OS, it sales were immediately cannibalized by its licensees.
When Palm did the same thing, the same thing happened.
Name one company besides Microsoft that's succeeded financially in licensing its stuff to others. I can think of one, but they were already in a dominant market position and were/are licensing to stave off anti-trust issues.
Go ahead, find examples and post them.
Now everyone knows why Microsoft was sitting on all that cash: they're gonna buy their way outta trouble. Why not? It's easier for them to buy their way out of trouble.
Andd after all that, they'll still have billions and billions lying around to cross-subsidize their money-losing ventures. Those money-losing ventures, of course, include almost everything Microsoft does except Windows and Office.
Microsoft's new slogan should be "Innovating financial solutions to legal problems."
Funny, I was just thinking today about all those people complaining about Apple's 1-button mouse usually are the same ones that use the 102-button mouse on a daily basis. They'll never be happy with less.
It's expensive, but it works - at least on the Mac side. Not sure, but it looks like the Windows version grew a brain.
EP on the Mac is very nice, it looks good and lets you think less so you can think better.
As a metric goes, this one has held up pretty well for me over the years. Back in the day, if a feature took around one day to code it took around 4 days of UI work to get a UI that was decent. I'd never claim my UIs were good, but they got the job done and were clear and straight-forward.
That includes testing the UI, BTW. Not user testing, just making sure the UI actually worked.
That five-to-one ratio is an average. For simple CLIs the ratio is less (there's no UI, but there's a usage statement) and for complicated features the ratio is more.
"I am reminded at this point of a fellow I used to know whose name was Henry, only to give you an idea of what a individualist he was, he spelled it H-E-N-3-R-Y. The three was silent, you see."
- Tom Lehrer, from "An Evening (Wasted) with Tom Lehrer, 1959