Any discussion of biometrics without discussing the crossover rate (or Equal Error Rate) is woefully incomplete. see this explanation: https://security.stackexchange.com/questions/57589/determining-the-accuracy-of-a-biometric-system The crossover rate is that point in the sensitivity settings of the system that yield minimum errors, where the False Acceptance Rate = the False Rejection Rate. In layman's terms, you're letting in unauthorized bad guys at the same rate you're keeping authorized good guys out. Any biometric system that doesn't list their crossover rate is pure snakeoil. Run away. Another data point few consider. A Large Theme Park used biometrics a few years back for their annual ticket holders. It soon became known as the "identical twins two-for-one sale". Can your biometrics discern identical twins? Few can.
One of my business partners took this stuff about two years ago and completely lost her sense of smell within 30 minutes. Yes, it's dangerous. Yes, it needs to be off the market. Yes, the manufacturer has known of the dangers for years, having been sued numerous times. Yes, they should be held accountable.
I did a presentation at the Atlanta Unix Users' Group this month that is a more up-to-date comparison. It's available in Open Office format. You can also get to it from my home page. I did a similar talk almost four years ago. My conclusion is that MySQL has closed the feature gap with PostgreSQL in recent years. I still give PostgreSQL the edge in features, and MySQL the edge in out-of-the-box untuned performance. I also discuss replication and clustering.
Ok, these 'market analysts' look at spreadsheets of market shares, etc. Look at the technology under the hood: Microsoft uses all Windows products. Yahoo uses BSD and PHP as their environment. I'm sure Gates and company would LOVE to be running such a large, critical portion of their business on OSS! Or throw all Yahoo's code away and re-write in.NET? Right!! From a platform point of view, anybody who thinks about this for more than 30 seconds will see that this is a non-starter. Nothing here. Move along.
I've been in gateway stores a time or two. They were almost completely useless. If you wanted a price on particular configuration, the sales droids directed you to an internet-connected machine where you could go to the gateway web site and get a price quote. Duh... I could have done that from home.
When I asked them about Linux, or a machine without an O/S, I was told "Microsoft won't let us do that." Double Duh.
In short, it was about the most useless showroom experience I've had in a long time. Gateway will lose nothing by shutting these turkeys (cows?) down.
Reference: Jerome H. Saltzer, and Michael D. Schroeder. The Protection of Information in Computer Systems. (invited tutorial paper) Proceedings of the IEEE 63, 9 (September 1975) pages 1278-1308. Reprinted in David D. Clark and David D. Redell, editors. Protection of Information in Computer Systems. IEEE 1975 CompCon tutorial. IEEE # 75CH1050-4. Also reprinted in Rein Turn, editor. Advances in Computer System Security. ArTech House, Dedham, MA, 1981, pages 105-135. ISBN 0-89006-096-7 Also reprinted in Marvin S. Levin, Steven B. Lipner, and Paul A. Karger. Protecting Data & Information: A Workshop in Computer & Data Security. Digital Equipment Corporation, 1982. This paper was originally prepared off-line. In 1997, Norman Hardy kindly rendered it into World-Wide Web form.
here
I'll agree on Tex. I remember the day I gave up on it. I attended a lecture by Knuth himself on abstract graph theory. Guess what he used to generate his overhead transparencies with? Colored felt-tipped markers. Here is the great Knuth himself, the creator of TeX with near-infinite computing resouces available, and he hand-draws equations with felt-tipped markers!! At that moment, I knew TeX was dead.
In general I found that virtual domains were a bit trickier to set up in postfix than in sendmail. Ordinary aliases were just as easy (read identical). My sites don't do enough volume to tell any difference in performance. The build/install process was probably a bit easier for postfix, i.e. didn't have to monkey around with M4. So as a sendmail admin of more years than I care to think about, postfix seems about as easy to administer as sendmail on a day-to-day basis.
Do the math, SCO will be bankrupt in December
on
SCO Execs Dumping Stock
·
· Score: 3, Insightful
According to the story, SCO burned through $14 million last fiscal year. They had $10 million in the bank in April. This is about 8 1/2 months worth of fuel. This is assuming that this year is the same as last in terms of revenue (probably going down) and expenses (probably going up). So 8 1/2 months from April puts them running on fumes in December. I doubt this saga will drag on for years.
Check your history. Guess who funded most of the BSD development? Right. The US Government. Who funded development of TCP/IP? Right again. Are these open source? Yes. Were they funded by Government for the Common Good? Yes. This is nothing new. This has been going on for a couple of decades now.
The Air and Space Museum is part of the Smithsonian. You walk in the main entrance and see the Wright Brothers flier, the Spirit of St. Louis, the X1, the Voyager, etc. all in one spot. It'll take your breath away. BTW, you can also touch a moon rock right there. An absolutely amazing place. My three sons and I can spend hours and hours there. If you get tired, walk across the mall to see the Museum of Natural History and look at the Hope Diamond. Best of all, all these places are FREE! That's right, just walk right in.
For those interested in historical perspectives on airline accidents and the ongoing debate on the risks of electronics, see the archives of the moderated netnews groups comp.risks. It's moderated by Peter G. Neumann at sri. One can get the archive at such places as
Google Groups
Not only in-flight risks, but all types are discussed here. It's one of the more lucid discussions on the net. I've been following this newsgroup for the better part of 20 years.
funding via hardware/software symbiosis
on
Funding Open Source?
·
· Score: 5, Informative
Back in my audio engineering days, there was a company that made transformers (hardware). In order to get people to buy their transformers, they gave away schematics (software) of how to build an insanely great preamp using their hardware.
Several Open Source projects are nicely funded doing the same thing. Take for instance the OSS telephone project Asterisk. The software is made available to enable more people to buy and use a particular telephone line interface card. Other cards are supported in the software, but the sponsoring company's is obviously supported first.
So, one avenue is to partner with a hardware maker, in the case of the PC to PDA sync, partner with an up-and-coming desktop hardware manufacturer, or a similar PDA maker.
The case against patents
on
Steal This Idea
·
· Score: 5, Informative
An older reference to patents in general can be found at Don Lancaster's site Tinaja.com. There's a pdf of the original paper, and some e-book links. Don's been an active author in the technology world for several decades. His site has some other amusing opinion pieces as well.
Enjoy!
in the old days, all flight attendants had to be registered nurses. Even recently, hiring preference was given to medically-trained applicants. Is this still the case? Does anyone have the statistics of what percentage of flight attendants have formal medical training (LPN, RN, etc)? From that we could figure the probability that one of the attendants is already capable of accurately checking vital signs, or using one of these telemedicne devices.
We need more information about your product. Is your product aimed directly at developers (i.e. Kylix), or at the mass market (i.e. Palm). I gather by reading between the lines, it may be mass-market. If it's a mass-market product, you want developers to develop third-party add-ons (aftermarket products). The main way to do this is to show them the money. How many units are you planning to ship in the first 6/12/24 months? Are you going to allow developers to co-market to your customer base? How much will the product cost? This drives how much one can charge for add-ons (never more that the product). All the neat development tools in the world won't draw developers unless you can show them how to make money by writing for your platform. Similarly, if the market is huge, developers will be willing to put up with less-than-optimal tools just to get a piece of the pie. You might also want to consider a forum for third-party developers to learn if somebody's already working on an idea similar to theirs. Of course you should treat your developers to the next release of the platform software well in advance of general availability so they can get their products working with it. Oracle charges a nominal fee to be part of their developer's network to cover the cost of quarterly software updates.
Slashdot Mirror
Any discussion of biometrics without discussing the crossover rate (or Equal Error Rate) is woefully incomplete. see this explanation: https://security.stackexchange.com/questions/57589/determining-the-accuracy-of-a-biometric-system
The crossover rate is that point in the sensitivity settings of the system that yield minimum errors, where the False Acceptance Rate = the False Rejection Rate. In layman's terms, you're letting in unauthorized bad guys at the same rate you're keeping authorized good guys out. Any biometric system that doesn't list their crossover rate is pure snakeoil. Run away.
Another data point few consider. A Large Theme Park used biometrics a few years back for their annual ticket holders. It soon became known as the "identical twins two-for-one sale". Can your biometrics discern identical twins? Few can.
A friend of mine used to work there until yesterday. He said they closed the place down, more like 95% layoffs. It's gone.
Could somebody post the original article that this post summarizes? e.g. Where can we get further information?
One of my business partners took this stuff about two years ago and completely lost her sense of smell within 30 minutes. Yes, it's dangerous. Yes, it needs to be off the market. Yes, the manufacturer has known of the dangers for years, having been sued numerous times. Yes, they should be held accountable.
I did a presentation at the Atlanta Unix Users' Group this month that is a more up-to-date comparison. It's available in Open Office format. You can also get to it from my home page. I did a similar talk almost four years ago. My conclusion is that MySQL has closed the feature gap with PostgreSQL in recent years. I still give PostgreSQL the edge in features, and MySQL the edge in out-of-the-box untuned performance. I also discuss replication and clustering.
Ok, these 'market analysts' look at spreadsheets of market shares, etc. Look at the technology under the hood: Microsoft uses all Windows products. Yahoo uses BSD and PHP as their environment. I'm sure Gates and company would LOVE to be running such a large, critical portion of their business on OSS! Or throw all Yahoo's code away and re-write in .NET? Right!! From a platform point of view, anybody who thinks about this for more than 30 seconds will see that this is a non-starter. Nothing here. Move along.
If I recall my log scales correctly, "two times quieter" would equal 3db quieter. Not exactly revolutionary, but a step in the right direction.
When I asked them about Linux, or a machine without an O/S, I was told "Microsoft won't let us do that." Double Duh.
In short, it was about the most useless showroom experience I've had in a long time. Gateway will lose nothing by shutting these turkeys (cows?) down.
Reference: Jerome H. Saltzer, and Michael D. Schroeder. The Protection of Information in Computer Systems. (invited tutorial paper) Proceedings of the IEEE 63, 9 (September 1975) pages 1278-1308. Reprinted in David D. Clark and David D. Redell, editors. Protection of Information in Computer Systems. IEEE 1975 CompCon tutorial. IEEE # 75CH1050-4. Also reprinted in Rein Turn, editor. Advances in Computer System Security. ArTech House, Dedham, MA, 1981, pages 105-135. ISBN 0-89006-096-7 Also reprinted in Marvin S. Levin, Steven B. Lipner, and Paul A. Karger. Protecting Data & Information: A Workshop in Computer & Data Security. Digital Equipment Corporation, 1982. This paper was originally prepared off-line. In 1997, Norman Hardy kindly rendered it into World-Wide Web form. here
I'll agree on Tex. I remember the day I gave up on it. I attended a lecture by Knuth himself on abstract graph theory. Guess what he used to generate his overhead transparencies with? Colored felt-tipped markers. Here is the great Knuth himself, the creator of TeX with near-infinite computing resouces available, and he hand-draws equations with felt-tipped markers!! At that moment, I knew TeX was dead.
In general I found that virtual domains were a bit trickier to set up in postfix than in sendmail. Ordinary aliases were just as easy (read identical). My sites don't do enough volume to tell any difference in performance. The build/install process was probably a bit easier for postfix, i.e. didn't have to monkey around with M4. So as a sendmail admin of more years than I care to think about, postfix seems about as easy to administer as sendmail on a day-to-day basis.
According to the story, SCO burned through $14 million last fiscal year. They had $10 million in the bank in April. This is about 8 1/2 months worth of fuel. This is assuming that this year is the same as last in terms of revenue (probably going down) and expenses (probably going up). So 8 1/2 months from April puts them running on fumes in December. I doubt this saga will drag on for years.
Check your history. Guess who funded most of the BSD development? Right. The US Government. Who funded development of TCP/IP? Right again. Are these open source? Yes. Were they funded by Government for the Common Good? Yes. This is nothing new. This has been going on for a couple of decades now.
The Air and Space Museum is part of the Smithsonian. You walk in the main entrance and see the Wright Brothers flier, the Spirit of St. Louis, the X1, the Voyager, etc. all in one spot. It'll take your breath away. BTW, you can also touch a moon rock right there. An absolutely amazing place. My three sons and I can spend hours and hours there. If you get tired, walk across the mall to see the Museum of Natural History and look at the Hope Diamond. Best of all, all these places are FREE! That's right, just walk right in.
Not only in-flight risks, but all types are discussed here. It's one of the more lucid discussions on the net. I've been following this newsgroup for the better part of 20 years.
Several Open Source projects are nicely funded doing the same thing. Take for instance the OSS telephone project Asterisk. The software is made available to enable more people to buy and use a particular telephone line interface card. Other cards are supported in the software, but the sponsoring company's is obviously supported first.
So, one avenue is to partner with a hardware maker, in the case of the PC to PDA sync, partner with an up-and-coming desktop hardware manufacturer, or a similar PDA maker.
Yes, but what's the bandwidth of a minivan full of CDROMs? I get 235 Mb/sec. Enjoy.
An older reference to patents in general can be found at Don Lancaster's site Tinaja.com. There's a pdf of the original paper, and some e-book links. Don's been an active author in the technology world for several decades. His site has some other amusing opinion pieces as well. Enjoy!
in the old days, all flight attendants had to be registered nurses. Even recently, hiring preference was given to medically-trained applicants. Is this still the case? Does anyone have the statistics of what percentage of flight attendants have formal medical training (LPN, RN, etc)? From that we could figure the probability that one of the attendants is already capable of accurately checking vital signs, or using one of these telemedicne devices.
We need more information about your product. Is your product aimed directly at developers (i.e. Kylix), or at the mass market (i.e. Palm). I gather by reading between the lines, it may be mass-market. If it's a mass-market product, you want developers to develop third-party add-ons (aftermarket products). The main way to do this is to show them the money. How many units are you planning to ship in the first 6/12/24 months? Are you going to allow developers to co-market to your customer base? How much will the product cost? This drives how much one can charge for add-ons (never more that the product). All the neat development tools in the world won't draw developers unless you can show them how to make money by writing for your platform. Similarly, if the market is huge, developers will be willing to put up with less-than-optimal tools just to get a piece of the pie. You might also want to consider a forum for third-party developers to learn if somebody's already working on an idea similar to theirs. Of course you should treat your developers to the next release of the platform software well in advance of general availability so they can get their products working with it. Oracle charges a nominal fee to be part of their developer's network to cover the cost of quarterly software updates.