As a professional photographer, I have to chime in here.
Just because Canon can put out sensors with higher density doesn't mean it does a bit of good. Even the high-end SLR lenses from Canon and Nikon don't have the autofocus accuracy or resolving power to get per-pixel sharpness on dense sensors. Slight camera shake (even with stabilization) translates to slight blur. To get sharp 1:1 crops on hi-res sensors, you need a big camera with a big sensor and a big lens on a tripod.
The real frustrations for professionals and consumers alike are:
Noise in low light
Dependable automatic white balance
Wide dynamic range (aka HDR)
Low light performance on full-frame SLRs is really getting good, and has pretty much exceeded film. White balance accuracy is pretty much staying even. The dynamic range of digital is still nowhere near film, but improving slowly.
Getting to 24-bit/pixel dynamic range will pretty much take care of the other two problems. The current best for DSLRs is 14-bit, and on (tens of thousands of dollar) medium format, it's 16-bit (so far as I can tell). Those medium format cameras are at roughly the same sensor pixel densities as the 21mp 1Ds Mark III, probably because pixel density is one of the biggest factors that contributes to noise.
This would be relevant if Episode 1 was an expansion, but it's more like a sequel. From the Steam store:
Episode One is the first in a series of games that reveal the aftermath of Half-Life 2 and launch a journey beyond City 17. Also features two multiplayer games. Half-Life 2 not required.
This argument is repeated time and again here on Slashdot and the fact is it is rediculous. Want to know why? Because Novell's customers want it. In fact, they want Suse Linux to run on whatever white-box thrown-together-component list they decide, and having vendors supply drivers to reach that goal makes Novell a more attractive company.
The argument isn't against vendors supplying drivers, it's against binary-only drivers. To be honest, I don't like compiling drivers every time I upgrade my kernel, but at least I have the option with a source driver. With binary drivers, I have to wait for the vendor to supply an update. If the hardware is old, that might never happen.
I've never understood why ALL drivers aren't distributed as source. Hardware vendors don't make a dime off of driver updates, so why should they care if the driver code is open source?
I used to think that Red Hat was Linux too (from a commercial/marketing perspective, that is). But now I'm starting to think that it's not someone like RedHat (or even Fedora) who are going to make Linux mainstream.
What's (arguably) the most popular and widely known open-source app out there? Firefox. Is that because of big corporate backing? Nope (well, corporate money, but not marketing or support).
Just yesterday I looked at Ubuntu seriously for the first time, and I'm amazed. It looks like the Firefox of linux distros. User-oriented, simple, and with no tech-speak on the web site. I actually had to hunt for information on what package manager it uses.
Two asides: 1) I think that the first distro that can be consumer-friendly like Ubuntu and capitalize on the exploding computer-lifestyle (social networking, blogs, messaging, photos, television/movies) craze will pull a whole lot of users. 2) The only thing that I think will hold Ubuntu back is the name. Despite its good intentions, "Ubuntu" is not an (American) consumer-friendly name.
I used to work in ET (Educational Technology) for a public university in Texas. We used a major commercial course management system. The software was certainly pre-packaged, but given that it's not the only application we used on campus (we had registration systems, authentication systems, assessment software, but also blogs, videoconferencing, gradebooks), we inevitably had to do some custom configuration. I'd be willing to bet that any university IT department worth its salt strives towards at least *some* level of integration between its apps, if for no other reason to keep the students from calling their intstructors with confusions, and in turn having the instructors call IT. We also didn't want to have to upload a data file every day to populate the course lists and rosters.
We had good sysadmins and custom-app staff around anyway, so we used them to adjust the system to do what we needed. We repeatedly came upon "unsupported" customization needs. The APIs didn't do what we wanted. The version of the system that we licensed wasn't extensible enough.
That system cost us somewhere in the neighborhood of $65,000 a year. The one with a more flexible extension API cost (IIRC) about twice that. And it still didn't do what we wanted out of the box, nor did the APIs exist to do some of the advanced integration we wanted to do.
We stopped, took a look around, and found the open-source Sakai Project (http://www.sakaiproject.org./ We took our $130K per year license fee and hired a programmer with half of it, paid $10K/year for project partnership (revision control & forums access, plus project input), and after a year of development (or so I hear, I've since left), had a system that does exactly what we needed, was more flexible, had a better support structure (talk to the developer directly!), and cost about half as much to run each year. Plus, we had a good programmer on staff who could work on other projects too.
On a related note, http://www.a-hec.org/open%20source%20030106%20clar ifications.html explains who the study funding came from (Sun, Unicon, and SCT). They're all course management systems vendors, and one of them actually distributes/supports Sakai commercially. It also mentions the BlackBoard/WebCT merger (the two most-important course management software companies) specifically as a compelling reason that universities are considering open source.
I also can't help but point out - Open Source basically started in universities, and the internet is based on that software. If it isn't ready for prime time, it's a little too late to turn back.
There's a lot of talk of adding an upper curve on the related stories, but it makes more sense to me to have them look like the main headlines: White text on a green background. That way they just look like collapsed story boxes, and stand on their own. Using any other coloring will make them look like they are part of the story box above, as footnotes of some sort.
In general, I think there are advantages to taking courses like RedHat's, as well as a do-it-yourself approach:
Instructor-led: Pro: There's someone there to answer your questions. Pro: You can dedicate yourself to the material without interruption. Pro: You won't be led off on tangential tasks. Pro: You'll have something to put on your resume. Con: You might be overly isolated from the nuts-and-bolts
Do-it-yourself: Pro: You'll learn where to go to find the answers on your own. Pro: You can focus on topics that are important to you. Pro: You're exposed to the full breadth or possilities. Con: You'll have to figure out on your own what's signal and what's noise.
FWIW, I've looked into the RedHat courses for some of my employees, and I think they're fairly good for someone who wants to learn useful Linux (for business) administration. The topics are a good mix of basics and enterprise-level stuff.
To help you with the self-taught portion of your learning experience, I highly recommend Linux Administration Handbook. It will lead you from low-level (booting, etc.) to high-level (web servers, backups), and it's concise enough to use as a reference later on.
I know this comment won't even get read, but I just have to add another ME TOO in hopes that this will be one of the questions answered. I've discussed with my RH sales rep the fact that Windows Server is cheaper after edu discounts, and the desktop Windows (Windows XP, for now) is included in the Campus Agreement. He offered a discount on 11 or more copies of RHEL, but only off of the "Standard" version.
I'm sure many sysadmins don't find themselves calling for support (even good tech support is a hassle) - what we really need is updates and stability.
What I would love: 1. Deep EDU discounts on RHEL ES and WS 2. Perhaps a site-license style agreement for these OSs if we run our own Up2Date server (thus consuming less RH resources) 3. Pay per-incident support or an affordable service-level agreement 4. A program that offers discounted or expedited support when the school funnels support requests through an on-staff RHCE (Dell offers something like this for hardware support)
In my experience, most (not all, of course) Open Source projects aren't concerned with backward compatibility outside of the scope of the project itself. Regression testing in OSS is folded into the bug testing.
That's one of the downsides of OSS. The biggest example: When you upgrade your libc, you have to recompile all of your dependent apps. One thing Windows and Solaris have going for them is that you can run a 5 year old binary on the newest version of the OS and it will almost always work.
It's a big burden on the development team to provide support for old interfaces, but this sort of thing is where OSS has a long way to go. It gets really expensive for individual persons/companies to support (bugfix, etc.) packages that are a few revisions old.
This is waaaay too complicated. Your social security card should have two numbers on it:
An identifier (000-00-0000) and An authenticator (AAA-AA-AAAA)
The identifier can be used to uniquely identify you (until we reach a population of 1,000,001), and the authenticator can be used to authenticate your identity. Provide a public system that can be used to authenticate identifiers (perhaps something similar to what credit card networks use and well-logged/monitored for abuse). Banks, creditors, or even your university could access the system when appropriate. Make it illegal to store authenticators. Provide a system to allow you to (perhaps for a small fee) change your authenticator when your card gets stolen.
This is, after all, a proven system that every slashdot reader uses regularly - good ol' username and password. And most people have already become accustomed to things requiring one, so it shouldn't be a difficult thing for the public to use.
(and I even previewed it. Try again)
An odd coincidence: this paper was posted to the NANOG (North American Network Operators' Group) mailing list just today. The first paragraph reads:
This document defines a Best Current Practice to minimize pollution of the Internet by various types of abuse, using the community's own measures in the absence of effective legal, regulatory and technical measures.
An odd coincidence: this was posted to the NANOG (North American Network Operators' Group) mailing list just today.
The first paragraph reads:
This document defines a Best Current Practice to minimize pollution of
the Internet by various types of abuse, using the community's own
measures in the absence of effective legal, regulatory and technical
measures.
First, I have to express my sincerest thanks for such a great show. You've been more help to my cooking than any other person.
What I want to know is if you've ever been approached by another network/company like PBS? The reason I ask is that of all the shows on Food TV, you're one of the only ones who really TEACHES cooking. Emeril's on every 3rd or 4th timeslot, and he cooks things that most people won't ever make and learn very little from. The closest thing I've seen recently to your caliber of program is Rick Bayless' "Mexico, one plate at a time" (on PBS) - one with some background on the process and history of the food we eat, and why it's prepared the way it is.
I like your show and I'd love to see you get more exposure.
Sounds to me like you've got some other problem. You might want to consider investing in an UPS, or at least a power conditioner, if you don't already have one. Or perhaps your power supply is to blame, if you've lost all of these drive on the same one/brand.
At a company I used to work for, we used to lose drives like crazy (usually a max life of 1 year). Once our vendor changed power-supply brands, we never had problems again.
The truth helps. Just keep speaking the truth, and tell your friends, people on the bus, folks at work.
"Um, Mr. Bus Driver sir, there's a madman at the back of the bus who won't quit talking about how 'The zero-day virus' is a government conspiracy to make us give up our rights. He sounds like one of those 'hackers' you always hear about...you know like on that movie, 'Hackers'?"
Not to steal Alton Brown's thunder (I own his book, it's awesome), but can anyone suggest any similar books for those of us who just can't justify culinary school, but want to become great amateur chefs? I've searched, but the cost and age of most of the books I've found make me want an opinion before buying.
(perhaps I'm offtopic, but please mod up responses)
Configuration files were made to be edited by hand! This is why Linux is so popular, flexibility.
I would hardly say that most people want to sit and jack around with config files
Creating a more organized and standardized way of formatting config files (anyone heard on something called "XML"?) would help programmers, users, *and* lusers.
Imagine a human-readable, programatically-ediable, self-documenting config file. Might make installers and package managers a little easier to write.
If someone is determined enough, they can get on your WLAN. MAC addresses can be spoofed, WEP keys can be sniffed. All you can do is authenticate and log.
I recently spoke to some keen fellows from Baylor University that have created an OpenBSD-based firewall/logging/authentication system that takes the poster's info page one step further. Everyone authenticates via an SSL-encrypted web site in order to join the network. DHCP leases are handed out in conjunction with a login session, so you can track who does what. Logging in also opens up your firewall to allow the newly-leased IP address through.
Your finger will touch the letter keys nearby, but that's OK. That's how Fastap technology works... so you don't need to be careful.
Like most new human interfaces, this will surely take some getting used to, but is this really good ergonomic design? How can you consistently avoid registering a press on nearby buttons? Even if the device is smart enough to tell the difference between intentional and accidental key-presses, every other device out there requires the user to consistently and squarely press each button involved in an input operation. People are used to that, and it's not likely to change.
I don't know about anyone else, but it's gonna be really difficult for me to feel comfortable using an interface where fat-fingering the keypad is normal.
Slashdot Mirror
As a professional photographer, I have to chime in here.
Just because Canon can put out sensors with higher density doesn't mean it does a bit of good. Even the high-end SLR lenses from Canon and Nikon don't have the autofocus accuracy or resolving power to get per-pixel sharpness on dense sensors. Slight camera shake (even with stabilization) translates to slight blur. To get sharp 1:1 crops on hi-res sensors, you need a big camera with a big sensor and a big lens on a tripod.
The real frustrations for professionals and consumers alike are:
Low light performance on full-frame SLRs is really getting good, and has pretty much exceeded film. White balance accuracy is pretty much staying even. The dynamic range of digital is still nowhere near film, but improving slowly.
Getting to 24-bit/pixel dynamic range will pretty much take care of the other two problems. The current best for DSLRs is 14-bit, and on (tens of thousands of dollar) medium format, it's 16-bit (so far as I can tell). Those medium format cameras are at roughly the same sensor pixel densities as the 21mp 1Ds Mark III, probably because pixel density is one of the biggest factors that contributes to noise.
This would be relevant if Episode 1 was an expansion, but it's more like a sequel. From the Steam store:
This argument is repeated time and again here on Slashdot and the fact is it is rediculous. Want to know why? Because Novell's customers want it. In fact, they want Suse Linux to run on whatever white-box thrown-together-component list they decide, and having vendors supply drivers to reach that goal makes Novell a more attractive company.
The argument isn't against vendors supplying drivers, it's against binary-only drivers. To be honest, I don't like compiling drivers every time I upgrade my kernel, but at least I have the option with a source driver. With binary drivers, I have to wait for the vendor to supply an update. If the hardware is old, that might never happen.
I've never understood why ALL drivers aren't distributed as source. Hardware vendors don't make a dime off of driver updates, so why should they care if the driver code is open source?
I used to think that Red Hat was Linux too (from a commercial/marketing perspective, that is). But now I'm starting to think that it's not someone like RedHat (or even Fedora) who are going to make Linux mainstream.
What's (arguably) the most popular and widely known open-source app out there? Firefox. Is that because of big corporate backing? Nope (well, corporate money, but not marketing or support).
Just yesterday I looked at Ubuntu seriously for the first time, and I'm amazed. It looks like the Firefox of linux distros. User-oriented, simple, and with no tech-speak on the web site. I actually had to hunt for information on what package manager it uses.
Two asides: 1) I think that the first distro that can be consumer-friendly like Ubuntu and capitalize on the exploding computer-lifestyle (social networking, blogs, messaging, photos, television/movies) craze will pull a whole lot of users. 2) The only thing that I think will hold Ubuntu back is the name. Despite its good intentions, "Ubuntu" is not an (American) consumer-friendly name.
If you RTFA, a Mozilla board member says that the quoted figure of $72 million is too high
Uh, actually if you RTFA, he says that $72 million is not correct. They could be making considerably more than that.
I used to work in ET (Educational Technology) for a public university in Texas. We used a major commercial course management system. The software was certainly pre-packaged, but given that it's not the only application we used on campus (we had registration systems, authentication systems, assessment software, but also blogs, videoconferencing, gradebooks), we inevitably had to do some custom configuration. I'd be willing to bet that any university IT department worth its salt strives towards at least *some* level of integration between its apps, if for no other reason to keep the students from calling their intstructors with confusions, and in turn having the instructors call IT. We also didn't want to have to upload a data file every day to populate the course lists and rosters.
r ifications.html explains who the study funding came from (Sun, Unicon, and SCT). They're all course management systems vendors, and one of them actually distributes/supports Sakai commercially. It also mentions the BlackBoard/WebCT merger (the two most-important course management software companies) specifically as a compelling reason that universities are considering open source.
We had good sysadmins and custom-app staff around anyway, so we used them to adjust the system to do what we needed. We repeatedly came upon "unsupported" customization needs. The APIs didn't do what we wanted. The version of the system that we licensed wasn't extensible enough.
That system cost us somewhere in the neighborhood of $65,000 a year. The one with a more flexible extension API cost (IIRC) about twice that. And it still didn't do what we wanted out of the box, nor did the APIs exist to do some of the advanced integration we wanted to do.
We stopped, took a look around, and found the open-source Sakai Project (http://www.sakaiproject.org./ We took our $130K per year license fee and hired a programmer with half of it, paid $10K/year for project partnership (revision control & forums access, plus project input), and after a year of development (or so I hear, I've since left), had a system that does exactly what we needed, was more flexible, had a better support structure (talk to the developer directly!), and cost about half as much to run each year. Plus, we had a good programmer on staff who could work on other projects too.
On a related note, http://www.a-hec.org/open%20source%20030106%20cla
I also can't help but point out - Open Source basically started in universities, and the internet is based on that software. If it isn't ready for prime time, it's a little too late to turn back.
There's a lot of talk of adding an upper curve on the related stories, but it makes more sense to me to have them look like the main headlines: White text on a green background. That way they just look like collapsed story boxes, and stand on their own. Using any other coloring will make them look like they are part of the story box above, as footnotes of some sort.
In general, I think there are advantages to taking courses like RedHat's, as well as a do-it-yourself approach:
Instructor-led:
Pro: There's someone there to answer your questions.
Pro: You can dedicate yourself to the material without interruption.
Pro: You won't be led off on tangential tasks.
Pro: You'll have something to put on your resume.
Con: You might be overly isolated from the nuts-and-bolts
Do-it-yourself:
Pro: You'll learn where to go to find the answers on your own.
Pro: You can focus on topics that are important to you.
Pro: You're exposed to the full breadth or possilities.
Con: You'll have to figure out on your own what's signal and what's noise.
FWIW, I've looked into the RedHat courses for some of my employees, and I think they're fairly good for someone who wants to learn useful Linux (for business) administration. The topics are a good mix of basics and enterprise-level stuff.
To help you with the self-taught portion of your learning experience, I highly recommend Linux Administration Handbook. It will lead you from low-level (booting, etc.) to high-level (web servers, backups), and it's concise enough to use as a reference later on.
I know this comment won't even get read, but I just have to add another ME TOO in hopes that this will be one of the questions answered. I've discussed with my RH sales rep the fact that Windows Server is cheaper after edu discounts, and the desktop Windows (Windows XP, for now) is included in the Campus Agreement. He offered a discount on 11 or more copies of RHEL, but only off of the "Standard" version.
I'm sure many sysadmins don't find themselves calling for support (even good tech support is a hassle) - what we really need is updates and stability.
What I would love:
1. Deep EDU discounts on RHEL ES and WS
2. Perhaps a site-license style agreement for these OSs if we run our own Up2Date server (thus consuming less RH resources)
3. Pay per-incident support or an affordable service-level agreement
4. A program that offers discounted or expedited support when the school funnels support requests through an on-staff RHCE (Dell offers something like this for hardware support)
tried submitting this when I learned about it:
http://www.longbets.org/
It's a similar concept, and there are some very familiar names making wagers (Dave Winer, Eric Schmidt, Vint Cerf, even Ted Danson)
In my experience, most (not all, of course) Open Source projects aren't concerned with backward compatibility outside of the scope of the project itself. Regression testing in OSS is folded into the bug testing.
That's one of the downsides of OSS. The biggest example: When you upgrade your libc, you have to recompile all of your dependent apps. One thing Windows and Solaris have going for them is that you can run a 5 year old binary on the newest version of the OS and it will almost always work.
It's a big burden on the development team to provide support for old interfaces, but this sort of thing is where OSS has a long way to go. It gets really expensive for individual persons/companies to support (bugfix, etc.) packages that are a few revisions old.
This is waaaay too complicated. Your social security card should have two numbers on it:
An identifier (000-00-0000) and
An authenticator (AAA-AA-AAAA)
The identifier can be used to uniquely identify you (until we reach a population of 1,000,001), and the authenticator can be used to authenticate your identity. Provide a public system that can be used to authenticate identifiers (perhaps something similar to what credit card networks use and well-logged/monitored for abuse). Banks, creditors, or even your university could access the system when appropriate. Make it illegal to store authenticators. Provide a system to allow you to (perhaps for a small fee) change your authenticator when your card gets stolen.
This is, after all, a proven system that every slashdot reader uses regularly - good ol' username and password. And most people have already become accustomed to things requiring one, so it shouldn't be a difficult thing for the public to use.
An odd coincidence: this paper was posted to the NANOG (North American Network Operators' Group) mailing list just today. The first paragraph reads: Not an answer, but certainly relevant.
First, I have to express my sincerest thanks for such a great show. You've been more help to my cooking than any other person.
What I want to know is if you've ever been approached by another network/company like PBS? The reason I ask is that of all the shows on Food TV, you're one of the only ones who really TEACHES cooking. Emeril's on every 3rd or 4th timeslot, and he cooks things that most people won't ever make and learn very little from. The closest thing I've seen recently to your caliber of program is Rick Bayless' "Mexico, one plate at a time" (on PBS) - one with some background on the process and history of the food we eat, and why it's prepared the way it is.
I like your show and I'd love to see you get more exposure.
[...] but there is no major reason to switch over.
Ha! Here are 8 reasons to start with. 16 more if you're using IE 5.5.
Sounds to me like you've got some other problem. You might want to consider investing in an UPS, or at least a power conditioner, if you don't already have one. Or perhaps your power supply is to blame, if you've lost all of these drive on the same one/brand.
At a company I used to work for, we used to lose drives like crazy (usually a max life of 1 year). Once our vendor changed power-supply brands, we never had problems again.
Geeks eat pretty much the same food as everyone else
I must disagree. The highest modded comments here, as well as the Ars Technica Cookbook Bachelor Chow, all contain one or more of:
- Crushed chips
- 2 or more cans of different kinds of soup
- Greater than 50% cheese
I consider myself a geek, but apparently I don't eat like one...even when I'm feeding other geeks.The truth helps. Just keep speaking the truth, and tell your friends, people on the bus, folks at work.
"Um, Mr. Bus Driver sir, there's a madman at the back of the bus who won't quit talking about how 'The zero-day virus' is a government conspiracy to make us give up our rights. He sounds like one of those 'hackers' you always hear about...you know like on that movie, 'Hackers'?"
Not to steal Alton Brown's thunder (I own his book, it's awesome), but can anyone suggest any similar books for those of us who just can't justify culinary school, but want to become great amateur chefs? I've searched, but the cost and age of most of the books I've found make me want an opinion before buying.
(perhaps I'm offtopic, but please mod up responses)
These two things are not mutually exclusive:
Creating a more organized and standardized way of formatting config files (anyone heard on something called "XML"?) would help programmers, users, *and* lusers.
Imagine a human-readable, programatically-ediable, self-documenting config file. Might make installers and package managers a little easier to write.
If someone is determined enough, they can get on your WLAN. MAC addresses can be spoofed, WEP keys can be sniffed. All you can do is authenticate and log.
I recently spoke to some keen fellows from Baylor University that have created an OpenBSD-based firewall/logging/authentication system that takes the poster's info page one step further. Everyone authenticates via an SSL-encrypted web site in order to join the network. DHCP leases are handed out in conjunction with a login session, so you can track who does what. Logging in also opens up your firewall to allow the newly-leased IP address through.
According to the interactive demo:
Like most new human interfaces, this will surely take some getting used to, but is this really good ergonomic design? How can you consistently avoid registering a press on nearby buttons? Even if the device is smart enough to tell the difference between intentional and accidental key-presses, every other device out there requires the user to consistently and squarely press each button involved in an input operation. People are used to that, and it's not likely to change.
I don't know about anyone else, but it's gonna be really difficult for me to feel comfortable using an interface where fat-fingering the keypad is normal.
So if the money isn't used to buy taxed software, it just evaporates, right?
A penny saved is a penny earned, an a penny earned is a penny taxed. And usually at a higher rate.
- Course materials
- Course registration
- Library catalog
- Journals (medical, etc)
- Job/intership listings
- Email (web based)
And more. I'm sure other universities have similar offerings. Just look around.