Hmmm, how far could SS1 actually get if it hadn't steered straight up? It would still take an hour to get up to detachment altitude, but if he only flew up to, say 15 km and did a long glide, could it at least get from US West Coast to Europe in a couple hours?
Just curious, is it ok for every *nixboxen on the network to send SMTP? Why should SMTP be blocked for everybody? Seems to me there could be many perfectly legitimate reasons for that in many networks, so preventing address-spoofing is a much better approach.
When AMD and Intel go dual-care relatively soon, I wouldn't be surprised if a dual processor/4-core AMD/Intel outperforms the 12 Efficeon's both per dollar and per watt.
I agree entirely that there's no reason why the attachment of various meta-data to files should be implemented in the file system itself. Instead, add a layer on top of the file system, but below applications and store the file attributes in a database. What would be the big gain from combining the two?
Applications that don't need the metadata could go directly to the FS with less overhead, apps that do need it can use the attribute db. Possibly we'll need the FS to be aware of the attribute database layer, but that can be a pretty thin interface and have near zero impact on performance.
The attribute database can be stored in a different databases and use a different underlying file systems as needed or desired. That kind of flexibility is what we love about Linux, isn't it?
The Linux community places a great deal of trust in Linus, including trusting him to know when new development is likely to be disruptive enough to justify forking to a new development branch.
However, Andrew's vision of leaving final stabilization to the distributors makes me hope it will be Linus making the when-to-fork decision, not Andrew (assuming Jonathan's summarization of Andrew's thinking is accurate). It may be true that the distributors serve this funcion, but it's a corrosive attitude for a core developer to take.
Normally, I think its reasonable to expect readers to figure out which UML or which ASP (to name two overloaded acronyms) from the context, but putting UML in a headline with a database strongly suggested to me that the UML is not the one that it turned out to be. Spelling out "user mode linux" would have been nice, but hey, the most anybody was hurt is that they wasted some time.
The "sideways" applications give you a view of what's in the window. Because it's distorted in a way that the human eye/mind is used to processing from the real world, it should be much easier to look at a number (lets say 20) of 3d slanted windows with app/document names on their spines and pick out the one your looking for.
I haven't seen any other way of presenting a large number of windows that is as easily scanned. Conventionally you've got to (semi-)consciously match an icon to an application, read a document title/filename, and/or scan a thumbnail (a distortion of a document which is much less natural to process than looking at a panel from an angle) in order to find what you're looking for.
Sure, while I'm working on a single project, I want that taking up all the screen I've got (minus a small amount of space I'll allocate to things I want to monitor), but if I could hold down a key to show me a 3D view for switching to something else, that would be genuinely useful.
Do you really need a hard disk for each unit? Boot all those puppies from the lan and save your money for the electricity, which will still be high, but not as high.
It seems to me that the whole "price of linux" thing is just wrong-headed. When I buy SUSE 9.1, I can install it on more than one computer, right? I can't do that with Windows. So why would I buy 5 copies of linux rather than 1? Isn't the real cost only 1/10 of MS?
In all of the rest of these the open source software is not the primary money maker...
Well, yeah. That's the whole point. You don't make money by giving away stuff for free. Ever. These business models are ones where it is to a company's advantage (usually indirect) to develop and distribute open source software.
Of course you can be a consultant without writing open source software. I would think that point is so obvious, no one would misconstrue the article as suggesting anything else.
It isn't hard to try to find for a profit making enterprise NOT to work on open source projects. These are strategies which provide a reason TO work on them. Even hosting provides a reason to contribute to open source projects (so that you're improvements to core code get into the open source version and you can then take advantage of other improvements to the open source version without having to redo your modifications), but I'd say its a much weaker one.
Huh? All except one involve distributing open source code:
Optimization - give away good, sell better code. Dual license - give away good code, sell same code to anyone not wanting to GPL their modifications. Consulting - give away code so people will hire you to customize it. Subscription - give away code, but get people to pay for more convenient distribution of it. Patronage - find someone else to pay you to give away code. Hosted - (this is the one) write code based on free software that only you use. Embedded - Give away code tailored to hardware you sell.
I think the point is that ramping up speeds is no longer the priority it was. Performance gains will come from including muliple cores on one chip instead of ever higher clock speeds (and the power consumption, cooling requirements, and noise that come with them).
Since having many processes and threads is the rule in today's computers (I have, let me check, 46 processes and 466 threads going at the moment, and I ain't really doing anything), this seems like a pretty good plan to me.
Ok, so I exagerrated a bit, you can get mini-ITX, (I didn't know there was more than one). But I can choose oodles of motherboards in various form factors if I'm willing to use a CPU that can also serve as a heater for a small room.
I did read recently that AMD processors can give you decent performance with low heat output if underclocked and undervolted, which I might try in the future.
I think Shirky's point is more than that good design is important, but that design for small groups is inherently better -- and I think it's wrong because of that. The main benefit of targetting small groups is that it makes step 0, which you didn't list but is the MOST important step, much harder to screw up:
0. Identify who, and what situations, you're designing for
If you don't do that, you'll design crap that doesn't do anything well for anybody. If you identify your user and what they're doing, you at least have a shot at making your software handle that situation magnificently. And if you choose your user/situation right, your core users will be ecstatic and the users who need to accomodate software tailored to someone else will be the users who are most up to the job.
Most of these ideas bear an uncanny similarity to those in "The inmates are running the asylum" which is also a pretty amusing read.
Why is it that ease-of-use, style and brand name are considered things that you shouldn't have to pay for? Style and brand name may be a bit harder to justify, but ease-of-use is pretty darn important. Does it really make sense to buy a harder to use product just because it's specs are better? Ooooh, I'm not going to by this easy-use-MP3 player, I'm going to buy one for the same price that has 10x as much storage as I need instead of just 2x.
I don't own an iPod, but I've spent enough time with my son's to say that it is an exquisitely designed device, far more usable than any other handheld devices I've had experience with. It may not be cheap, but its not overpriced either. Good design is not just "cool," it's valuable and those of us building or programming information management of any sort need to learn that.
The underlying assumption that you buy a company to get its technology is so wrong, even for a technology company. Brand or name recognition, distribution and marketing infrastructure, customer base, alliances / partnerships / connections are in all but the rarest cases far more valuable than any technology gained by a merger or acquisition.
Well, I think women earn more in the worlds oldest profession.
Re:a word of warning
on
Buying a New TV?
·
· Score: 2, Interesting
My highly scientific *cough* research doesn't support this. There are a lot of internet postings indicating people with no problems, but many people seem to have phosphor burn problems as well. (You also gotta wonder why RPTV manuals have dire warnings about phosphor burn if it's not a big deal. One manual I read in PDF, maybe Toshiba, had the warning every half dozen pages or so.)
Logos (aka bugs) on things like Discovery Channel and new channel crawl bands seem particularly problematic. Proper configuration (primarily reducing brightness and contrast to sane levels) may have a significant impact.
I've got no problems with my Samsung HDTV (reasonable price, looks good, shallowest cabinet out there, a bit over budget for OP), but its only been a couple months.
Most exciting new music I've heard in the last few years -- Christian McBride, Uri Caine, Ahmir Thompson. I got to see these cats in one of very few live gigs -- they all got too much else going on.
I'm also a big Don Pullen fan. He died too young (50s) a couple years back, but New Beginnings and the African Brazilian recordings (Kele Mo Bana) are among the high points.
And of course Diz, Bird, Miles, Herbie, Louis, Billie Holliday!, Trane, Brubeck, Duke,... so much good stuff.
And I'll second Coltrane and Johnny Hartman. Ahhhhh.
Slashdot Mirror
Hmmm, how far could SS1 actually get if it hadn't steered straight up? It would still take an hour to get up to detachment altitude, but if he only flew up to, say 15 km and did a long glide, could it at least get from US West Coast to Europe in a couple hours?
Just curious, is it ok for every *nixboxen on the network to send SMTP? Why should SMTP be blocked for everybody? Seems to me there could be many perfectly legitimate reasons for that in many networks, so preventing address-spoofing is a much better approach.
When AMD and Intel go dual-care relatively soon, I wouldn't be surprised if a dual processor/4-core AMD/Intel outperforms the 12 Efficeon's both per dollar and per watt.
I agree entirely that there's no reason why the attachment of various meta-data to files should be implemented in the file system itself. Instead, add a layer on top of the file system, but below applications and store the file attributes in a database. What would be the big gain from combining the two?
Applications that don't need the metadata could go directly to the FS with less overhead, apps that do need it can use the attribute db. Possibly we'll need the FS to be aware of the attribute database layer, but that can be a pretty thin interface and have near zero impact on performance.
The attribute database can be stored in a different databases and use a different underlying file systems as needed or desired. That kind of flexibility is what we love about Linux, isn't it?
The Linux community places a great deal of trust in Linus, including trusting him to know when new development is likely to be disruptive enough to justify forking to a new development branch.
However, Andrew's vision of leaving final stabilization to the distributors makes me hope it will be Linus making the when-to-fork decision, not Andrew (assuming Jonathan's summarization of Andrew's thinking is accurate). It may be true that the distributors serve this funcion, but it's a corrosive attitude for a core developer to take.
Normally, I think its reasonable to expect readers to figure out which UML or which ASP (to name two overloaded acronyms) from the context, but putting UML in a headline with a database strongly suggested to me that the UML is not the one that it turned out to be. Spelling out "user mode linux" would have been nice, but hey, the most anybody was hurt is that they wasted some time.
Wasted time. Slashdot. Hmmm....
The "sideways" applications give you a view of what's in the window. Because it's distorted in a way that the human eye/mind is used to processing from the real world, it should be much easier to look at a number (lets say 20) of 3d slanted windows with app/document names on their spines and pick out the one your looking for.
I haven't seen any other way of presenting a large number of windows that is as easily scanned. Conventionally you've got to (semi-)consciously match an icon to an application, read a document title/filename, and/or scan a thumbnail (a distortion of a document which is much less natural to process than looking at a panel from an angle) in order to find what you're looking for.
Sure, while I'm working on a single project, I want that taking up all the screen I've got (minus a small amount of space I'll allocate to things I want to monitor), but if I could hold down a key to show me a 3D view for switching to something else, that would be genuinely useful.
Do you really need a hard disk for each unit? Boot all those puppies from the lan and save your money for the electricity, which will still be high, but not as high.
Well, it may be worse than nothing, but its better than Voyager.
It seems to me that the whole "price of linux" thing is just wrong-headed. When I buy SUSE 9.1, I can install it on more than one computer, right? I can't do that with Windows. So why would I buy 5 copies of linux rather than 1? Isn't the real cost only 1/10 of MS?
Doh!
P2 should begin "It isn't hard to try to find reasons for..."
It isn't hard to try to find for a profit making enterprise NOT to work on open source projects. These are strategies which provide a reason TO work on them. Even hosting provides a reason to contribute to open source projects (so that you're improvements to core code get into the open source version and you can then take advantage of other improvements to the open source version without having to redo your modifications), but I'd say its a much weaker one.
Huh? All except one involve distributing open source code:
Optimization - give away good, sell better code.
Dual license - give away good code, sell same code to anyone not wanting to GPL their modifications.
Consulting - give away code so people will hire you to customize it.
Subscription - give away code, but get people to pay for more convenient distribution of it.
Patronage - find someone else to pay you to give away code.
Hosted - (this is the one) write code based on free software that only you use.
Embedded - Give away code tailored to hardware you sell.
It's been a while since I looked at it, but doesn't that violate the (soon-to-be-discarded) Yast license?
I think the point is that ramping up speeds is no longer the priority it was. Performance gains will come from including muliple cores on one chip instead of ever higher clock speeds (and the power consumption, cooling requirements, and noise that come with them).
Since having many processes and threads is the rule in today's computers (I have, let me check, 46 processes and 466 threads going at the moment, and I ain't really doing anything), this seems like a pretty good plan to me.
I did read recently that AMD processors can give you decent performance with low heat output if underclocked and undervolted, which I might try in the future.
Better yet, build your own desktop using a Pentium M or Athlon Mobile motherboard.
Oh wait. you can't get them. Never mind.
I think Shirky's point is more than that good design is important, but that design for small groups is inherently better -- and I think it's wrong because of that. The main benefit of targetting small groups is that it makes step 0, which you didn't list but is the MOST important step, much harder to screw up:
0. Identify who, and what situations, you're designing for
If you don't do that, you'll design crap that doesn't do anything well for anybody. If you identify your user and what they're doing, you at least have a shot at making your software handle that situation magnificently. And if you choose your user/situation right, your core users will be ecstatic and the users who need to accomodate software tailored to someone else will be the users who are most up to the job.
Most of these ideas bear an uncanny similarity to those in "The inmates are running the asylum" which is also a pretty amusing read.
I don't own an iPod, but I've spent enough time with my son's to say that it is an exquisitely designed device, far more usable than any other handheld devices I've had experience with. It may not be cheap, but its not overpriced either. Good design is not just "cool," it's valuable and those of us building or programming information management of any sort need to learn that.
The underlying assumption that you buy a company to get its technology is so wrong, even for a technology company. Brand or name recognition, distribution and marketing infrastructure, customer base, alliances / partnerships / connections are in all but the rarest cases far more valuable than any technology gained by a merger or acquisition.
Well, I think women earn more in the worlds oldest profession.
My highly scientific *cough* research doesn't support this. There are a lot of internet postings indicating people with no problems, but many people seem to have phosphor burn problems as well. (You also gotta wonder why RPTV manuals have dire warnings about phosphor burn if it's not a big deal. One manual I read in PDF, maybe Toshiba, had the warning every half dozen pages or so.)
Logos (aka bugs) on things like Discovery Channel and new channel crawl bands seem particularly problematic. Proper configuration (primarily reducing brightness and contrast to sane levels) may have a significant impact.
I've got no problems with my Samsung HDTV (reasonable price, looks good, shallowest cabinet out there, a bit over budget for OP), but its only been a couple months.
Cyberonic gets generally good reviews at broadband reports - 1.5/768 month to month for $50.
Ok, name one.
Most exciting new music I've heard in the last few years -- Christian McBride, Uri Caine, Ahmir Thompson. I got to see these cats in one of very few live gigs -- they all got too much else going on.
I'm also a big Don Pullen fan. He died too young (50s) a couple years back, but New Beginnings and the African Brazilian recordings (Kele Mo Bana) are among the high points.
And of course Diz, Bird, Miles, Herbie, Louis, Billie Holliday!, Trane, Brubeck, Duke,... so much good stuff.
And I'll second Coltrane and Johnny Hartman. Ahhhhh.