A lot of people purchased the PSP because it was new and cool, and for its multimedia features. Now that the PSP isn't so new anymore and the UMD format is falling out of favor, I fully expect that the DS will begin to pull ahead of the PSP in sales. The reason is that DS sales are driven by the titles themselves (e.g. Nintendogs, Metroid Prime, Mario Kart DS, Tetris DS, etc.) rather than the hype of the hardware. As a result, the Nintendo DS is likely to gain momentum as long as Nintendo keeps producing blockbuster titles for it.
In addition, the Nintendo DS targets a far larger market (adult men, teenage men, children, and women) than the PSP (young adult men with disposable cash) and competes heavily at a lower price point. Customizations like "pink" hardware help push it with the alternative markets.
Basically, Nintendo has a winner on their hands, and will do well as long as they don't screw it up. Sony has a first attempt on their hands that did exceptionally well. We'll see if they follow it up with a more focused device.
Anyone who has actually had an encounter with Apple's customer service would know that they're exactly the same as any other manufacturer.
I've had experience on several occasions, and I have to say that I disagree wholeheartedly. While the Apple reps behind the counter of their "genius bar" are no smarter than the reps for any other manufacturer, they DO go out of their way for the customer. For example, my iBook had a warped hinge thanks to a bus driver stopping suddenly, and a lady standing up at the time. The screen was actually bent too far back, but it still worked. It just closed with some trouble.
In any case, I eventually had to have the logic board replaced under their faulty board recall. The unit was beyond the warranty, and I hadn't chosen to purchase Apple Care. However, they not only replaced the board, but they also replaced the case at no charge. I walked home with a laptop that was in brand new condition. Now that's customer service!
You do realize that an Apple II remake would sell like hotcakes to the retro community, right? The Atari Flashback 2 has been flying off the shelves (in part because it can be modded to take 2600 cartridges), and homebrewers have been sapping up the C64 sticks to add keyboards to them!
You may think you're being funny, but there are a surprising number of people out there who wish you were right.
Cool. Now how about some details? MHz, MIPS, FLOPS, instruction set, JVM version support, OSes ported, what exactly "Network attached processing" services are, etc, would all be nice to know. Is this just a really fancy DRMAA machine, or what?
Don't worry, it will get better. You can forget ever sleeping in again, but you will eventually be able to get 8 hours of unbroken sleep. Just make sure you go to bed early...
It would seem to me, that a CPU's workload is roughly limited by the number of transistors it has multiplied by it's MHz speed.
The number of transistors can go up for a variety of reasons. Chief among them is designs that utilize complex performance enhancements. To name a few:
Superscalar processing
Branch prediction
Hyperthreading
Out of order instructions
Pipelining
The secondary source of transistor usage is coprocessors like Floating Point Units and SIMD Units.
The latest craze in processor design is to simplify the microprocessor back down to the most basic level. From there, the processors are ramped up through shear numbers of parallel pipelines (i.e. threads) and cores as opposed to ramping up the individual CPU horsepower. These multi-core chips typically share coprocessors among a pipelines or cores, and may even have entire cores dedictated to specific tasks like SIMD. As a result, a properly designed program will be able to execute within a very short period of time, thanks to the parallel nature of the multi-core architecture.
Now the only problem is in finding these "properly written programs".
You're far too trusting. Remember, this is the same "online journal" that printed the Google Office Confirmed scoop right after a press conference that announced nothing of the sort.
The Inquirer is taking Azul's word for it at the moment, which is probably why the article is so light on details. About a billion questions pop in my mind when I hear a story like this. The only answer I get is that, "Sun is banging on Azul's door for IP infringement."
Sure.
Does anyone have any real info on these guys? About all I can find is an unsourced Wikipedia article titled Network Attached Processing. It is also lacking in details. (A shill, perhaps?)
Now, if this is updated and it goes from 2 to 3 pages, then now you have a page "2, 3 and 4". Fine, but the next physical page would need to be page 5 now. The only way to fix that, is to reprint EVERYTHING. No ammount of well-planned document formatting will fix this need (unless, as the gp noted, you start doing things like having a page 3 then 3a or some other such nonsense).
If we were talking about "patching" paper books, then you'd be correct. But that's not what the ggp said. He was talking about "re-editing" and "reindexing" to make sure that page numbers still matched. Hyperlinks exist to make that sort of thing redundant. You place the anchors and hyperlinks in the document, then the generated version automatically replaces the links with the correct page number.
There's no real reason not to reprint the entire work. The paper itself is pretty cheap. A spiral bound or glue bound version would easily meet the lower cost requirements for printing. If you're only updating 20% of the book in electronic form, it should be possible to expend far fewer resources on rechecking the book. The problem is that the hand-layouts used today would require that it be treated as a completely new book rather than an update. That just isn't right.
Look at xbox 360. Tons of units sitting on the shelves in Japan, and people can't even find one in the US.
You can't? I was under the impression that they were also sitting on the shelves in the States. At least, that's the impression I get from local game stores...
If publishers would use more digital tools, this wouldn't be a problem. Using tags to identify indexes can allow a computer to precisely compute the page location of modified material. Run it through a layout program configured with the settings for the book, and you can have a professionally done work all ready for POD printing within a few minutes of making the modifications.
This is exactly the type of stuff that technology was designed to solve. The fact that everyone still writes technical books in Microsoft Word so that they can be later laid out by hand is just shocking.
tech books have small audiences, and a short shelf life.
Yet this shouldn't be the case. Books on advanced data structures, OS Design, compiler theory, CPU architectures, language introductions, 3D Graphics theory, Artificial Intelligence, Virtual Machine Design, File System Design, etc, etc, etc, can sit on the shelf for YEARS. There's no inherent reason why computer books are so transient other than the fact that wanna-be programmers want a book on every little, unstable API in existence.
Why do they want books on these subjects? Because they skip learning the basics, then they try to skip learning how to read documentation and standards. All of which means their heads are filling with marketing mush rather than useful information on how to design computer programs. ("Reading the W3C standards is too hard. Whaa!" Be a man/woman, suck it up, and figure it out! You'll get a lot more out of a few hours with the standards than you'll get out of hundreds of hours with fluffy books.)
You have to ask yourself, do you have the FREE manuals for the x86 and AMD64 architectures sitting on your bookshelf? (Other architectures count too, but their documentation isn't usually free.) Have you read them? Why not? The information these books contain can help you understand exactly what your processor is doing, even at the 50,000 ft level of Java or Visual Basic.
So if you find yourself with loads of books on outdated materials, but very few (or none) books on timeless basics, throw away your collection and start looking up the stuff you really need.
Tech books started as an extreme niche market, with well-researched books chock full of useful information. Because of the amount of time and resources that it would take to put a book together (compared to the number of sales over its lifetime), books used to be more expensive.
Somewhere along the way, though, publishers got an idea. If they could fill 300 pages with the literary equivalent of shovelware, they could sell you the book for the same amount of money (since buyers were used to it), but save huge amounts on the research of the title. Thus you ended up with books on VR that did nothing but describe commercial software packages, then in the appendix copy the instructions for a headboom from a far better book. That way they could advertise it as a "build your own VR system!" book, without really doing anything.
This worked so well that publishers got another idea. They could publish even more of these books, and make MORE money! People would still pay it. So they flooded the shelves with whatever was popular at the moment. Be it the Sound Blaster, PERL, Java, XML, LDAP, whatever. It got to the point where if it could be extended from a magazine article, it went to a book form.
And that's how we got to today. If someone can write an entire book on XML DTDs consisting of 30 pages of content, plus 250 pages of source code, manual pages, and descriptions of available parser packages, they will. As a result, the signal to noise ratio is pretty low. If the wannabe programmers would stop buying this crud, we might be able to send a message to the publishers that we want real books. Until then, though, you can only try to sift through the mess of garbage for the good stuff. Check out Bruce Perens' books. I can't vouch for all the content, but at least you can preview them online to see if they're worth the purchase.
Oh, and in case you want to save a tree: Free Online Books (That are worth more than the paper they're printed on.)
I think you're going to see wildly varying answers regarding sizes of teams, depending upon site complexity, etc. The real issue here is that it looks like you need to learn to push back.
That's a nice sentiment, except for one problem: He's a manager, not a coder. He doesn't need to push back, he needs to spend his time managing. Which means that instead of coding, he needs to spend his time doing other things like:
Market the idea to the rest of company. Sending out mockups and ROI case studies of other companies can entice your manager and/or his manager, and do a lot to help sell why more resources are needed.
Use your budget more effectively. Your company may not have given you leave to hire full-time employees, but that doesn't necessarily mean that you can't hire contractors to build tricky or time-consuming portions. Bonus points if you can get stuff auto-generated.
Build trust. You need to gain a reputation as someone who gets things done, and can be trusted with a task. If you build that trust, you'll be trusted that you'll use extra resources wisely rather than empire-building. Yeah, it's difficult with limited resources. Figure it out. You're a manager now, so you'd better find a way.
Don't make excuses. Learn to put a positive spin on timetables, instead. Using tools like Microsoft Project (blech) can allow you to chart out how many man-hours something will take. It can also help you show how it will get done faster if you have more resources.
Don't commit to a project unless you and your superiors are agreed on the timetables. Eveyone expects some slippage on large projects, but too much will cost you dearly. If you already agreed to a messed up timetable (or didn't give one!), then you may need to eat some crow when you present a realistic projection. You should still give that projection, though! Without it, you'll just look incompetent. With it, you'll at least admit to a mistake and ask to correct it.
All in all, I don't hold very much hope for the story submitter. Being a manager is very different from being a programmer. If he's been in his position for two years and hasn't learned how to play the game yet, then he may not be cut out for it. Being a manager is a cut-throat business, and there are only two ways to survive: Either be really good, or be really good at brown nosing. The former is usually preferrable; especially if your bosses are no slouches.
David Cutler (project lead for DEC VMS and RSX11) walked out of DEC when management canceled the x86 VMS port.
It wasn't an x86 port. It was a brand new RISC architecture with a complex new version of VMS. The project was called Parallel Reduced Instruction Set Machine, or PRISM for short.
Supposedly he took the VMS (PRISM) source with him
Actually, it was slightly less illegal than that. Cutler took his entire team with him as a condition for working for Microsoft. They then proceded to redo much of the software work they'd done on PRISM. As you mentioned previously, Digital sued, but ended up settling on the condition that Windows NT be ported to the Alpha. (Fat lot of good that did.)
it was adapted to run DOS and OS/2 applications
I forget the exact terminology, but the kernel has pluggable "personalities" that allow it to function with different user modes attached. There is a decent Wikipedia article on its architecture.
You forgot launches. If launch vehicles were free from gravity (or at least experienced reduced downward acceleration), they could reach orbit on far less fuel. In theory, we could build Star Trek style shuttlecraft with such technology. (Don't get me started on the "Shuttle Pods", though. B&B had their heads up their rears when they dreamed those up. Then again, when aren't their heads up their rears?)
How about they just use a working kernel like say linux, something from a bsd, etc.
Generally speaking, the Windows NT Kernel is a superb piece of code. The problems come in when Microsoft abuses the kernel rather than working with it. The fact that everything runs with Administrator permissions (because all the users run as administrators) is not the original intent of the kernel. Windows Terminal Server Systems tend to be a little more on track, as they default deny administrator privleges to regular users. Unfortunately, they also feel extremely unweildly due to the lack of SUDO-type permission elevation, and the fact that individual desktops are only partly separated from each other. (e.g. Installing new programs is often just as hard as on Unix X Sessions. Many programs don't allow you to install for only one user.)
But, but, but... "When you combine people and technology, you have a very powerful combination." Fashion designer (and part time Microsoft shill) Tommy Hilfiger said so! So it MUST be true! Vista is the future! Viva la Vista!
Seriously, though. A voice command from your cellphone to check email? What are these guys smoking?
We do know what "Worst" is. "Worst" is when your so-called Word file fails to open in Word.
No, worst is when Word decides to open Internet Explorer, which decides the document is a Word file so it opens Word, which decides it's an HTML document so it opens Internet Explorer, which decides it's a Word document...
Unlike Nutscrape Layers, the form.* API is universally supported by every browser made that even pretended to have Javascript support.
For Netscape 4 compatbility. Don't expect it to remain in the future.
I'm not talking about Javascript HTML APIs -- I'm talking about the stuff that everyone uses to manipulate form values that is only 50% standardized.
The document.forms API *is* an HTML API that was part of Netscape's Client Side Scripting. You're not supposed to use it anymore. If you want a form, you're supposed to chose one of three options:
1. Use getElementsByTagName() to find the form by its name. 2. Place an id on it and find the form by getElementById. 3. Walk the tree until you find what you're looking for.
Of course, the form is probably not what you're interested in. What you really want are the form fields. You can use the same three options to find the fields. The resulting object will be exactly the same as if you had used document.forms[i].elements["name"].
In complex forms, most people don't even bother with these options anymore. Since the form is usually created in Javascript, the fields are placed in a datastructure ahead of time. Which means they can be referenced without trying to distinguish them from the rest of the DOM elements.
And let's not even get into window.* -- it's only the proprietary-loving assholes who ever use that
Heh. I agree with you there. The spec says you should be using HTMLFrameElement. So that's what we use.:-)
And Blink has been standardized. See.sig:)
Ok, you got me there. Although, I think the only reason it's defined is because the IETF specs for HTML defined it. According to the spec, "Conforming user agents may simply not blink the text. Note that not blinking the text is one technique to satisfy checkpoint 3.3 of WAI-UAAG." As a result, many browsers ignore the blink tag. Mozilla supports it though. (For whatever reason.) People seem to miss the fact that the blink tag was a joke. *sigh*
Actually it's a pleasant surprise to me, as the W3C has historically focused on hypertext documents and has done almost nothing to improve the state of the art of web applications.
*cough* *sputter* *gag*
Hello? DOM? Document Object Model? The enter API that makes complex web applications possible?
We're talking about a feature introduced in 1999 that has almost universal support already.
Mozilla implemented it for Mozilla 1.0 and Netscape 7. Safari implemented it in OS X 10.3. Opera added support in version 8.0. Microsoft may have implemented it years ago, but the rest of the market hasn't wanted it standardized until recently. Ergo, it wasn't. Those of us who wanted that functionality used hidden iframes instead.
Maybe they can even get around to completing standardization of the form.* API (from Netscape 2.0) one of these days!
I supposed next you're going to demand NS4 layer support? It should go well with the "blink" tag support that I'm sure you're going to request. What are you thinking? The W3C would bust your kneecaps if they caught you using the form.* API!
Three words: Document. Object. Model. You're supposed to be grabbing everything through the DOM, not through Javascript. That's why Netscape's latest specifications for JavaScript 1.5 removed all those web apis. This brings it in line with the ECMA-262 specification. You are coding your Javascript to the ECMA-262 specification, right? Otherwise, you're just creating "legacy" coding problems for yourself that won't port to all current and future browsers.
For someone who thinks Netscape sucks, you sure are married to it.:-P
Slashdot Mirror
Yep, it's April Fools again. And here I thought that the Saturday date would mean that I'd miss Slashdot's bad humor this year.
Thanks Taco. (Like, totally not. Fur sure.)
But here in the U.S., it's a neck-and-neck race.
A lot of people purchased the PSP because it was new and cool, and for its multimedia features. Now that the PSP isn't so new anymore and the UMD format is falling out of favor, I fully expect that the DS will begin to pull ahead of the PSP in sales. The reason is that DS sales are driven by the titles themselves (e.g. Nintendogs, Metroid Prime, Mario Kart DS, Tetris DS, etc.) rather than the hype of the hardware. As a result, the Nintendo DS is likely to gain momentum as long as Nintendo keeps producing blockbuster titles for it.
In addition, the Nintendo DS targets a far larger market (adult men, teenage men, children, and women) than the PSP (young adult men with disposable cash) and competes heavily at a lower price point. Customizations like "pink" hardware help push it with the alternative markets.
Basically, Nintendo has a winner on their hands, and will do well as long as they don't screw it up. Sony has a first attempt on their hands that did exceptionally well. We'll see if they follow it up with a more focused device.
Anyone who has actually had an encounter with Apple's customer service would know that they're exactly the same as any other manufacturer.
I've had experience on several occasions, and I have to say that I disagree wholeheartedly. While the Apple reps behind the counter of their "genius bar" are no smarter than the reps for any other manufacturer, they DO go out of their way for the customer. For example, my iBook had a warped hinge thanks to a bus driver stopping suddenly, and a lady standing up at the time. The screen was actually bent too far back, but it still worked. It just closed with some trouble.
In any case, I eventually had to have the logic board replaced under their faulty board recall. The unit was beyond the warranty, and I hadn't chosen to purchase Apple Care. However, they not only replaced the board, but they also replaced the case at no charge. I walked home with a laptop that was in brand new condition. Now that's customer service!
I mean, how many times have you been able to say that your new computer model was a thousand times faster than your old one?
:-P
Apple II = 1 MHz 6502 Processor
Mac = 2.2 GHz Intel Core Duo Processor
2,200,000,000 / 1,000,000 = 2,200x
You were only off by about a factor of two.
You do realize that an Apple II remake would sell like hotcakes to the retro community, right? The Atari Flashback 2 has been flying off the shelves (in part because it can be modded to take 2600 cartridges), and homebrewers have been sapping up the C64 sticks to add keyboards to them!
You may think you're being funny, but there are a surprising number of people out there who wish you were right.
DRMAA is a standard for distributed computing. By your response, I take it that your hardware isn't intended for that function. :-)
;-)
So what exactly is "network attached processing?"
BTW, thanks for responding. Sorry to bug you during your relaxing downtime in JFK.
Cool. Now how about some details? MHz, MIPS, FLOPS, instruction set, JVM version support, OSes ported, what exactly "Network attached processing" services are, etc, would all be nice to know. Is this just a really fancy DRMAA machine, or what?
:-)
Inquiring minds want to know!
no to mention those nice sleep hours...
Don't worry, it will get better. You can forget ever sleeping in again, but you will eventually be able to get 8 hours of unbroken sleep. Just make sure you go to bed early...
What I need is a life.
;-)
What you need is kids. They'll take care of your free time for you, dontcha worry 'bout that.
The number of transistors can go up for a variety of reasons. Chief among them is designs that utilize complex performance enhancements. To name a few:
The secondary source of transistor usage is coprocessors like Floating Point Units and SIMD Units.
The latest craze in processor design is to simplify the microprocessor back down to the most basic level. From there, the processors are ramped up through shear numbers of parallel pipelines (i.e. threads) and cores as opposed to ramping up the individual CPU horsepower. These multi-core chips typically share coprocessors among a pipelines or cores, and may even have entire cores dedictated to specific tasks like SIMD. As a result, a properly designed program will be able to execute within a very short period of time, thanks to the parallel nature of the multi-core architecture.
Now the only problem is in finding these "properly written programs".
You're far too trusting. Remember, this is the same "online journal" that printed the Google Office Confirmed scoop right after a press conference that announced nothing of the sort.
The Inquirer is taking Azul's word for it at the moment, which is probably why the article is so light on details. About a billion questions pop in my mind when I hear a story like this. The only answer I get is that, "Sun is banging on Azul's door for IP infringement."
Sure.
Does anyone have any real info on these guys? About all I can find is an unsourced Wikipedia article titled Network Attached Processing. It is also lacking in details. (A shill, perhaps?)
Now, if this is updated and it goes from 2 to 3 pages, then now you have a page "2, 3 and 4". Fine, but the next physical page would need to be page 5 now. The only way to fix that, is to reprint EVERYTHING. No ammount of well-planned document formatting will fix this need (unless, as the gp noted, you start doing things like having a page 3 then 3a or some other such nonsense).
If we were talking about "patching" paper books, then you'd be correct. But that's not what the ggp said. He was talking about "re-editing" and "reindexing" to make sure that page numbers still matched. Hyperlinks exist to make that sort of thing redundant. You place the anchors and hyperlinks in the document, then the generated version automatically replaces the links with the correct page number.
There's no real reason not to reprint the entire work. The paper itself is pretty cheap. A spiral bound or glue bound version would easily meet the lower cost requirements for printing. If you're only updating 20% of the book in electronic form, it should be possible to expend far fewer resources on rechecking the book. The problem is that the hand-layouts used today would require that it be treated as a completely new book rather than an update. That just isn't right.
Look at xbox 360. Tons of units sitting on the shelves in Japan, and people can't even find one in the US.
You can't? I was under the impression that they were also sitting on the shelves in the States. At least, that's the impression I get from local game stores...
If publishers would use more digital tools, this wouldn't be a problem. Using tags to identify indexes can allow a computer to precisely compute the page location of modified material. Run it through a layout program configured with the settings for the book, and you can have a professionally done work all ready for POD printing within a few minutes of making the modifications.
This is exactly the type of stuff that technology was designed to solve. The fact that everyone still writes technical books in Microsoft Word so that they can be later laid out by hand is just shocking.
tech books have small audiences, and a short shelf life.
Yet this shouldn't be the case. Books on advanced data structures, OS Design, compiler theory, CPU architectures, language introductions, 3D Graphics theory, Artificial Intelligence, Virtual Machine Design, File System Design, etc, etc, etc, can sit on the shelf for YEARS. There's no inherent reason why computer books are so transient other than the fact that wanna-be programmers want a book on every little, unstable API in existence.
Why do they want books on these subjects? Because they skip learning the basics, then they try to skip learning how to read documentation and standards. All of which means their heads are filling with marketing mush rather than useful information on how to design computer programs. ("Reading the W3C standards is too hard. Whaa!" Be a man/woman, suck it up, and figure it out! You'll get a lot more out of a few hours with the standards than you'll get out of hundreds of hours with fluffy books.)
You have to ask yourself, do you have the FREE manuals for the x86 and AMD64 architectures sitting on your bookshelf? (Other architectures count too, but their documentation isn't usually free.) Have you read them? Why not? The information these books contain can help you understand exactly what your processor is doing, even at the 50,000 ft level of Java or Visual Basic.
So if you find yourself with loads of books on outdated materials, but very few (or none) books on timeless basics, throw away your collection and start looking up the stuff you really need.
Tech books started as an extreme niche market, with well-researched books chock full of useful information. Because of the amount of time and resources that it would take to put a book together (compared to the number of sales over its lifetime), books used to be more expensive.
Somewhere along the way, though, publishers got an idea. If they could fill 300 pages with the literary equivalent of shovelware, they could sell you the book for the same amount of money (since buyers were used to it), but save huge amounts on the research of the title. Thus you ended up with books on VR that did nothing but describe commercial software packages, then in the appendix copy the instructions for a headboom from a far better book. That way they could advertise it as a "build your own VR system!" book, without really doing anything.
This worked so well that publishers got another idea. They could publish even more of these books, and make MORE money! People would still pay it. So they flooded the shelves with whatever was popular at the moment. Be it the Sound Blaster, PERL, Java, XML, LDAP, whatever. It got to the point where if it could be extended from a magazine article, it went to a book form.
And that's how we got to today. If someone can write an entire book on XML DTDs consisting of 30 pages of content, plus 250 pages of source code, manual pages, and descriptions of available parser packages, they will. As a result, the signal to noise ratio is pretty low. If the wannabe programmers would stop buying this crud, we might be able to send a message to the publishers that we want real books. Until then, though, you can only try to sift through the mess of garbage for the good stuff. Check out Bruce Perens' books. I can't vouch for all the content, but at least you can preview them online to see if they're worth the purchase.
Oh, and in case you want to save a tree: Free Online Books (That are worth more than the paper they're printed on.)
That's a nice sentiment, except for one problem: He's a manager, not a coder. He doesn't need to push back, he needs to spend his time managing. Which means that instead of coding, he needs to spend his time doing other things like:
All in all, I don't hold very much hope for the story submitter. Being a manager is very different from being a programmer. If he's been in his position for two years and hasn't learned how to play the game yet, then he may not be cut out for it. Being a manager is a cut-throat business, and there are only two ways to survive: Either be really good, or be really good at brown nosing. The former is usually preferrable; especially if your bosses are no slouches.
I know microsoft got into trouble a few year ago for charging PC distributers for a copy of Windows
Microsoft also agreed to never use a per-processor pricing model. And yet, Windows NT Workstation can't seem to use more than two...
David Cutler (project lead for DEC VMS and RSX11) walked out of DEC when management canceled the x86 VMS port.
It wasn't an x86 port. It was a brand new RISC architecture with a complex new version of VMS. The project was called Parallel Reduced Instruction Set Machine, or PRISM for short.
Supposedly he took the VMS (PRISM) source with him
Actually, it was slightly less illegal than that. Cutler took his entire team with him as a condition for working for Microsoft. They then proceded to redo much of the software work they'd done on PRISM. As you mentioned previously, Digital sued, but ended up settling on the condition that Windows NT be ported to the Alpha. (Fat lot of good that did.)
it was adapted to run DOS and OS/2 applications
I forget the exact terminology, but the kernel has pluggable "personalities" that allow it to function with different user modes attached. There is a decent Wikipedia article on its architecture.
You forgot launches. If launch vehicles were free from gravity (or at least experienced reduced downward acceleration), they could reach orbit on far less fuel. In theory, we could build Star Trek style shuttlecraft with such technology. (Don't get me started on the "Shuttle Pods", though. B&B had their heads up their rears when they dreamed those up. Then again, when aren't their heads up their rears?)
How about they just use a working kernel like say linux, something from a bsd, etc.
Generally speaking, the Windows NT Kernel is a superb piece of code. The problems come in when Microsoft abuses the kernel rather than working with it. The fact that everything runs with Administrator permissions (because all the users run as administrators) is not the original intent of the kernel. Windows Terminal Server Systems tend to be a little more on track, as they default deny administrator privleges to regular users. Unfortunately, they also feel extremely unweildly due to the lack of SUDO-type permission elevation, and the fact that individual desktops are only partly separated from each other. (e.g. Installing new programs is often just as hard as on Unix X Sessions. Many programs don't allow you to install for only one user.)
Vista not People Ready
But, but, but... "When you combine people and technology, you have a very powerful combination." Fashion designer (and part time Microsoft shill) Tommy Hilfiger said so! So it MUST be true! Vista is the future! Viva la Vista!
Seriously, though. A voice command from your cellphone to check email? What are these guys smoking?
We do know what "Worst" is. "Worst" is when your so-called Word file fails to open in Word.
No, worst is when Word decides to open Internet Explorer, which decides the document is a Word file so it opens Word, which decides it's an HTML document so it opens Internet Explorer, which decides it's a Word document...
Unlike Nutscrape Layers, the form.* API is universally supported by every browser made that even pretended to have Javascript support.
:-)
.sig :)
For Netscape 4 compatbility. Don't expect it to remain in the future.
I'm not talking about Javascript HTML APIs -- I'm talking about the stuff that everyone uses to manipulate form values that is only 50% standardized.
The document.forms API *is* an HTML API that was part of Netscape's Client Side Scripting. You're not supposed to use it anymore. If you want a form, you're supposed to chose one of three options:
1. Use getElementsByTagName() to find the form by its name.
2. Place an id on it and find the form by getElementById.
3. Walk the tree until you find what you're looking for.
Of course, the form is probably not what you're interested in. What you really want are the form fields. You can use the same three options to find the fields. The resulting object will be exactly the same as if you had used document.forms[i].elements["name"].
In complex forms, most people don't even bother with these options anymore. Since the form is usually created in Javascript, the fields are placed in a datastructure ahead of time. Which means they can be referenced without trying to distinguish them from the rest of the DOM elements.
And let's not even get into window.* -- it's only the proprietary-loving assholes who ever use that
Heh. I agree with you there. The spec says you should be using HTMLFrameElement. So that's what we use.
And Blink has been standardized. See
Ok, you got me there. Although, I think the only reason it's defined is because the IETF specs for HTML defined it. According to the spec, "Conforming user agents may simply not blink the text. Note that not blinking the text is one technique to satisfy checkpoint 3.3 of WAI-UAAG." As a result, many browsers ignore the blink tag. Mozilla supports it though. (For whatever reason.) People seem to miss the fact that the blink tag was a joke. *sigh*
Actually it's a pleasant surprise to me, as the W3C has historically focused on hypertext documents and has done almost nothing to improve the state of the art of web applications.
:-P
*cough* *sputter* *gag*
Hello? DOM? Document Object Model? The enter API that makes complex web applications possible?
We're talking about a feature introduced in 1999 that has almost universal support already.
Mozilla implemented it for Mozilla 1.0 and Netscape 7. Safari implemented it in OS X 10.3. Opera added support in version 8.0. Microsoft may have implemented it years ago, but the rest of the market hasn't wanted it standardized until recently. Ergo, it wasn't. Those of us who wanted that functionality used hidden iframes instead.
Maybe they can even get around to completing standardization of the form.* API (from Netscape 2.0) one of these days!
I supposed next you're going to demand NS4 layer support? It should go well with the "blink" tag support that I'm sure you're going to request. What are you thinking? The W3C would bust your kneecaps if they caught you using the form.* API!
Three words: Document. Object. Model. You're supposed to be grabbing everything through the DOM, not through Javascript. That's why Netscape's latest specifications for JavaScript 1.5 removed all those web apis. This brings it in line with the ECMA-262 specification. You are coding your Javascript to the ECMA-262 specification, right? Otherwise, you're just creating "legacy" coding problems for yourself that won't port to all current and future browsers.
For someone who thinks Netscape sucks, you sure are married to it.