I can download an album in less time than it takes to listen to it. [...] A movie, on the other hand, I'm probably going to have to leave to download overnight.
About 10 years ago, I was asking some people on IRC about this newfangled MP3 thing. "So, can you download them as fast as you listen to them?"
"Almost, sometimes" came the answer.
Five years from now, a two hour 1080p movie with 5.1 audio will take less than 30 minutes to download on a typical home internet connection. I bet.
CDs and their predecessors are collections of individual performances, with a few exceptions.
You've just broken the heart of every artist that's ever agonised over the running order of their album.
Good music can be produced for next to nothing, whereas it is much more difficult to do that with movies. A song or album can be credibly done by an INDIVIDUAL, or maybe a band and a few extra people to produce. Ten people, tops, unless they're padding it.
I agree up to a point, and I happen to prefer, on the whole, cheaply recorded music.
But consider that lots of people like the expensive stuff. The mainstream superstars spend millions on studio time with extremely high end equipment, studios with expensively built acoustics, engineers and mic technicians and session musicians who charge professional rates. How much do think it costs to hire a 40 piece orchestra for a day?
Still cheaper than a typical movie, but not what you can afford to do with your disposable income.
The guy that thinks that Stackoverflow, essentially a simplified web forum that could be designed by a semi-literate PHP monkey in 48 hours of work, is a major feat of software engineering?
It's like a swear box. You know, in an attempt to get out of the habit of swearing, you put a dollar in box every time you swear. The contents of the box goes to charity.
This is exactly the same, except that in this case the habit you're trying to get out of is releasing bad code.
We all sneak out bad code from time to time - "it's ugly but it works; I can clean it up, or I can ship it and have an extra hour doing [insert recreation of choice]". The 'swear box' makes cleaning it up seem more attractive. And if you don't, a worthy cause benefits.
The analogy to carbon offsets is pretty weak, but I guess it's wry humour of a sort.
Aside from the familiarity of Java, what benefits would Java offer for web services?
From my perspective, the mainstream application server frameworks are all Java-centric. JBoss, WebSphere etc.
There's a laundy list of features you get from these containers - clustering, connection pooling, caching, load balancing, distributed deployment, etc..Net probably gives you all this too, but personally I prefer not to be locked into MS (Mono notwithstanding). The Java stuff is reasonably open, so you can migrate between app servers if you need to.
The mainstream Java IDEs all have plenty of support for web service development.
The big news nowadays is that the Java platform is a lot more friendly than it used to be for polyglot programming. Groovy, Jython, Clojure, JRuby are all lovely languages for writing the kind of stuff that Java tended to make fiddly.
they are talking about Service Oriented Architecture and that translates to the WWW
This is factually wrong, and is a pretty good example of the sort of "talking about something you don't understand" problem that readers have with your post.
Is it? Cloud computing, it is all the rage, it is Software As Service, it is rent me version of software and then there is this Wikipedia article that does seem to support my assertion.
The fact that Web mashups make use of SOA does not mean that SOA 'translates to' WWW.
You could build a whole enterprise around SOA principles, without a single Web browser being involved. Possibly without HTTP being involved. One very robust way of doing enterprise SOA is for your clients and services to talk to each other over some asynch messaging protocol (JMS, Apache ActiveMQ, Websphere MQ, etc.)
If every browser was as standards-compliant as Opera, Google wouldn't have needed to push Chrome. Yes; you should be able to use Google sites to their full function in Opera, once they migrate off Gears to HTML5.
Only problem is in locations that operate on download caps, not bandwidth (Australia for example).
I'm quite sure that problem is going to solve itself, somehow, with market forces. The market is going to have to provide the bandwidth and high/absent caps that people need to use the Internet services they want.
That said, you could use a Web based word processor all day, every day for a month before getting anywhere near a single music download.
It's quite common knowledge that Chrome OS will be locked down. There's even *already* been announcements that it will be the worst piece of *DRM* ever in front of security. If *anything* is changed in the system, the OS downloads it and replaces it again.
Here's how it will actually work, showing how if you want to hack on your own hardware, you can:
1. Verify the partition table on the filesystem looks sane.
2. Load kernel A from the filesystem.
3. Verify signature of kernel.
4. If signature is invalid:
1. If this is kernel A, retry with kernel B.
2. Else this is kernel B. Both kernels are bad, so set the recovery-mode cookie non-volatile register and reboot into recovery firmware.
5. If kernel was signed with a public key not known to the boot loader, this is a developer kernel:
1. Initialize the display.
2. Display scary developer mode warning to user. For example: "DEVELOPER MODE. If you don't know what this means, press X now, and we'll fix your system automatically. If you are a l33t h4x0r, wait 30 seconds or press ASCII 0x44."
3. Wait for keypress or 30-second delay before continuing.
4. If key pressed was not D (ASCII 0x44), jump to Recovery Firmware.
6. Continue booting the kernel.
any company that gets BOFH on their customers, is going to see those customers go elsewhere
Actually, there are lots of customers who would pay for the privilege of not being responsible for the technical maintenance of a computing facility. There are companies that have built their entire revenue stream on that fact.
Absolutely. It's the B and FH from BOFH that nobody wants to pay for. Think of that - someone who does your technical maintenance, yet holds themselves to high standards of customer service!
I tend to go by the thickness of Crockford's book, vs the thickness of any "Complete Javascript" book when determining how much "good stuff" the language has.
I believe there are two reasons for this:
1. Crockford's writing is concise and to the point. It assumes prior programming knowledge.
2. Crockford's book does not concern itself with the DOM
So I believe a good chunk of the extra stuff in the fatter books is "here's what a loop is", "here's what if() does", and a bigger chunk yet is about HTML and CSS.
Are we necessarily bound to a physical presence? Why not push this to go digital along with everything else in our era?
Well, I'm not a wargamer. But I am a board gamer, and I think the pleasure of touching physical things would a shame to lose. Who doesn't prefer to play chess with a beautifully crafted wooden chess set, than on a computer screen?
Can we simplify the models? Because if we don't need any specific, exact kind of orc we could probably get something going to where generic figures are produced that could be customized later. I'm picturing a series blank figures made from carve-able plastic and a standard set of accessories to be glued onto them (armor, weapons, and the like.) Customizable, and there-by open.
It's been done in the proprietary world -- in the late 80s Games Workshop came up with plastic skeletons and dwarves. They were sold as a cheap way to get a big, nondescript army, and they were also ripe for customisation.
Lead is easier for a hobbyist to work with, however. That's why lead soldiers have been around for so long.
Maybe I'm old and stuck in my ways, but having my important data stored somewhere else and not locally makes me jumpy and nervous.
I'm quite the opposite. I think that Google having three copies of my data, with at least one on a different frame, is a lot safer than my single copy that might get backed up once a month but probably won't.
Yeah, I could improve my backup process, but why should I, when someone else is doing it for me for free?
As the number of users increases, the probability of them all hitting the service at once decreases. So you get a better ratio of hardware to users. Still, you need masses of servers for masses of users, so you get economies of scale (bulk buy discounts etc.).
Build a server room for one user -> bad VFM. Build a server room for 100 users -> potentially worthwhile Build a server room for 1000 users -> getting there. A server facility for a million users -> cheap
If you're IBM, you can run your own cloud for your own employees. If you're a two man startup, you're better off outsourcing it.
Slashdot Mirror
I can download an album in less time than it takes to listen to it.
[...]
A movie, on the other hand, I'm probably going to have to leave to download overnight.
About 10 years ago, I was asking some people on IRC about this newfangled MP3 thing. "So, can you download them as fast as you listen to them?"
"Almost, sometimes" came the answer.
Five years from now, a two hour 1080p movie with 5.1 audio will take less than 30 minutes to download on a typical home internet connection. I bet.
I believe that for some formats, the player needs to peek at the start and end of the file, before chewing through from the start.
Hence, many Bittorrent implementations prioritise the start and end of the file.
For as-good-as-streaming, of course you can fetch the beginning, fetch the end, then fetch the rest beginning at the front.
If everyone does that it kills torrent swarm performance as a whole, but hey, we're not advocating P2P here, right ;)
When was the last time a new movie came out that I was even mildly interested in seeing? Donnie Darko? How long ago was that??
Donnie Darko - 2001. Which means you haven't seen Primer - 2004. You need to fix that now.
CDs and their predecessors are collections of individual performances, with a few exceptions.
You've just broken the heart of every artist that's ever agonised over the running order of their album.
Good music can be produced for next to nothing, whereas it is much more difficult to do that with movies. A song or album can be credibly done by an INDIVIDUAL, or maybe a band and a few extra people to produce. Ten people, tops, unless they're padding it.
I agree up to a point, and I happen to prefer, on the whole, cheaply recorded music.
But consider that lots of people like the expensive stuff. The mainstream superstars spend millions on studio time with extremely high end equipment, studios with expensively built acoustics, engineers and mic technicians and session musicians who charge professional rates. How much do think it costs to hire a 40 piece orchestra for a day?
Still cheaper than a typical movie, but not what you can afford to do with your disposable income.
The guy that thinks that Stackoverflow, essentially a simplified web forum that could be designed by a semi-literate PHP monkey in 48 hours of work, is a major feat of software engineering?
Your point is addressed here:
http://www.codinghorror.com/blog/archives/001284.html
If stackoverflow is so simple, why is there no other site that does the job as well?
It seems like a lot of people don't get this.
It's like a swear box. You know, in an attempt to get out of the habit of swearing, you put a dollar in box every time you swear. The contents of the box goes to charity.
This is exactly the same, except that in this case the habit you're trying to get out of is releasing bad code.
We all sneak out bad code from time to time - "it's ugly but it works; I can clean it up, or I can ship it and have an extra hour doing [insert recreation of choice]". The 'swear box' makes cleaning it up seem more attractive. And if you don't, a worthy cause benefits.
The analogy to carbon offsets is pretty weak, but I guess it's wry humour of a sort.
Ugh, I'm contributing to this thread terribly late.
I think the GP has it exactly right.
Scientology is a religion in the same way as Dunkin Donuts is a restaurant
... because you could certainly argue that Dunkin' Donuts is a restaurant. It serves food. It has seating.
So it's all a matter of semantics. Is Scientology a religion? Depends on your definitions.
Aside from the familiarity of Java, what benefits would Java offer for web services?
From my perspective, the mainstream application server frameworks are all Java-centric. JBoss, WebSphere etc.
There's a laundy list of features you get from these containers - clustering, connection pooling, caching, load balancing, distributed deployment, etc. .Net probably gives you all this too, but personally I prefer not to be locked into MS (Mono notwithstanding). The Java stuff is reasonably open, so you can migrate between app servers if you need to.
The mainstream Java IDEs all have plenty of support for web service development.
The big news nowadays is that the Java platform is a lot more friendly than it used to be for polyglot programming. Groovy, Jython, Clojure, JRuby are all lovely languages for writing the kind of stuff that Java tended to make fiddly.
they are talking about Service Oriented Architecture and that translates to the WWW
This is factually wrong, and is a pretty good example of the sort of "talking about something you don't understand" problem that readers have with your post.
Is it? Cloud computing, it is all the rage, it is Software As Service, it is rent me version of software and then there is this Wikipedia article that does seem to support my assertion.
The fact that Web mashups make use of SOA does not mean that SOA 'translates to' WWW.
You could build a whole enterprise around SOA principles, without a single Web browser being involved. Possibly without HTTP being involved. One very robust way of doing enterprise SOA is for your clients and services to talk to each other over some asynch messaging protocol (JMS, Apache ActiveMQ, Websphere MQ, etc.)
I'm no web designer so perhaps I'm misunderstanding TFA, but is offline script caching one of the features of HTML5?
Yes.
http://www.whatwg.org/specs/web-apps/current-work/multipage/offline.html#offline
Somehow, I have hard time believing that.
Why?
If every browser was as standards-compliant as Opera, Google wouldn't have needed to push Chrome. Yes; you should be able to use Google sites to their full function in Opera, once they migrate off Gears to HTML5.
Of course offline GMail will use HTML5. That's *strongly* implied by "we are abandoning Gears in favour of HTML 5"
Offline caching is a feature of HTML5.
http://www.whatwg.org/specs/web-apps/current-work/multipage/offline.html#offline
HTML 5 does exactly what it says it does.
Dive into HTML 5 tells you what that is, and whether your browser supports it.
It's up to developers to apply it. Google is doing so.
Only problem is in locations that operate on download caps, not bandwidth (Australia for example).
I'm quite sure that problem is going to solve itself, somehow, with market forces.
The market is going to have to provide the bandwidth and high/absent caps that people need to use the Internet services they want.
That said, you could use a Web based word processor all day, every day for a month before getting anywhere near a single music download.
Why should I use Chromium OS when I can download any Linux distro and install Chrome on it.
*You* probably don't need to.
You or I might say that Knoppix (or whatever) "just works". But I know people who could screw up Knoppix; I bet you do too.
The whole point of ChromeOS is that it takes "It just works" and "it keeps on just working" to new levels. And of course there are tradeoffs involved.
HTML5 renders Gears (almost) obsolete. Any browser that implements HTML5 will be capable of local storage.
It's quite common knowledge that Chrome OS will be locked down. There's even *already* been announcements that it will be the worst piece of *DRM* ever in front of security. If *anything* is changed in the system, the OS downloads it and replaces it again.
Here's how it will actually work, showing how if you want to hack on your own hardware, you can:
http://sites.google.com/a/chromium.org/dev/chromium-os/chromiumos-design-docs/firmware-boot-and-recovery :
Pseudocode:
1. Verify the partition table on the filesystem looks sane.
2. Load kernel A from the filesystem.
3. Verify signature of kernel.
4. If signature is invalid:
1. If this is kernel A, retry with kernel B.
2. Else this is kernel B. Both kernels are bad, so set the recovery-mode cookie non-volatile register and reboot into recovery firmware.
5. If kernel was signed with a public key not known to the boot loader, this is a developer kernel:
1. Initialize the display.
2. Display scary developer mode warning to user. For example: "DEVELOPER MODE. If you don't know what this means, press X now, and we'll fix your system automatically. If you are a l33t h4x0r, wait 30 seconds or press ASCII 0x44."
3. Wait for keypress or 30-second delay before continuing.
4. If key pressed was not D (ASCII 0x44), jump to Recovery Firmware.
6. Continue booting the kernel.
any company that gets BOFH on their customers, is going to see those customers go elsewhere
Actually, there are lots of customers who would pay for the privilege of not being responsible for the technical maintenance of a computing facility. There are companies that have built their entire revenue stream on that fact.
Absolutely. It's the B and FH from BOFH that nobody wants to pay for. Think of that - someone who does your technical maintenance, yet holds themselves to high standards of customer service!
I tend to go by the thickness of Crockford's book, vs the thickness of any "Complete Javascript" book when determining how much "good stuff" the language has.
I believe there are two reasons for this:
1. Crockford's writing is concise and to the point. It assumes prior programming knowledge.
2. Crockford's book does not concern itself with the DOM
So I believe a good chunk of the extra stuff in the fatter books is "here's what a loop is", "here's what if() does", and a bigger chunk yet is about HTML and CSS.
Are we necessarily bound to a physical presence? Why not push this to go digital along with everything else in our era?
Well, I'm not a wargamer. But I am a board gamer, and I think the pleasure of touching physical things would a shame to lose. Who doesn't prefer to play chess with a beautifully crafted wooden chess set, than on a computer screen?
Can we simplify the models? Because if we don't need any specific, exact kind of orc we could probably get something going to where generic figures are produced that could be customized later. I'm picturing a series blank figures made from carve-able plastic and a standard set of accessories to be glued onto them (armor, weapons, and the like.) Customizable, and there-by open.
It's been done in the proprietary world -- in the late 80s Games Workshop came up with plastic skeletons and dwarves. They were sold as a cheap way to get a big, nondescript army, and they were also ripe for customisation.
Lead is easier for a hobbyist to work with, however. That's why lead soldiers have been around for so long.
Maybe I'm old and stuck in my ways, but having my important data stored somewhere else and not locally makes me jumpy and nervous.
I'm quite the opposite. I think that Google having three copies of my data, with at least one on a different frame, is a lot safer than my single copy that might get backed up once a month but probably won't.
Yeah, I could improve my backup process, but why should I, when someone else is doing it for me for free?
The cloud is a mysterious black box which owes you nothing.
It's a service with a contract, acceptable use policy, terms of service.
You don't like your cloud provider's backup policy? Buy a service that suits you better.
This is the book that'll make you realise Javascript is OK:
http://oreilly.com/catalog/9780596517748
It's not afraid to call out the bad parts, and to show you how to work around them. That's down to a rushed standardization process.
It doesn't deal with the DOM at all - after all, that's not part of JS.
It leaves you thinking JS is pretty neat, if you use it right.
You don't see it?
As the number of users increases, the probability of them all hitting the service at once decreases. So you get a better ratio of hardware to users. Still, you need masses of servers for masses of users, so you get economies of scale (bulk buy discounts etc.).
Build a server room for one user -> bad VFM.
Build a server room for 100 users -> potentially worthwhile
Build a server room for 1000 users -> getting there.
A server facility for a million users -> cheap
If you're IBM, you can run your own cloud for your own employees.
If you're a two man startup, you're better off outsourcing it.