Is it just me, or does the CGI look like your now cliche Saturday morning cartoon? Plastic toys brought to life? Or a bad tie-in Video game's cutscenes?
No care in character design? No fluidity of momevemnt? No rendering time for clean textures? No unique style like Genndy Tartakovsky's "Clone Wars"?
Assassin's Creed looks better than that, and its rendered in real time on a single X-Box!
Please remember, Linus is primarily a kernel maintainer. He's responsible for the under-the-line stuff that makes it such a great server OS.
But the user experience is largely the purvue of the Distros, their window managers, application suites, etc. And Linus is right, these are a disaster.
But saying he's divorced is silly, its never been his area of expertise or the area where he works.
"Installation Information for a User Product means any methods, procedures, authorization keys, or other information required to install and execute modified versions of a covered work in that User Product from a modified version of its Corresponding Source. The information must suffice to ensure that the continued functioning of the modified object code is in no case prevented or interfered with solely because modification has been made.
If you convey an object code work under this section in, or with, or specifically for use in, a User Product, and the conveying occurs as part of a transaction in which the right of possession and use of the User Product is transferred to the recipient in perpetuity or for a fixed term (regardless of how the transaction is characterized), the Corresponding Source conveyed under this section must be accompanied by the Installation Information. But this requirement does not apply if neither you nor any third party retains the ability to install modified object code on the User Product (for example, the work has been installed in ROM)."
QED: You can't distribute an XO containing GPLv3 software without giving the authorization key. Thus an education ministry can't use GPLv3 software on an XO while still maintaiting the "only run signed code" model, as they DO have the authorization key for the laptop.
Ah, RM "Proprietary is Never Good" Stallman...
on
Richard Stallman on OLPC
·
· Score: 2, Insightful
The XO's networking capabilities is fantastic. It gets far better range (thanks to its dual rabbit ears), has ultra low power mesh networking, and a bunch of other capabilities.
But because it uses binary blobs for the driver and firmware, RMS fees it is hopelessly compromised?!
Does RMS not drive a car built in the past 20 years because you aren't supposed to change the computer running the engine? What about fly in a commercial airliner?
Also, the XO can never use GPLv3 code. For the US market, they will give the unlock key, but for the third world, this key is the responsibility of the educational ministry, which often needs to keep the software base consistent (among other things, this helps manage theft).
The iPhone is revolutionary because it just works.
I've looked at smartphones in the past, and play with them whenever I'm paynig my wireless bill at the store instead of the mail.
Other smartphones don't have web browsers that just works, they don't have email that just works, they don't connect to the computer in a way that just works they don't have a user interface so simple my mom can use it but so powerful I'd love to use it.
I don't have one yet, because I don't NEED a smartphone. But if I wanted a smartphone, rather than just a cellphone, the "It Just Works" factor make it the iPhone or nothing.
Guys, block structured languages you can trivially derive pretty-printing formatting. Which means you get the "easy to read" ability trivially. You should NEVER read the Silly Parenthesis in LISP, that gets taken care of for you by any one of a gazillion tools.
But you can't go the other way around.
EG, on C, you can cut and paste trivially. You can't on python, you need an editor which parses the language sufficiently to understand that "This is a code block going into a different code block".
Likewise, code generation. Its trivial to build a recursive code generator targeteing "pick your favorate block structure language". But in Python, you have to explicitly add a notion of "what depth is my recursive code generator actually operating at", and "woe to you if you forget a linefeed".
Next thing you know, slashdot will be full of defenders who insist that Make's treatment of tab and space is a good thing.
Like so much of Windows, whitespace-as-block-structure is a bug, not a feature.
I'm saying the editor fixes the indentation for you.
Have you ever used LISP? The editor reads the parenthesis, you read the indentation.
Indentation is a derived piece of structure, but should never have semantic meaning in a language except as a separator. Anytime you do so you ask for problems, both for editors, for code generators, and for users.
When a program is "GPLv2 or later" and "Contributions are signed back to the company", if the OSS software is useful the GPL's patent poison pill actually makes it a more attractive buyout target.
You see, when Apple buys out CUPS, or Sun buys out MySQL, they can distribute the code under whatever liscence they want as copyright holders.
Which means, for THEM, its non GPL. But for everyone else, it is GPLv2 or later.
Thus you can still get community support with copyright transfer, but you have a competitive advantage in selling/using it commmercially.
I wonder if the FSF realized that this is a side consequence of GPLv3 and the "or later" clause?
The arms race favors the MP/RI-AA in this, because in order for a P2P system to work for file exchange, you need to be able to actually get to the peers which have portions of the file.
You don't just take the tracker at its word, but instead actually verify the nodes. This means the graph is "correct", preventing the joe-job defense.
You aren't blocking the tracker, but the actual P2P communication within the system.
And Tor so throttles ones' performance that only the truely paranoid use it for their BitTorrent.
The only reasonably-robust defense strategies involve detecting and blocking the spiders, and distinguishing them from humans. This is a hard problem, as the RI/MP-AA could even use humans (outsourced to India/China for $1/hr) for any human-like behavior needed.
One other possibility might be some honeypot nodes, which if blocked act as early warning and you track which nodes were told about the honeynodes. The problem is if the ISP is cooperating, the spider node IP addresses can be volatile (unused DHCP leases with a quick rotation time, from a large range of addresses with actual users).
The only robust defense is closed-world piracy, where the spiders can't get in, but such networks of pirates, by their very definition, are less distributive in information.
If Google can find it, an MP/RI-AA spider could find it and spider the torrent.
If Google can't find it, the pirate users can't find it.
Oh, on the liability: according to the original article, this messes up one set of liability protection AT&T has, but they might still be able to retreat to the DMCA safe harbor provision, because they actually aren't making a decision about copyright, just enforcing someone else's decision.
But since they are enforcing someone else's decision, they can probably avoid liability if the decision is bogus.
Time-Warner cable supposidly has 50% of the bandwidth used by 5% of the users. Who wants to bet that of this bandwidth, it is almost all pirated material?
The strength of piracy on the Internet is the ease of getting the
pirated material, and the ease of distribution. Thus pirated material must be easy to find. So all the MP/RI-AA has to do is find it, and do something about it. Rather than
playing Whak-A-Mole on Torrent tracker servers (which are largely offshore),
with ISP cooperation from AT&T it becomes possible to play Whak-A-Mole on the
users of the torrents themselves...
So the MP/RI-AA or their contractor surfs the Torrent sites, and connects to the torrents
with a manipulated client, verifies that a particular torrent is a
copyright violation, maps the users of the torrent, and then sends an
automated list of the nodes to the ISP saying "This graph is bad, any
edge between two nodes in this graph should be killed", and the ISP simply
RST-flood any edge in the graph which crosses its network, or just put in a router ACL to drop that pair for a while. Because the strength of the system relies on it being public and P2P, the MP/RI-AA can easily get this information.
AT&T has multiple incentives to cooperate, and can probably do it safely. It has a second party (MP/RI-AA or a company they create/contract for) do the deciding, so they dont' have the liabliity.
It keeps the content providers happy for when they are negotiating their compete-with-iTunes/Netflix video on demand and cable TV services.
It keeps the content providers from pushing through very draconian legislation, or at least draconian legislation you aren't happy with. (It can F-up your competitors, but thats just a bonus)
Its very easy to implement (short-lived router ACLs which are automatically injected and revoked).
And it drops their bandwidth bills by 30-50% by eliminating a large amount of deliberately-noncacheable (both politically and because of bittorrent encryption) traffic.
I wouldn't take it as a guarentee, but I'd almost be willing to bet that AT&T does something like this in the next year. Who wouldn't leap at a chance to reduce your costs by 30%, keep a group of "partners" you have to deal with happy, and without any real work on your part (just an SNMP-manager program)?
This won't stop closed-world pirates, but those are far less annoying
to the ISPs simply because there are so many fewer of them, and less
important to the MP/RI-AA because they are less likely to be users you
can convert to paying customers if you make the illegal content
sources unusable.
Its too late. The writing is on the wall. With almost all studios having defected to Blu-Ray primary/Blu-Ray only, anyone who's been sitting out the format war to date is not going to jump at this.
Especially since, lets face it, you'd only care about Blu-Ray/HD-DVD in the first place if you drop $1k-2k+ on the TV itself, and another $200-1K on the stereo system.
I use WPA. Why? Because on my parents network, they want to use file sharing between their desktop and their laptop. On both mine and my parents, there are networked printers.
But I write down the password on the router, and anyone who visits in person is welcome to use it.
Does Bruce not use a home printer? Share files between home computers?
A company which is trying to rake in millions by providing a "You paid more so you can skip the line" service, which promised shoe scanners etc, has to resort to trying to give a (rather small, given the need to get TSA approval) prize purse to make their business model work?
Re:The classmate hardware SUCKS, at least...
on
Negroponte vs Intel
·
· Score: 2, Interesting
I hardly say I've drunk the OLPC cool-aid. I think it will be a failure, and the current software is a disaster.
But the hardware design really is vastly better. The hardware on the XO is brilliant. You could make a "big keyboard" version and sell it as is to the military for $2000 a copy, its that rugged.
Just some of the weaknesses of the Classmate's design in comparison (beyond the 50% higher price tag):
a: Cooling openings are a weakness and unnecessary in a device which should be passively cooled. (Dust, dirt, and debris is the enemy. Even with the faster processor, a metal slug & fins could have been put into the case. It will get dirty, even in the 1st world. I've had computers die due to dust before.) I think it is actively cooled, which makes the problem worse!
b: The keyboard is not as ruggedized. (Probably feels a lot better, however.)
c: The monitor hinge is a weakness (as I said, 1 semi-dead laptop myself).
d: The screen is no where near as good. (You can really read text on the XO's screen, ranging from full sunlight to full-dark. But an 800x480 LCD? Ick).
e: No support for mesh networking.
f: Significantly more power requirements.
g: Even the lid-closure isn't as robust, compared with the XO's design (which covers all ports but power when closed, yet ups WiFi range when open)
The OLPC's politics annoy me, the software needs real work (I'd rather run WinNT 3.1 over the current OLPC software), but the hardware really is vastly superior. Intel's hardware is a bad joke in comparison.
The classmate hardware SUCKS, at least...
on
Negroponte vs Intel
·
· Score: 5, Informative
IF targeting the 3rd world, the classmate sucks:
a) There are cooling holes on it! Hello dirt and debris.
b) The keyboard is non ruggedized, at least compared to the XO.
c) It uses a conventional montior arrangement rather than the OLPC "behind the monitor" arrangement. This means that it has a complex, wire heavy connector through the hinge rather than just a USB and power connection.
I don't see how the classmate could last 6 months in a third world environment.
I question some of the OLPC's intent, but their hardware design blows away that Intel POS its not even funny.
a: This is very old news, from back in October, just rehashed to get more clicks.
b: It is irrelevant. Even if Linus loved the GPLv3, there is so much code contributed to the Linux kernel without a transfer of copyright and under GPLv2 only terms that it couldn't be changed anyway.
Although I believe copyright should be much shorter (15-50 years, rather than "forever on the Installment Plan"), this reeks of "I Want A Pony". It will NEVER happen. Never. Ever. Ever.
I've yet to find a system which is easier to change the drive on than a MacBook...
Three screws and the drive carrier slides out. Four Torq screws to move the carrier to a new drive. Pop the new drive in and go. I know, I just upped mine from 80 to 160 GB.
Its the kind of repair which should never have been sent in at all, either a Do It Thyself or do it in a shop, as it really is a 5 minute job.
By making it a Trike (3 wheel) its largely counted as a motorcycle so all the crash testing requirements go out the window (including side impact, which would shread this little egg).
But the bigger worry is that trikes are far less stable in a turn, because it is at a much earlier point that they start to roll over.
The Corbin Sparrow had a real tendency to roll over. Alpina may be better by having a wider front wheel footprint, but the pod shape has a higher center of gravity. I hope either they have really REALLY good Dynamic Stability Control/Electonic Stability Control or really REALLY good laywers.
Lost in the "Oh goody, non embrionic stem cells" congradulatory bit on the part of the zealots is they forget that this is also "big step towards human cloning".
Slashdot Mirror
Is it just me, or does the CGI look like your now cliche Saturday morning cartoon? Plastic toys brought to life? Or a bad tie-in Video game's cutscenes?
No care in character design? No fluidity of momevemnt? No rendering time for clean textures? No unique style like Genndy Tartakovsky's "Clone Wars"?
Assassin's Creed looks better than that, and its rendered in real time on a single X-Box!
Star wars used to have some class, like the Holiday Special.
Please remember, Linus is primarily a kernel maintainer. He's responsible for the under-the-line stuff that makes it such a great server OS.
But the user experience is largely the purvue of the Distros, their window managers, application suites, etc. And Linus is right, these are a disaster.
But saying he's divorced is silly, its never been his area of expertise or the area where he works.
"Installation Information for a User Product means any methods, procedures, authorization keys, or other information required to install and execute modified versions of a covered work in that User Product from a modified version of its Corresponding Source. The information must suffice to ensure that the continued functioning of the modified object code is in no case prevented or interfered with solely because modification has been made.
If you convey an object code work under this section in, or with, or specifically for use in, a User Product, and the conveying occurs as part of a transaction in which the right of possession and use of the User Product is transferred to the recipient in perpetuity or for a fixed term (regardless of how the transaction is characterized), the Corresponding Source conveyed under this section must be accompanied by the Installation Information. But this requirement does not apply if neither you nor any third party retains the ability to install modified object code on the User Product (for example, the work has been installed in ROM)."
QED: You can't distribute an XO containing GPLv3 software without giving the authorization key. Thus an education ministry can't use GPLv3 software on an XO while still maintaiting the "only run signed code" model, as they DO have the authorization key for the laptop.
The XO's networking capabilities is fantastic. It gets far better range (thanks to its dual rabbit ears), has ultra low power mesh networking, and a bunch of other capabilities.
But because it uses binary blobs for the driver and firmware, RMS fees it is hopelessly compromised?!
Does RMS not drive a car built in the past 20 years because you aren't supposed to change the computer running the engine? What about fly in a commercial airliner?
Also, the XO can never use GPLv3 code. For the US market, they will give the unlock key, but for the third world, this key is the responsibility of the educational ministry, which often needs to keep the software base consistent (among other things, this helps manage theft).
The iPhone is revolutionary because it just works.
I've looked at smartphones in the past, and play with them whenever I'm paynig my wireless bill at the store instead of the mail.
Other smartphones don't have web browsers that just works, they don't have email that just works, they don't connect to the computer in a way that just works they don't have a user interface so simple my mom can use it but so powerful I'd love to use it.
I don't have one yet, because I don't NEED a smartphone. But if I wanted a smartphone, rather than just a cellphone, the "It Just Works" factor make it the iPhone or nothing.
Ah, I love the python defenders...
Guys, block structured languages you can trivially derive pretty-printing formatting. Which means you get the "easy to read" ability trivially. You should NEVER read the Silly Parenthesis in LISP, that gets taken care of for you by any one of a gazillion tools.
But you can't go the other way around.
EG, on C, you can cut and paste trivially. You can't on python, you need an editor which parses the language sufficiently to understand that "This is a code block going into a different code block".
Likewise, code generation. Its trivial to build a recursive code generator targeteing "pick your favorate block structure language". But in Python, you have to explicitly add a notion of "what depth is my recursive code generator actually operating at", and "woe to you if you forget a linefeed".
Next thing you know, slashdot will be full of defenders who insist that Make's treatment of tab and space is a good thing.
Like so much of Windows, whitespace-as-block-structure is a bug, not a feature.
I'm saying the editor fixes the indentation for you.
Have you ever used LISP? The editor reads the parenthesis, you read the indentation.
Indentation is a derived piece of structure, but should never have semantic meaning in a language except as a separator. Anytime you do so you ask for problems, both for editors, for code generators, and for users.
Instead of the stupid "whitespace has semantic meaning", which Fs up editors, cut & paste, and code generators?
Well, execpt that The Estonian "Cyber War" was really internal-to-Estonia script kiddiez.
When a program is "GPLv2 or later" and "Contributions are signed back to the company", if the OSS software is useful the GPL's patent poison pill actually makes it a more attractive buyout target.
You see, when Apple buys out CUPS, or Sun buys out MySQL, they can distribute the code under whatever liscence they want as copyright holders.
Which means, for THEM, its non GPL. But for everyone else, it is GPLv2 or later.
Thus you can still get community support with copyright transfer, but you have a competitive advantage in selling/using it commmercially.
I wonder if the FSF realized that this is a side consequence of GPLv3 and the "or later" clause?
The arms race favors the MP/RI-AA in this, because in order for a P2P system to work for file exchange, you need to be able to actually get to the peers which have portions of the file.
You don't just take the tracker at its word, but instead actually verify the nodes. This means the graph is "correct", preventing the joe-job defense.
You aren't blocking the tracker, but the actual P2P communication within the system.
And Tor so throttles ones' performance that only the truely paranoid use it for their BitTorrent.
The only reasonably-robust defense strategies involve detecting and blocking the spiders, and distinguishing them from humans. This is a hard problem, as the RI/MP-AA could even use humans (outsourced to India/China for $1/hr) for any human-like behavior needed.
One other possibility might be some honeypot nodes, which if blocked act as early warning and you track which nodes were told about the honeynodes. The problem is if the ISP is cooperating, the spider node IP addresses can be volatile (unused DHCP leases with a quick rotation time, from a large range of addresses with actual users).
The only robust defense is closed-world piracy, where the spiders can't get in, but such networks of pirates, by their very definition, are less distributive in information.
If Google can find it, an MP/RI-AA spider could find it and spider the torrent.
If Google can't find it, the pirate users can't find it.
Oh, on the liability: according to the original article, this messes up one set of liability protection AT&T has, but they might still be able to retreat to the DMCA safe harbor provision, because they actually aren't making a decision about copyright, just enforcing someone else's decision.
But since they are enforcing someone else's decision, they can probably avoid liability if the decision is bogus.
Time-Warner cable supposidly has 50% of the bandwidth used by 5% of the users. Who wants to bet that of this bandwidth, it is almost all pirated material?
The strength of piracy on the Internet is the ease of getting the pirated material, and the ease of distribution. Thus pirated material must be easy to find. So all the MP/RI-AA has to do is find it, and do something about it. Rather than playing Whak-A-Mole on Torrent tracker servers (which are largely offshore), with ISP cooperation from AT&T it becomes possible to play Whak-A-Mole on the users of the torrents themselves...
So the MP/RI-AA or their contractor surfs the Torrent sites, and connects to the torrents with a manipulated client, verifies that a particular torrent is a copyright violation, maps the users of the torrent, and then sends an automated list of the nodes to the ISP saying "This graph is bad, any edge between two nodes in this graph should be killed", and the ISP simply RST-flood any edge in the graph which crosses its network, or just put in a router ACL to drop that pair for a while. Because the strength of the system relies on it being public and P2P, the MP/RI-AA can easily get this information.
AT&T has multiple incentives to cooperate, and can probably do it safely. It has a second party (MP/RI-AA or a company they create/contract for) do the deciding, so they dont' have the liabliity.
It keeps the content providers happy for when they are negotiating their compete-with-iTunes/Netflix video on demand and cable TV services.
It keeps the content providers from pushing through very draconian legislation, or at least draconian legislation you aren't happy with. (It can F-up your competitors, but thats just a bonus)
Its very easy to implement (short-lived router ACLs which are automatically injected and revoked).
And it drops their bandwidth bills by 30-50% by eliminating a large amount of deliberately-noncacheable (both politically and because of bittorrent encryption) traffic.
I wouldn't take it as a guarentee, but I'd almost be willing to bet that AT&T does something like this in the next year. Who wouldn't leap at a chance to reduce your costs by 30%, keep a group of "partners" you have to deal with happy, and without any real work on your part (just an SNMP-manager program)?
This won't stop closed-world pirates, but those are far less annoying to the ISPs simply because there are so many fewer of them, and less important to the MP/RI-AA because they are less likely to be users you can convert to paying customers if you make the illegal content sources unusable.
Its too late. The writing is on the wall. With almost all studios having defected to Blu-Ray primary/Blu-Ray only, anyone who's been sitting out the format war to date is not going to jump at this.
Especially since, lets face it, you'd only care about Blu-Ray/HD-DVD in the first place if you drop $1k-2k+ on the TV itself, and another $200-1K on the stereo system.
I use WPA. Why? Because on my parents network, they want to use file sharing between their desktop and their laptop. On both mine and my parents, there are networked printers.
But I write down the password on the router, and anyone who visits in person is welcome to use it.
Does Bruce not use a home printer? Share files between home computers?
A company which is trying to rake in millions by providing a "You paid more so you can skip the line" service, which promised shoe scanners etc, has to resort to trying to give a (rather small, given the need to get TSA approval) prize purse to make their business model work?
I hardly say I've drunk the OLPC cool-aid. I think it will be a failure, and the current software is a disaster.
But the hardware design really is vastly better. The hardware on the XO is brilliant. You could make a "big keyboard" version and sell it as is to the military for $2000 a copy, its that rugged.
Just some of the weaknesses of the Classmate's design in comparison (beyond the 50% higher price tag):
a: Cooling openings are a weakness and unnecessary in a device which should be passively cooled. (Dust, dirt, and debris is the enemy. Even with the faster processor, a metal slug & fins could have been put into the case. It will get dirty, even in the 1st world. I've had computers die due to dust before.) I think it is actively cooled, which makes the problem worse!
b: The keyboard is not as ruggedized. (Probably feels a lot better, however.)
c: The monitor hinge is a weakness (as I said, 1 semi-dead laptop myself).
d: The screen is no where near as good. (You can really read text on the XO's screen, ranging from full sunlight to full-dark. But an 800x480 LCD? Ick).
e: No support for mesh networking.
f: Significantly more power requirements.
g: Even the lid-closure isn't as robust, compared with the XO's design (which covers all ports but power when closed, yet ups WiFi range when open)
The OLPC's politics annoy me, the software needs real work (I'd rather run WinNT 3.1 over the current OLPC software), but the hardware really is vastly superior. Intel's hardware is a bad joke in comparison.
IF targeting the 3rd world, the classmate sucks:
a) There are cooling holes on it! Hello dirt and debris.
b) The keyboard is non ruggedized, at least compared to the XO.
c) It uses a conventional montior arrangement rather than the OLPC "behind the monitor" arrangement. This means that it has a complex, wire heavy connector through the hinge rather than just a USB and power connection.
I don't see how the classmate could last 6 months in a third world environment.
I question some of the OLPC's intent, but their hardware design blows away that Intel POS its not even funny.
The Linux Kernel GPLv2 deliberately leaves off the "or later", because that gives control of your liscence to some other entity (the FSF).
a: This is very old news, from back in October, just rehashed to get more clicks.
b: It is irrelevant. Even if Linus loved the GPLv3, there is so much code contributed to the Linux kernel without a transfer of copyright and under GPLv2 only terms that it couldn't be changed anyway.
Although I believe copyright should be much shorter (15-50 years, rather than "forever on the Installment Plan"), this reeks of "I Want A Pony". It will NEVER happen. Never. Ever. Ever.
I've yet to find a system which is easier to change the drive on than a MacBook...
Three screws and the drive carrier slides out. Four Torq screws to move the carrier to a new drive. Pop the new drive in and go. I know, I just upped mine from 80 to 160 GB.
Its the kind of repair which should never have been sent in at all, either a Do It Thyself or do it in a shop, as it really is a 5 minute job.
By making it a Trike (3 wheel) its largely counted as a motorcycle so all the crash testing requirements go out the window (including side impact, which would shread this little egg).
But the bigger worry is that trikes are far less stable in a turn, because it is at a much earlier point that they start to roll over.
The Corbin Sparrow had a real tendency to roll over. Alpina may be better by having a wider front wheel footprint, but the pod shape has a higher center of gravity. I hope either they have really REALLY good Dynamic Stability Control/Electonic Stability Control or really REALLY good laywers.
See this old Slashdot article on how servers can detect such modifications when they happen by using a bit of Javascript as an integrity checker.
(Disclaimer, I'm one of the authors of the work)
Lost in the "Oh goody, non embrionic stem cells" congradulatory bit on the part of the zealots is they forget that this is also "big step towards human cloning".
I want my clone damnit!