Some TLD registries have requirements that you be a citizen of that country, follow certain guidelines or keep costs high enough to discourage domain campers.
Other registries like Verisign who manage.com really don't give a flying fuck, they have poor dispute policies and are very much in it for the money (with ICANN's approval) - their policies are "if you have money we will give you a domain"; this is bad, but still bearable because Verisign are limited to the.com TLD.
This is turning ICANN into an open top-level registry, gone will be the stringent entry requirements (you must be a country or have a good viable cause to require a TLD), they are effectively becoming a registry and registrar all in one with top-picks in the global namespace.
ICANN is an American corporation with no governmental or world-wide oversight, it's role is supposed to be technical and not commercial.
I'm thinking long-term when ICANN has re-couped the costs and have ironed out the bugs, then the price drops from $100k to $10k or even lower as ICANN see an opportunity to profit and the overall cost to business of running their own TLD falls below what they already pay.
From a technical point of view I see this as a failure. It's putting more weight on on the root name servers, how long until there are 10,000 TLDs, 100,000 of them? a million? Then there's all the people running the TLDs who may or may not have the infrastructure, technical expertise or long-term stability to properly run one.
"free market" != stable, there will be thousands of different ways to register sub-domains, some where registrations will not be allowed, some which cost excessive amounts of money and eventually whole TLDs setup to Google-bomb.
Sure it means people will be able to type in "apple", but what gives one specific company the right over a global use of that trademark name? What about "apple" print design services, "apple" car mechanics all of whom have their own trademarks in that industry.
I'm just an idealist, and would much prefer ".uk.org.mycompany.www" style domains (remember, thats how it was supposed to work in the beginning but somebody fucked up), but all I can see happening from opening up TLDs is annoyance and instability.
I'm not even going to comment on ICANN doing this for the money, capitalist pigs.
Well, ReiserFs handled this by tail packing, while traditionally you're limited to whole FS blocks with ReiserFS you can store multiple small files or the tails of files into a single block.
It's a little bit of a problem though because the FS is now doing more to write files, but from a performance perspective is very good for read performance (stat & file content often stored in the same block). That's very good for creating tar archives of large directory structures quickly which solves half the problem, but it's take longer to extract (I don't have the numbers for exactly how much longer).
Just as an example, ReiserFS + RAID1 is amazing in the right setting with less than ~10% writes, such as large content caches esp. with a large write-through cache. (*wonders what'll happen to the FS he loves now that Mr Reiser is in prison*)
The only way to get optimal performance for lots of small files is to be sequentially stored one after another, preceded by the directory structure and stat information in one big block that's loaded into memory. This is really for read-only data like uh.. tape drives, rom filesystems (seen cramfs?).
I guess what I'm trying to say is that filesystems have adapted to compromise between read & write ratios, and any that favors one too much is no longer general purpose or very useful to many people. There will be breakthroughs every now & then, and increasingly fast storage (flash? in-ram? holographic? quantum?) is invented and becomes cheaper, but there are only so many different ways you can reinvent the wheel.
I can only hope that Sun's going to move all development to the GPL branch in future.
To take Digium Asterisk as an example of a disastrous dual licensing scheme, they release a GPL version of their code and keep their proprietary version closed source (which means they can use no GPL code in it). All contributions made are given to Digium or put in the public domain - stripping the original author of his or her copyright.
So it's open source and you can take the code, branch it off and do whatever you want, but possible contributors are less willing to push upstream, and in a few instances Digium have refused to take really great innovative contributions because they use GPL code from other projects which cant be re-licensed and given to the company.
Back to the topic... it's not such an extreme situation as people can still make contributions under a dual CDDL/GPL license, but it's harmful in the long run because it greatly restricts how you can mix code/projects into Java without forking the dev effort.
I think we should get the rumor mill started with things like "actively discouraged open-source" and so on, after all Sun are doing a good thing yet it seems this one company have been holding it up with an over-zealous attitude to I.P.
The fact that he went to such great lengths to cheat his way through high school, and more importantly to pursue further education, makes me hate this person almost instantly. He's clearly not interested in working and learning, he just wants a degree so he can defraud his future employers.
We need sneaky little buggers like him, if I ran a security consulting firm I'd offer him freelance contracts and visa sponsorship if he needed.
I was just trying to push some arguments that there was no way this could happen unless Microsoft specifically had policies in place to not test with latest browsers.
The developers there are nice people if you ever meet them, and the incompetence is unlikely to be on their behalf.
With a large dev team, lots of server admins, lots of marketing and a massive user base - you mean they don't test at all with other browsers while their in beta?
The only way I could see this happening is if theres an internal policy to use _only_ IE for browsing (unlikely), or if developers were not allowed to.
I'm going to have to second this, but probably more towards the charity side.
It's quite easy for computer recycling charities to get working computers, but because of data security policies at a lot of companies they are not allowed to recycle hard-drives. This means that a disproportionate number of computers to hard-drives float around until they're finally scrapped (which overall costs the charity more time, effort and money).
For example, I have a 9gb and a 26gb drive in my main development machine - with a few 40gb and 125gb drives waiting for me to upgrade to (80%/, 67%/home used so far) - without working drives it's next to worthless and unusable.
Capcom I think released a fighting game similar to Mortal Combat in the mid 90s, and distributed with it was a code book - matte black paper with shiny black codes which would just show up all black if photocopied.
To play the game it'd say something like "Page 3 row 5 column C", then look it up in the code book and squint with the light in the right direction to get the code.
Each Vega2 processor has 48 cores, 768 cores in just 16 processors is pretty good and you can be certain a number of those are reserved for system use on such a large-scale machine; these are already fairly lightweight hardware threads and I can only presume more hardware threads per-core and you'd get some serious I/O starvation issues.
I've found the Photoshop support in the latest point releases to still have quite a few issues.
Photoshop 7 is well supported and has been for a while. CS is well supported although there are a few quirks. CS2 works well enough to be usable, but activation is broken for numerous reasons (although a solution has been worked out). CS3 doesn't work at all.
Considering I've spent a great deal of time and money on training and software, and regularly depend on the features of CS2 and CS3, only being able to use CS and Photoshop 7 is a real downer.
My housemate has a machine setup for bittorrent, when we first moved in together it was very annoying as he seemed oblivious that running it all the time meant that my connections were slow, dropping all the time & unusable.
So I spoke to him, you know - in a rational way. It's now scheduled for the nights & days when we're either asleep or at work with a few hours in between & most of the weekends where it's either throttled down to 10k/s (by uTorrent) or stopped completely.
On top of that we've got a Smoothwall box with packet prioritization for ssh/web/email/im etc. but no bandwidth throttling.
At the end of the day, if you cant come to an agreement then it's probably just gonna get worse for you two and there's nothing you can do to stop him being an asshole.
Have you also noticed a trend in growing numbers with each new technology adoption.
First it was Napster, fairly popular, but still restricted to a smallish number of users.
Then Limewire whos user base is slowly growing, although still relatively small.
Then it was Kazaa, limited to Windows computers and spammed with viruses, but still got quite a few users and was very widespread.
And now it's BitTorrent which is easy to use, available for every platform imaginable and has a massive user base.
Because of this I can't see the **AA, associated industries and local/non-us counterparts adopting it widespread until the next tech move comes along which will no doubt help them bounce back from slowing sales trends.
While at the same time they're lobbying the governments to restrict P2P, tighten up copyright laws and and generally make it difficult for anybody involved. Isn't this counterproductive?
Surely they should be encouraging new services, new ways of distributing and using content to get the ball rolling - then jump in afterwards after they've seen different techniques tried & tested and use their financial might, industry and marketing experience?
Oh sorry, what have I been smoking, I'm obviously not thinking like an estoric PHB.
That sounds similar to my experience, but both Acrobat and the freebie Acrobat Reader were dog slow.
Fairly recently I realized Evince was available, it's blazing fast and supports all the stuff I need.
I've had a few documents that were amazingly slow to read in any PDF viewer, one comes to mind that was particuarly awful, a ~300 page spec that was watermarked with 'Confidential' in big grey letters across the page. The pages would load very quickly, then you'd see some flickering and it'd go blank while the watermark was slowly displayed, then the text of the page would flicker and it'd finally be done.
However sudo & graphical equivilents will ask you for your own password before allowing the program to continue, which implicitly makes you think a little and read the message.
Now the big difference: sudo has a grace period of 5 minutes, meaning you get less of these messages, I think on my development machine I may get one or two a day if I do any sort of admin tasks - otherwise none.
Whereas Vista pops up for anything and everything, just requiring you to click a button to continue. Within 3 hours of using Vista I was blindly clicking through them. It's not normal to have UAC pop-up 4-5 times during the same install session!
Some TLD registries have requirements that you be a citizen of that country, follow certain guidelines or keep costs high enough to discourage domain campers.
Other registries like Verisign who manage .com really don't give a flying fuck, they have poor dispute policies and are very much in it for the money (with ICANN's approval) - their policies are "if you have money we will give you a domain"; this is bad, but still bearable because Verisign are limited to the .com TLD.
This is turning ICANN into an open top-level registry, gone will be the stringent entry requirements (you must be a country or have a good viable cause to require a TLD), they are effectively becoming a registry and registrar all in one with top-picks in the global namespace.
ICANN is an American corporation with no governmental or world-wide oversight, it's role is supposed to be technical and not commercial.
Can we send Bush and Cheney with him? Everybody has to die someday right...
Because there's still order to the chaos, which in this case is the TLD maintainer.
I'm thinking long-term when ICANN has re-couped the costs and have ironed out the bugs, then the price drops from $100k to $10k or even lower as ICANN see an opportunity to profit and the overall cost to business of running their own TLD falls below what they already pay.
Yeah, lets setup a server to server copy mechanism to transfer the "DNS" file with all the domain names in to keep everybody up to date.
It's almost ironic that it went from a flat hosts file being copied around, only to head back that way.
From a technical point of view I see this as a failure. It's putting more weight on on the root name servers, how long until there are 10,000 TLDs, 100,000 of them? a million? Then there's all the people running the TLDs who may or may not have the infrastructure, technical expertise or long-term stability to properly run one.
"free market" != stable, there will be thousands of different ways to register sub-domains, some where registrations will not be allowed, some which cost excessive amounts of money and eventually whole TLDs setup to Google-bomb.
Sure it means people will be able to type in "apple", but what gives one specific company the right over a global use of that trademark name? What about "apple" print design services, "apple" car mechanics all of whom have their own trademarks in that industry.
I'm just an idealist, and would much prefer ".uk.org.mycompany.www" style domains (remember, thats how it was supposed to work in the beginning but somebody fucked up), but all I can see happening from opening up TLDs is annoyance and instability.
I'm not even going to comment on ICANN doing this for the money, capitalist pigs.
Well, ReiserFs handled this by tail packing, while traditionally you're limited to whole FS blocks with ReiserFS you can store multiple small files or the tails of files into a single block.
It's a little bit of a problem though because the FS is now doing more to write files, but from a performance perspective is very good for read performance (stat & file content often stored in the same block). That's very good for creating tar archives of large directory structures quickly which solves half the problem, but it's take longer to extract (I don't have the numbers for exactly how much longer).
Just as an example, ReiserFS + RAID1 is amazing in the right setting with less than ~10% writes, such as large content caches esp. with a large write-through cache. (*wonders what'll happen to the FS he loves now that Mr Reiser is in prison*)
The only way to get optimal performance for lots of small files is to be sequentially stored one after another, preceded by the directory structure and stat information in one big block that's loaded into memory. This is really for read-only data like uh.. tape drives, rom filesystems (seen cramfs?).
I guess what I'm trying to say is that filesystems have adapted to compromise between read & write ratios, and any that favors one too much is no longer general purpose or very useful to many people. There will be breakthroughs every now & then, and increasingly fast storage (flash? in-ram? holographic? quantum?) is invented and becomes cheaper, but there are only so many different ways you can reinvent the wheel.
I can only hope that Sun's going to move all development to the GPL branch in future.
To take Digium Asterisk as an example of a disastrous dual licensing scheme, they release a GPL version of their code and keep their proprietary version closed source (which means they can use no GPL code in it). All contributions made are given to Digium or put in the public domain - stripping the original author of his or her copyright.
So it's open source and you can take the code, branch it off and do whatever you want, but possible contributors are less willing to push upstream, and in a few instances Digium have refused to take really great innovative contributions because they use GPL code from other projects which cant be re-licensed and given to the company.
Back to the topic... it's not such an extreme situation as people can still make contributions under a dual CDDL/GPL license, but it's harmful in the long run because it greatly restricts how you can mix code/projects into Java without forking the dev effort.
When you know who the company is, you can then start asking the right questions like:
iirc it might have been the midi component of the sound system, but any more information would be great.
Does anybody know who they are?
I think we should get the rumor mill started with things like "actively discouraged open-source" and so on, after all Sun are doing a good thing yet it seems this one company have been holding it up with an over-zealous attitude to I.P.
We need sneaky little buggers like him, if I ran a security consulting firm I'd offer him freelance contracts and visa sponsorship if he needed.
I was just trying to push some arguments that there was no way this could happen unless Microsoft specifically had policies in place to not test with latest browsers.
The developers there are nice people if you ever meet them, and the incompetence is unlikely to be on their behalf.
With a large dev team, lots of server admins, lots of marketing and a massive user base - you mean they don't test at all with other browsers while their in beta?
The only way I could see this happening is if theres an internal policy to use _only_ IE for browsing (unlikely), or if developers were not allowed to.
Not to mention high-end graphics software such as Autodesk Maya, Softimage XSI and Apple Shake, all of which have a very deep stronghold in Hollywood.
I interpreted it as Linux now fulfills the hard real-time requirements with these vendor-specific bundles that Sun and Microsoft don't.
I'm going to have to second this, but probably more towards the charity side.
/, 67% /home used so far) - without working drives it's next to worthless and unusable.
It's quite easy for computer recycling charities to get working computers, but because of data security policies at a lot of companies they are not allowed to recycle hard-drives. This means that a disproportionate number of computers to hard-drives float around until they're finally scrapped (which overall costs the charity more time, effort and money).
For example, I have a 9gb and a 26gb drive in my main development machine - with a few 40gb and 125gb drives waiting for me to upgrade to (80%
Capcom I think released a fighting game similar to Mortal Combat in the mid 90s, and distributed with it was a code book - matte black paper with shiny black codes which would just show up all black if photocopied.
To play the game it'd say something like "Page 3 row 5 column C", then look it up in the code book and squint with the light in the right direction to get the code.
Similar-ish idea, but much less retarded.
I've tried this and reguarly end up using Photoshop in VirtualBox and VMWare, but it's quite slow compared to Wine.
Each Vega2 processor has 48 cores, 768 cores in just 16 processors is pretty good and you can be certain a number of those are reserved for system use on such a large-scale machine; these are already fairly lightweight hardware threads and I can only presume more hardware threads per-core and you'd get some serious I/O starvation issues.
:)
How I'd love to have one of these boxes
I've found the Photoshop support in the latest point releases to still have quite a few issues.
Photoshop 7 is well supported and has been for a while.
CS is well supported although there are a few quirks.
CS2 works well enough to be usable, but activation is broken for numerous reasons (although a solution has been worked out).
CS3 doesn't work at all.
Considering I've spent a great deal of time and money on training and software, and regularly depend on the features of CS2 and CS3, only being able to use CS and Photoshop 7 is a real downer.
My housemate has a machine setup for bittorrent, when we first moved in together it was very annoying as he seemed oblivious that running it all the time meant that my connections were slow, dropping all the time & unusable.
So I spoke to him, you know - in a rational way. It's now scheduled for the nights & days when we're either asleep or at work with a few hours in between & most of the weekends where it's either throttled down to 10k/s (by uTorrent) or stopped completely.
On top of that we've got a Smoothwall box with packet prioritization for ssh/web/email/im etc. but no bandwidth throttling.
At the end of the day, if you cant come to an agreement then it's probably just gonna get worse for you two and there's nothing you can do to stop him being an asshole.
Have you also noticed a trend in growing numbers with each new technology adoption.
Because of this I can't see the **AA, associated industries and local/non-us counterparts adopting it widespread until the next tech move comes along which will no doubt help them bounce back from slowing sales trends.
While at the same time they're lobbying the governments to restrict P2P, tighten up copyright laws and and generally make it difficult for anybody involved. Isn't this counterproductive?
Surely they should be encouraging new services, new ways of distributing and using content to get the ball rolling - then jump in afterwards after they've seen different techniques tried & tested and use their financial might, industry and marketing experience?
Oh sorry, what have I been smoking, I'm obviously not thinking like an estoric PHB.
To put this into perspect, in the UK people are starting to complain that it's costing them £80 a time to fill up, thats nearly $160 USD.
We don't have so much of a car culture or the need to drive obscene distances every day, so it's much less of a problem.
That sounds similar to my experience, but both Acrobat and the freebie Acrobat Reader were dog slow.
Fairly recently I realized Evince was available, it's blazing fast and supports all the stuff I need.
I've had a few documents that were amazingly slow to read in any PDF viewer, one comes to mind that was particuarly awful, a ~300 page spec that was watermarked with 'Confidential' in big grey letters across the page. The pages would load very quickly, then you'd see some flickering and it'd go blank while the watermark was slowly displayed, then the text of the page would flicker and it'd finally be done.
However sudo & graphical equivilents will ask you for your own password before allowing the program to continue, which implicitly makes you think a little and read the message.
Now the big difference: sudo has a grace period of 5 minutes, meaning you get less of these messages, I think on my development machine I may get one or two a day if I do any sort of admin tasks - otherwise none.
Whereas Vista pops up for anything and everything, just requiring you to click a button to continue. Within 3 hours of using Vista I was blindly clicking through them. It's not normal to have UAC pop-up 4-5 times during the same install session!