I never managed to understand why people keep harping about new TLDs.
While i can see the need for a directory service where one can type the same of a company and have a choice of web sites for companies with that name/service/etc. (yahoo works very well for me when i want to achieve that btw).
I apears to me that the creation of new top level TLDs is quite silly... the value today of.com is that there is not hierarchy... one may assume that a domain name ends with.com and thus it is unnecessary to query for it.
Creating new TLDs, assuming they would be succesful would defeat that value... one would be forced to figure out in which TLD a name resides.
Of corse since users do not care about new TLDs and assume that the Web site for foo is foo.com, new TLDs are largely a bust.
I've been hearing about new TLDs for years and years and i could simply never figure out what could the rational behind it be.
DNS is not a directory service, and would suck as one... assume there are 100 top level TLVs. You need a directory service to search them... then wouldn't it be simpler to start by building a directory service ?
As for alternative roots... that is an even sillier discussion. Anyone can have a DNS server that claims to be root... and that will be 100% useless unless everybody else recognises it.
DNS seems to attract all sorts of insane people... which as every certified lunatic proceed to imagine their own parallel universe with alternative root servers, etc.
The article reminds me of the CLNP mandate that european and american goverments instituted back in the early 90s. CLNP was going to replace IPv4 mostly because of two reasons:
- It had longer addresses and pretty much the same functionality
- It wasn't a DARPA project. Chauvinism is still a major factor and while the EU and Japan feel that they contributed to CLNP/IPv6 development. IPv4 was seen as an american defense project and thus politically unplesant.
If you review the american GOSIP, the IAB declaration of that CLNP was going to be the next-generation IP protocol and the european political efforts in the same direction, this seems a perfect carbon copy.
I guess than whenever clueless politicians get involved in chosing technology the results are always the same...
The economic cost of replacing IPv4 for the fashion of the moment (CLNP, IPv6) is astronomical and will not happen unless there are very significant benificts to those that pay the cost. Fact of the matter is that most people couldn't care less than the chinese do not have enought IP addresses. The internet is still centered around the US in political topology and economic terms.
I do not see US network and information service providers switching to CLNP anytime soon because there are not enought IP addresses in China. There simply is no economic justification to do so.
My major beef with IPv6 is that while it is a research toy it isn't a very interesting one... It is basically IPv4 with bigger addresses... for that we could have sticked with CLNP and TUBA the last time around.
I would say it is time the academic community stops wasting time on IPv6 and move on to research projects with real research value...
Slashdot Mirror
I never managed to understand why people keep harping about new TLDs.
.com is that there is not hierarchy... one may assume that a domain name ends with .com and thus it is unnecessary to query for it.
While i can see the need for a directory service where one can type the same of a company and have a choice of web sites for companies with that name/service/etc. (yahoo works very well for me when i want to achieve that btw).
I apears to me that the creation of new top level TLDs is quite silly... the value today of
Creating new TLDs, assuming they would be succesful would defeat that value... one would be forced to figure out in which TLD a name resides.
Of corse since users do not care about new TLDs and assume that the Web site for foo is foo.com, new TLDs are largely a bust.
I've been hearing about new TLDs for years and years and i could simply never figure out what could the rational behind it be.
DNS is not a directory service, and would suck as one... assume there are 100 top level TLVs. You need a directory service to search them... then wouldn't it be simpler to start by building a directory service ?
As for alternative roots... that is an even sillier discussion.
Anyone can have a DNS server that claims to be root... and that will be 100% useless unless everybody else recognises it.
DNS seems to attract all sorts of insane people... which as every certified lunatic proceed to imagine their own parallel universe with alternative root servers, etc.
The article reminds me of the CLNP mandate that european and american goverments instituted back in the early 90s. CLNP was going to replace IPv4 mostly because of two reasons:
- It had longer addresses and pretty much the same
functionality
- It wasn't a DARPA project. Chauvinism is still a major factor and while the EU and Japan feel that they contributed to CLNP/IPv6 development. IPv4 was seen as an american defense project and thus politically unplesant.
If you review the american GOSIP, the IAB declaration of that CLNP was going to be the next-generation IP protocol and the european political efforts in the same direction, this seems a perfect carbon copy.
I guess than whenever clueless politicians get involved in chosing technology the results are always the same...
The economic cost of replacing IPv4 for the fashion of the moment (CLNP, IPv6) is astronomical and will not happen unless there are very significant benificts to those that pay the cost.
Fact of the matter is that most people couldn't care less than the chinese do not have enought IP addresses. The internet is still centered around the US in political topology and economic terms.
I do not see US network and information service providers switching to CLNP anytime soon because there are not enought IP addresses in China. There simply is no economic justification to do so.
My major beef with IPv6 is that while it is a research toy it isn't a very interesting one...
It is basically IPv4 with bigger addresses... for that we could have sticked with CLNP and TUBA the last time around.
I would say it is time the academic community stops wasting time on IPv6 and move on to research projects with real research value...