Of course, IPv4 is just as capable as IPv6 of doing QoS, and in fact most provider IP networks outside the Internet already offer DiffServ QoS. The 'IPv6 enables QoS' myth is almost as strong as the 'IPv6 has better security' one - IPSec was invented for IPv6, but it's been deployed widely on IPv4.
Fortunately, the explosion in home networks, peer to peer, WiFi laptops and smartphones (particularly 3G) will be enough to make IPv6 happen, sooner or later.
Skype already does this, and is of course growing at an amazing rate (see skype.com) - because the signalling and voice data may go through another computer or two before reaching the final computer, all calls are encrypted end to end. Will be interesting to see how the lawmakers deal with Skype (which is Luxembourg based and hence not governed by US regulations, unless it sets up a US subsidiary).
Accents are spoken - you are talking about a dialect or variant of written English. I'm sure most Slashdot readers are smart enough to adapt to minor variations like that, just as the Brits somehow manage to adapt to the way Americans 'use' English...
Once you are on digital TV, whether cable, satellite or broadcast (terrestrial), TiVo's inability to tune into one TV channel while recording on another is very painful. It's really the digital set top box's fault - in the UK, Sky+ is a TiVo like device from the Sky satellite TV service that has dual digital tuners - but TiVo really suffers as a result, particularly in households where one person likes watching live TV and the other person wants to watch recorded stuff...
The lack of a wired link for remote commands between the STB and the TiVo is also painful - every now and then the Sky box switched itself off on receiving a channel change command, so I had to slow down the IR sending even more. Channel changing is truly glacial, even though the Sky remote can change channels incredibly fast - there've been few useful upgrades to the TiVo software for things like this.
I have an original Series 1 TiVo, and it's been incredibly slow for some time - the only thing that sped it up was deleting a huge number of season passes. I find it hard to work out why it's so incredibly slow - I suspect really bad database design, indexing and/or in-memory algorithms, as the amount of data it has to search is not that enormous. Intelligent algorithms and on-disk indexes should make it go a lot more quickly...
They should also support SCART connections for the European market, as well as PAL. Also, I couldn't work out why you had to select Composite Video vs. 5.1 sound at order time - these are orthogonal options (Composite Video doesn't include sound, I believe) and are normally configurable.
BT is already implementing this with Vodafone - basically, you take your mobile phone home and it starts using your land line, via Bluetooth.
Bluetooth is probably a better solution for battery life, at least at the moment, and a huge number of European phones support it. Bluetooth headsets stopped being a geek accessory about two years ago and are now used by a lot of people, particularly when driving.
I bought and junked a Philips Pronto Neo universal remote and have now got a Harmony H688 which works brilliantly - it has a web-based programming application that gets smarter as more people teach it about their equipment, and it involved far less effort than the Philips to get it working. Effort to get a remote working is probably the key factor for most people - even though I'm something of a geek I draw the line at spending tens of hours programming a remote, though I did spend about 4 hours tweaking and testing.
Somebody mod the parent up! There's a lot of mileage in simply switching Windows users to Firefox and Thunderbird as a way of getting into open source, and the reaction mentioned shows there's real demand for this.
This is a smaller step than going wholesale into Linux, and will lead to a certain proportion of Windows Mozilla users going for Linux as a result. More Mozilla usage will also reduce the number of worm-ridden PCs which is a benefit to everyone, whatever their OS.
Actually the deadline is this Thurs, July 15th, so they have a bit more time to finish it. Already looks like a good application - not much more needed in the sections I looked at.
I'm not sure how IPv6 would help with this... It sounds like you are talking about creating small multicast groups, or perhaps anycast, but this would also require an entirely new protocol, not just HTTP, to serve such pages. Multicast is not yet enabled in the IPv4 Internet because of its interesting security/management issues (imagine how easy it is to DoS a multicast) and billing problems (do you charge the same as unicast, even though one multicast could tie up thousands of links?). IPv6 will have the same issues as far as I can see. This is the first time I've heard of this scheme, so any pointers are welcome.
Something along the lines of P2P networking such as BitTorrent or Freenet, which automatically distributes content throughout the network, seems to me a better option.
IPv6 will arrive because of home networking and VoIP problems with NAT (though Skype handles NAT so well that may not be a driver in the end), and of course 3G and the gradual demand for IPv6 in Asia. I don't think schemes like this will drive IPv6 adoption, because they only work when mass-adopted; the real IPv6 adoption drivers work because 'islands' of IPv6 can get benefits from talking to other islands even though IPv4 remains predominant.
This is about e-commerce fraud - so all that is needed is for e-commerce sites to refuse to ship to Macedonian addresses. This would enable Macedonian internet users to still use the Web, email, etc, but just not do e-commerce with such sites.
Blocking the entire IP range is an extreme and stupid response considering it's so easy to just not do business with countries with high fraud rates.
See this eWeek article which says IE is too dangerous to keep using. Strong stuff from a mainstream publication - the bit about people potentially losing online banking and stock trading passwords is probably teh most effective at getting people to switch.
Auto replies would also get your address marked as 'confirmed valid' i.e. able to receive emails, even if you don't read the spam, so you'll probably just get even more spam.
Did you turn on SpamAssassin's Bayesian filtering? I found that this is generally good as long as you train it on enough ham and spam. SpamAssassin uses Bayesian filters, rule-based filters, black lists and Razor style services, so it's generally proof against spam that gets through several of these defences. I get something like a 99% hit rate these days.
I use SpamAssassin on my mail server, because it has Bayesian filtering, rules-based filtering (not just keywords but regexes and code to detect things such as use of Base64 encoding), and collaborative filtering (e.g. Razor).
Bayesian filters generally work OK, but every day there are a few spams that append some irrelevant text (a joke or just random unusual words) that would beat Bayesian filters if not for the rules-based filters. While they are less effective spams as you point out, there may still be people who click on the URLs and buy something, so this doesn't mean the death of spam. In fact only ISP-based filtering can really affect spam, since there are enough clueless people with no client side filtering that spam will get some responses.
SpamAssassin's integrated approach means that, with quite a few custom rules, I can get upwards of 99% accuracy and the only false positives are new commercial emails from e-commerce sites.
The solution is to install a firewall from CD before connecting to the Internet - unlike anti-virus software, firewalls don't really need updating, so the fact you are installing a CD version doesn't matter. A relative of mine runs an NT4 PC that doesn't have the latest security updates, but that doesn't matter since she has a firewall, so she has never been hit by a worm.
Then, install and update the anti-virus software, and run Windows Update, before using IE or Outlook for anything else.
And of course, using Mozilla Firefox and Thunderbird will avoid one class of future infections.
Non-greediness is not just an implementation detail, it is regex language feature (so to speak) that lets you write simpler regexes even if the implementation is greedy. You may well say that regexes are a mathematical concept, but in the real world it's useful to be able to write more expressive regexes.
egrep is rather old in regex terms, being invented perhaps 20+ years ago. Should regexes remain forever with a 1970s feature set, or is it OK to advance this? I vote for enhanced regexes, in which case Perl is one of the most powerful, though Java, Python and.NET have some nice features too. See http://regex.info for more info on a useful book.
Perl regexps are just more powerful than most - that's why many other languages have Perl-compatible regex support available as libraries or in the language (e.g. PCRE). To you it's cryptic; to people who know Perl, it's more powerful and more readable. See regex.info for Jeffrey Friedl's great book on regexes which covers many languages including Perl.
Perl is definitely more capable in regexes than most languages, and having regexes in the language makes it easier to do regex-heavy tasks, or use them where most languages would support imperative code.
Since Perl's regex support is so strong, it enables commented multi-line regexes using the '/x' modifier at the end of the regex, e.g. (with broken indentation and comment alignment due to Slash/HTML...):
$str =~/ (.*?) \& # Match zero or more characters up to '&', non-greedy (\w+) \( # Match one or more word characters up to '(' (.*?) \) # Match zero or more characters up to '(', non-greedy /gisx;
This is not so quick to type, but still less than the equivalent non-regex code typically, and makes it easier to see what is going on a year later. Perl also has great regex debugging available.
Perl is a sharp tool - it enables you to write very concise and unreadable code, just as it enables you to write slightly more verbose and more readable code.
With Perl 6 and its Parrot virtual machine implementation, users of Python and other languages should also be able to access the thousands of CPAN modules, bringing the amazing reach of CPAN to a wider audience that doesn't necessarily want the Perl syntax.
Slashdot Mirror
TWiki has a front-end for vim, so an Emacs frontend using a similar approach would not be hard - see this TWiki page on the subject.
Good coverage of exactly how Microsoft 'helped' Citrix in the thin-client market, and probably really helped thin-client Linux...
Kings get crowned, not 'coronated'. If this was a Wiki I could correct your post directly, of course...
Of course, IPv4 is just as capable as IPv6 of doing QoS, and in fact most provider IP networks outside the Internet already offer DiffServ QoS. The 'IPv6 enables QoS' myth is almost as strong as the 'IPv6 has better security' one - IPSec was invented for IPv6, but it's been deployed widely on IPv4.
Fortunately, the explosion in home networks, peer to peer, WiFi laptops and smartphones (particularly 3G) will be enough to make IPv6 happen, sooner or later.
Skype already does this, and is of course growing at an amazing rate (see skype.com) - because the signalling and voice data may go through another computer or two before reaching the final computer, all calls are encrypted end to end. Will be interesting to see how the lawmakers deal with Skype (which is Luxembourg based and hence not governed by US regulations, unless it sets up a US subsidiary).
Accents are spoken - you are talking about a dialect or variant of written English. I'm sure most Slashdot readers are smart enough to adapt to minor variations like that, just as the Brits somehow manage to adapt to the way Americans 'use' English...
Once you are on digital TV, whether cable, satellite or broadcast (terrestrial), TiVo's inability to tune into one TV channel while recording on another is very painful. It's really the digital set top box's fault - in the UK, Sky+ is a TiVo like device from the Sky satellite TV service that has dual digital tuners - but TiVo really suffers as a result, particularly in households where one person likes watching live TV and the other person wants to watch recorded stuff...
The lack of a wired link for remote commands between the STB and the TiVo is also painful - every now and then the Sky box switched itself off on receiving a channel change command, so I had to slow down the IR sending even more. Channel changing is truly glacial, even though the Sky remote can change channels incredibly fast - there've been few useful upgrades to the TiVo software for things like this.
I have an original Series 1 TiVo, and it's been incredibly slow for some time - the only thing that sped it up was deleting a huge number of season passes. I find it hard to work out why it's so incredibly slow - I suspect really bad database design, indexing and/or in-memory algorithms, as the amount of data it has to search is not that enormous. Intelligent algorithms and on-disk indexes should make it go a lot more quickly...
They should also support SCART connections for the European market, as well as PAL. Also, I couldn't work out why you had to select Composite Video vs. 5.1 sound at order time - these are orthogonal options (Composite Video doesn't include sound, I believe) and are normally configurable.
BT is already implementing this with Vodafone - basically, you take your mobile phone home and it starts using your land line, via Bluetooth.
Bluetooth is probably a better solution for battery life, at least at the moment, and a huge number of European phones support it. Bluetooth headsets stopped being a geek accessory about two years ago and are now used by a lot of people, particularly when driving.
For more information, see:
Somebody mod the parent up! There's a lot of mileage in simply switching Windows users to Firefox and Thunderbird as a way of getting into open source, and the reaction mentioned shows there's real demand for this.
This is a smaller step than going wholesale into Linux, and will lead to a certain proportion of Windows Mozilla users going for Linux as a result. More Mozilla usage will also reduce the number of worm-ridden PCs which is a benefit to everyone, whatever their OS.
Actually the deadline is this Thurs, July 15th, so they have a bit more time to finish it. Already looks like a good application - not much more needed in the sections I looked at.
I'm not sure how IPv6 would help with this... It sounds like you are talking about creating small multicast groups, or perhaps anycast, but this would also require an entirely new protocol, not just HTTP, to serve such pages. Multicast is not yet enabled in the IPv4 Internet because of its interesting security/management issues (imagine how easy it is to DoS a multicast) and billing problems (do you charge the same as unicast, even though one multicast could tie up thousands of links?). IPv6 will have the same issues as far as I can see. This is the first time I've heard of this scheme, so any pointers are welcome.
Something along the lines of P2P networking such as BitTorrent or Freenet, which automatically distributes content throughout the network, seems to me a better option.
IPv6 will arrive because of home networking and VoIP problems with NAT (though Skype handles NAT so well that may not be a driver in the end), and of course 3G and the gradual demand for IPv6 in Asia. I don't think schemes like this will drive IPv6 adoption, because they only work when mass-adopted; the real IPv6 adoption drivers work because 'islands' of IPv6 can get benefits from talking to other islands even though IPv4 remains predominant.
This is about e-commerce fraud - so all that is needed is for e-commerce sites to refuse to ship to Macedonian addresses. This would enable Macedonian internet users to still use the Web, email, etc, but just not do e-commerce with such sites.
Blocking the entire IP range is an extreme and stupid response considering it's so easy to just not do business with countries with high fraud rates.
See this eWeek article which says IE is too dangerous to keep using. Strong stuff from a mainstream publication - the bit about people potentially losing online banking and stock trading passwords is probably teh most effective at getting people to switch.
Auto replies would also get your address marked as 'confirmed valid' i.e. able to receive emails, even if you don't read the spam, so you'll probably just get even more spam.
Did you turn on SpamAssassin's Bayesian filtering? I found that this is generally good as long as you train it on enough ham and spam. SpamAssassin uses Bayesian filters, rule-based filters, black lists and Razor style services, so it's generally proof against spam that gets through several of these defences. I get something like a 99% hit rate these days.
I use SpamAssassin on my mail server, because it has Bayesian filtering, rules-based filtering (not just keywords but regexes and code to detect things such as use of Base64 encoding), and collaborative filtering (e.g. Razor).
Bayesian filters generally work OK, but every day there are a few spams that append some irrelevant text (a joke or just random unusual words) that would beat Bayesian filters if not for the rules-based filters. While they are less effective spams as you point out, there may still be people who click on the URLs and buy something, so this doesn't mean the death of spam. In fact only ISP-based filtering can really affect spam, since there are enough clueless people with no client side filtering that spam will get some responses.
SpamAssassin's integrated approach means that, with quite a few custom rules, I can get upwards of 99% accuracy and the only false positives are new commercial emails from e-commerce sites.
The solution is to install a firewall from CD before connecting to the Internet - unlike anti-virus software, firewalls don't really need updating, so the fact you are installing a CD version doesn't matter. A relative of mine runs an NT4 PC that doesn't have the latest security updates, but that doesn't matter since she has a firewall, so she has never been hit by a worm.
Then, install and update the anti-virus software, and run Windows Update, before using IE or Outlook for anything else.
And of course, using Mozilla Firefox and Thunderbird will avoid one class of future infections.
Non-greediness is not just an implementation detail, it is regex language feature (so to speak) that lets you write simpler regexes even if the implementation is greedy. You may well say that regexes are a mathematical concept, but in the real world it's useful to be able to write more expressive regexes.
egrep is rather old in regex terms, being invented perhaps 20+ years ago. Should regexes remain forever with a 1970s feature set, or is it OK to advance this? I vote for enhanced regexes, in which case Perl is one of the most powerful, though Java, Python and .NET have some nice features too. See http://regex.info for more info on a useful book.
Perl regexps are just more powerful than most - that's why many other languages have Perl-compatible regex support available as libraries or in the language (e.g. PCRE). To you it's cryptic; to people who know Perl, it's more powerful and more readable. See regex.info for Jeffrey Friedl's great book on regexes which covers many languages including Perl.
Perl already supports spaced-out multi-line regular expressions including comments - see my earlier comment in this thread