Yeah I think you're right... the client seems to be showing a completely random list of people, one is exactly opposite me on the globe and I'm a bit surprised that no-one on the Interational Space Station is on my list as being a close neighbour:)
As I said really - good idea. Unbelievably bad implementation. I expect it was designed for wireless networks and they haven't really thought through the wired network side of it.
The idea is sound, but the implementation is flawed. I mainly seem to get people in the US, even though I'm in the UK (6000+ miles is hardly close) and nobody on the list is in my country even though my friend about a mile away has it installed and running.
In my experience, trying to guess where people are by their IP address doesn't work very well. It would work much better if you could simply add the locations you are at most often by country/postcode or even just grid reference and it used those instead. Even people on the same subnet as me could be hundreds of miles away if they're dialling in. If people have sold of or subleased blocks of IP addresses to other countries, the records could even indicate the wrong location or wrong country anyway - so it may always think I'm in Finland when actually I'm in England... I'm not impressed.
I wrote an IM client myself which simply discovered people on the same subnet using broadcasts - and even that seemed to be more effective at finding my friends and colleagues than this (admittedly broadcasts don't usually work on public networks).
Can someone please answer this really obvious question? What is the point? With all the disadvantages mentionned above (stealing card details without you knowing, problems with multiple cards in a wallet meaning you have to take it out anyway etc etc) why would you want a proximity card anyway? There isn't any problem with normal chip cards as they are... (except they don't require a pin or anything yet). Why not just increase the security of normal chip cards first, which could probably be done without issuing anyone with a new card. The proximity concept doesn't solve any problems at all and simply adds more new ones.
> Physical, hardly. > Have you ever purchased anything online?
Yes, I seem to recall needing to physically see my card to do it and enter the numbers on a keyboard. The site did not simply sense the card in my wallet from a pop-up window and start charging things to it.
> All I need is your number, name and expiry and I can charge > your account all I want.
And how will you get those without seeing something with my card details on it (like my card)?
The argument here is that just walking past something/someone carrying a proximity reader could steal all the details off your card and possibly start using it unless it's also combined with some kind of compulsory PIN.
I presume the insertion of "IP" in the title of this article was a mistake or assumption made by a naeive author? You don't use IP to carry telephone calls on a phone network. Ever. IP is no good for carrying voice data and there are many better protocols around which were designed for this purpose. I presume that they really mean ATM or some other voice protocol? You need a small packet, circuit based protocol to handle large numbers of voice calls efficiently. Although IP could be made to work, it would be pretty difficult and is essentially rather like trying to put a sealed bus in a sea to try and make a ferry - why not just use a boat?
> In reality I think dotnet is what everyone thinks, a competitor > to Java.
Er, not *everyone* thinks it's a competitor to Java. Sure it has similarities, but that doesn't mean to say it's trying to compete with it. It's not even multi-platform - which is half the point of Java. In fact I've found that it's mainly slashdot readers who think like this. Most of the rest of the IT world think it's a very good new platform.
> How many highgrade professional games are written in Java currently?
None, but that's because Java is SLOW AS HELL..NET isn't slow - it's often faster than C/C++ in fact (although probably using more memory). Look at some benchmarks if you don't believe me, or at least run a big application and try and imagine the same thing in Java. Even Visual Studio.NET itself is written in C# and it's no slower than version 6 which is C++. I think you can even generate native binaries if you're really worried about the speed of your.NET app (util called ngen.exe).
I'm very sure that we'll start to see.NET games coming out quite soon (now that DirectX 9 supports.NET), and I'm pretty sure that most types of games won't suffer a performance hit. If they do, it's not hard to link to C++ code for the bits which need more optimisation (MMX etc).
I would have thought it was the other way round, if you've got three chips - you need to mix the three channels back together again. If it's all coming off one chip, it would all be done in the same place and properly calibrated?!
If three is better, how come really decent 6 megapixel+ digital SLR still cameras only have one CCD and get stunning colour accuracy?
> other way around bro. 1080 is horizontal, 1920 is vertical.
Nope... There are 1080 horizontal scan lines (making the vertical resolution 1080) and 1920 pixels horizontal resolution (you never refer to the horizontal resolution in lines).
I agree. I'm sure that if you want to edit your movies, you'd be better off with a normal Motion-JPEG camcorder that doesn't lose any quality if you edit it (each frame is stored seperately as a JPEG without relying on previous frames). Editing MPEG - especially if you cut it into small chunks, is going to considerably degrade the quality each time you mess with it. With M-JPEG it can go back on the tape in the exact same quality you recorded it in (unless you applied any weird effects that actually modified the frames -eg a crossfade or title).
Er, excuse my naievity, but what's wrong with a "single" ccd? I thought they only used multiple CCDs because they couldn't make a single one of that resolution fast enough. Now that technology has improved and you can have a high-res CCD with decent colour, surely thats BETTER than 3 joined together with dodgy prisms?
I certainly know that my latest digital video camera that only has one, is much better quality than my old one that had three...
> how the hackers so easily found the WLANs mail server once inside the network
It doesn't imply that they did find the WLAN's mail server (if it even had one). They just needed the Internet connection to anonymise themselves and used any existing open-relay on the Internet to send the mail, or more likely - they just sent them directly without using any SMTP server at all (a lot of spamming software seems to mail directly now if you look at the headers - with no intermediate relays).
All you need to do to send a mail is resolve the MX record for the domain name in question to get the IP address of the receiving host, then just open a TCP connection on port 25 directly to that machine straight from the computer that sent the mail. This is the most reliable way to deliver a message.
We (UK) don't however, have the ability to take landline numbers to mobile networks or vice versa as suggested by the article (not permanently anyway - you can redirect calls if you want but you end up paying for incoming calls)....which I'm very glad about because that would be a totally stupid idea. How on earth would you know if you're calling a landline (1p/min or perhaps free) or a mobile (10p/min or much more depending on your tariff). I can't believe they're even suggesting this. It's hard enough knowing whether you're calling someone on your own network, let alone the huge annoyance/expense of calling a mobile when you thought you were calling a landline.
Camcorder - where on earth did he get that from?! Did the guy that submitted that actaully read the article he was submitting?! It doesn't mention that in the article - mainly because it would be a stupid idea. That's maxiumum overkill if I've ever heard of it. That would be like using a camcorder to 'look' and see if the fridge door is shut!
And why use a piston to change the swing? What's wrong with an electromagnet which wouldn't need to actually touch the pendlum?
Windows XP even comes with the runtime pre-installed! If there weren't any platforms for it, nobody would be developing for it, but as it happens, hundreds of thousands of software developers are currently using.NET (language doesn't matter with.NET - it's the framework that counts).
At my university (graduated 1999) Classes tought with Java: 1 Classes tought with C/Prolog/Lisp/Perl/SML: 10
my point: since when did what you were taught at uni make any difference to what you use at your job?! Uni's use academic languages for teaching about languages, and pascal/c/c++ for teaching about software development. Their choice for the later is academic (if you'll excuse the pun). Hell - if you were taught Java and can't pick up C#, you've got to be the thickest software developer I've ever come across. They're almost identical - C# is just a neater version of Java with the annoyances polished out. Oh and it's a standard.
Slashdot Mirror
Yeah I think you're right... the client seems to be showing a completely random list of people, one is exactly opposite me on the globe and I'm a bit surprised that no-one on the Interational Space Station is on my list as being a close neighbour :)
As I said really - good idea. Unbelievably bad implementation. I expect it was designed for wireless networks and they haven't really thought through the wired network side of it.
Nick...
The idea is sound, but the implementation is flawed. I mainly seem to get people in the US, even though I'm in the UK (6000+ miles is hardly close) and nobody on the list is in my country even though my friend about a mile away has it installed and running.
In my experience, trying to guess where people are by their IP address doesn't work very well. It would work much better if you could simply add the locations you are at most often by country/postcode or even just grid reference and it used those instead. Even people on the same subnet as me could be hundreds of miles away if they're dialling in. If people have sold of or subleased blocks of IP addresses to other countries, the records could even indicate the wrong location or wrong country anyway - so it may always think I'm in Finland when actually I'm in England... I'm not impressed.
I wrote an IM client myself which simply discovered people on the same subnet using broadcasts - and even that seemed to be more effective at finding my friends and colleagues than this (admittedly broadcasts don't usually work on public networks).
Nick...
Can someone please answer this really obvious question? What is the point? With all the disadvantages mentionned above (stealing card details without you knowing, problems with multiple cards in a wallet meaning you have to take it out anyway etc etc) why would you want a proximity card anyway? There isn't any problem with normal chip cards as they are... (except they don't require a pin or anything yet). Why not just increase the security of normal chip cards first, which could probably be done without issuing anyone with a new card. The proximity concept doesn't solve any problems at all and simply adds more new ones.
Nick...
> Physical, hardly.
> Have you ever purchased anything online?
Yes, I seem to recall needing to physically see my card to do it and enter the numbers on a keyboard. The site did not simply sense the card in my wallet from a pop-up window and start charging things to it.
> All I need is your number, name and expiry and I can charge
> your account all I want.
And how will you get those without seeing something with my card details on it (like my card)?
The argument here is that just walking past something/someone carrying a proximity reader could steal all the details off your card and possibly start using it unless it's also combined with some kind of compulsory PIN.
Nick...
I presume the insertion of "IP" in the title of this article was a mistake or assumption made by a naeive author? You don't use IP to carry telephone calls on a phone network. Ever. IP is no good for carrying voice data and there are many better protocols around which were designed for this purpose. I presume that they really mean ATM or some other voice protocol? You need a small packet, circuit based protocol to handle large numbers of voice calls efficiently. Although IP could be made to work, it would be pretty difficult and is essentially rather like trying to put a sealed bus in a sea to try and make a ferry - why not just use a boat?
Nick...
It's alright - if you don't want to attach it to a PSU, you can connect it directly to an old lathe as shown the article photos.
> "If English was good enough for Jesus, it's good enough for everyone else."
Jesus didn't speak English. Is that supposed to be funny somehow?
> In reality I think dotnet is what everyone thinks, a competitor
.NET isn't slow - it's often faster than C/C++ in fact (although probably using more memory). Look at some benchmarks if you don't believe me, or at least run a big application and try and imagine the same thing in Java. Even Visual Studio.NET itself is written in C# and it's no slower than version 6 which is C++. I think you can even generate native binaries if you're really worried about the speed of your .NET app (util called ngen.exe).
.NET games coming out quite soon (now that DirectX 9 supports .NET), and I'm pretty sure that most types of games won't suffer a performance hit. If they do, it's not hard to link to C++ code for the bits which need more optimisation (MMX etc).
> to Java.
Er, not *everyone* thinks it's a competitor to Java. Sure it has similarities, but that doesn't mean to say it's trying to compete with it. It's not even multi-platform - which is half the point of Java. In fact I've found that it's mainly slashdot readers who think like this. Most of the rest of the IT world think it's a very good new platform.
> How many highgrade professional games are written in Java currently?
None, but that's because Java is SLOW AS HELL.
I'm very sure that we'll start to see
Nick... (ASP.NET developer)
I would have thought it was the other way round, if you've got three chips - you need to mix the three channels back together again. If it's all coming off one chip, it would all be done in the same place and properly calibrated?!
If three is better, how come really decent 6 megapixel+ digital SLR still cameras only have one CCD and get stunning colour accuracy?
> other way around bro. 1080 is horizontal, 1920 is vertical.
Nope... There are 1080 horizontal scan lines (making the vertical resolution 1080) and 1920 pixels horizontal resolution (you never refer to the horizontal resolution in lines).
I agree. I'm sure that if you want to edit your movies, you'd be better off with a normal Motion-JPEG camcorder that doesn't lose any quality if you edit it (each frame is stored seperately as a JPEG without relying on previous frames). Editing MPEG - especially if you cut it into small chunks, is going to considerably degrade the quality each time you mess with it. With M-JPEG it can go back on the tape in the exact same quality you recorded it in (unless you applied any weird effects that actually modified the frames -eg a crossfade or title).
Er, excuse my naievity, but what's wrong with a "single" ccd? I thought they only used multiple CCDs because they couldn't make a single one of that resolution fast enough. Now that technology has improved and you can have a high-res CCD with decent colour, surely thats BETTER than 3 joined together with dodgy prisms?
I certainly know that my latest digital video camera that only has one, is much better quality than my old one that had three...
Nick...
Anyone that thinks OpenOffice is good enough to deploy on that scale hasn't used OpenOffice.
Anyone that mods this down probably also hasn't used OpenOffice.
Remember: OpenSource is only free if your time has no value...
up to 100k narrowband
100k - 500k mediumband
500k+ broadband...
I certainly wouldn't call anything less than 500k broadband.
Nick..
> how the hackers so easily found the WLANs mail server once inside the network
It doesn't imply that they did find the WLAN's mail server (if it even had one). They just needed the Internet connection to anonymise themselves and used any existing open-relay on the Internet to send the mail, or more likely - they just sent them directly without using any SMTP server at all (a lot of spamming software seems to mail directly now if you look at the headers - with no intermediate relays).
All you need to do to send a mail is resolve the MX record for the domain name in question to get the IP address of the receiving host, then just open a TCP connection on port 25 directly to that machine straight from the computer that sent the mail. This is the most reliable way to deliver a message.
Nick...
We (UK) don't however, have the ability to take landline numbers to mobile networks or vice versa as suggested by the article (not permanently anyway - you can redirect calls if you want but you end up paying for incoming calls). ...which I'm very glad about because that would be a totally stupid idea. How on earth would you know if you're calling a landline (1p/min or perhaps free) or a mobile (10p/min or much more depending on your tariff). I can't believe they're even suggesting this. It's hard enough knowing whether you're calling someone on your own network, let alone the huge annoyance/expense of calling a mobile when you thought you were calling a landline.
Nick...
I know what I'd put in: the Internet. ...or at least - the copy of the Internet that google has :)
shaking in the *cold* I meant. DOH!
Time for 1st coffee...
Nick...
Does anyone else not quite get the point of an SDK for an opensource product?
:)
The product *is* the SDK!
Nick...
Camcorder - where on earth did he get that from?! Did the guy that submitted that actaully read the article he was submitting?! It doesn't mention that in the article - mainly because it would be a stupid idea. That's maxiumum overkill if I've ever heard of it. That would be like using a camcorder to 'look' and see if the fridge door is shut!
And why use a piston to change the swing? What's wrong with an electromagnet which wouldn't need to actually touch the pendlum?
Nick...
PS. spelling of "tought" copied so even *you* can understand.
It is of course, "taught".
If that's not screaming for a troll/flaimbait then I don't know what is....
ALL recent versions of Windows support C#!
Even Linux supports C# though the mono project.
Windows XP even comes with the runtime pre-installed! If there weren't any platforms for it, nobody would be developing for it, but as it happens, hundreds of thousands of software developers are currently using
At my university (graduated 1999)
Classes tought with Java: 1
Classes tought with C/Prolog/Lisp/Perl/SML: 10
my point: since when did what you were taught at uni make any difference to what you use at your job?! Uni's use academic languages for teaching about languages, and pascal/c/c++ for teaching about software development. Their choice for the later is academic (if you'll excuse the pun). Hell - if you were taught Java and can't pick up C#, you've got to be the thickest software developer I've ever come across. They're almost identical - C# is just a neater version of Java with the annoyances polished out. Oh and it's a standard.
Nick...
Looks like CmdrTaco *is* the April Fool!
Looks like CmdrTaco *is* the April Fool!