When Bill Thompson first started writing for BBC Online I thought it was cool, he seemed to actually know what he was talking about and convey it in a manner to get the masses informed.
Then all this stuff about government control started up.
I've now decided not to even read any of his future articles (never anything new in them given how much I read other sources of tech news), else my blood pressure will go through the roof.
Exactly what I'd been wondering. I can't see any reason for this not to work with things like clothes, but it would be a problem for any electronic goods...
Just because Chandra stated that was why HAL 'went nuts' doesn't mean it was, or was the sole cause. Real-world engineers aren't right 100% of the time.
What? Me getting narky at nitpickers of TV/films because they take anything stated by a character as 100% fact?
I think you didn't read the story. "3 times that of our solar system" or 17 light-hours refers to the radius of the star's orbit. That means that the radius of interior object is at most as large as the radius orbit, or the star would "get stuck" in the dence material.
Um, no:
The orbital attributes mean that the entire mass of the interior object, between 2.6 million and 3.7 million times that of the sun, is crammed within a space about three times the size of our solar system.
So, whilst what you said might be what they MEANT, it's not what they said.
On the other hand, I could say that the entire mass of our solar system is enclosed within a space 1000000000000 times the radius of the Earth's orbit, and be correct;).
-Ath
I don't think anyone's mentioned this yet (I did a quick search for a few key words).
You still need payphones for those people who may well have a cellphone, but it only works way back in the UK where they're from.
With the way cellphone coverage in the USA works, to my understanding, the same applies to someone who's simply hopped over to the next state to visit friends or whatever.
The whole thing is merely pointing out that some IP stacks don't give anything like enough randomness between connections (attempts) from different hosts.
The vendors of such stacks need to get a clue[tm].
There's not just hijacking the connect. If you know the necessary sequence number you could spoof a RST packet to make the connect drop, voila DoS.
Re:Don't make broadband suck to satisfy modemers!
on
The Modem Lives On
·
· Score: 1
Yes, drop prediction _for deciding on hits/kills_. You still use prediction to try and give each player a better view of things at that instant, but they know not to 100% rely on it being reality. It's better than seeing players freeze up anyway.
I may have a lovely 15ms ping to Barrysworld but I often play on upto 250ms ping and 10-20% packet loss for my Q3F clan, so I do experience it both ways.
Why should a modemer need slightly less _skill_ to kill me just because he gets greater compensation due to his higher ping?
Don't make broadband suck to satisfy modemers!
on
The Modem Lives On
·
· Score: 1
I recently took up playing CounterStrike so checked lots of sites for tips on tweaking HL performance. Whilst doing this I came across some info about the netcode in 1.1.0.1. Apparently Valve did in fact make it easier for modems, at the expense of broadband connections.
What they did was allow some of the client-side prediction from higher-ping players to over-ride what the server would have thought happened. Thus a broadband player could have gotten a kill denied that he SHOULD have gotten 'cos the prediction on the modemer's client said he got out of the way in time. Likewise a broadband user can get killed by a modemer due to this.
IMO this plain sucks. Leave it with things decided purely at the server and if someone wants 'fairer' get a better connection.
-Ath
Slashdot Mirror
When Bill Thompson first started writing for BBC Online I thought it was cool, he seemed to actually know what he was talking about and convey it in a manner to get the masses informed.
Then all this stuff about government control started up.
I've now decided not to even read any of his future articles (never anything new in them given how much I read other sources of tech news), else my blood pressure will go through the roof.
AIUI the problem is that this relies on you using segmented memory rather than whatever alternative model it is Linux et al use, right ?
So, you can do it on x86, but all current OSes have chosen not to use segmented memory for other reasons (I believe performance is a big one).
Exactly what I'd been wondering. I can't see any reason for this not to work with things like clothes, but it would be a problem for any electronic goods...
Just because Chandra stated that was why HAL 'went nuts' doesn't mean it was, or was the sole cause. Real-world engineers aren't right 100% of the time.
What? Me getting narky at nitpickers of TV/films because they take anything stated by a character as 100% fact?
I don't think anyone's mentioned this yet (I did a quick search for a few key words).
You still need payphones for those people who may well have a cellphone, but it only works way back in the UK where they're from.
With the way cellphone coverage in the USA works, to my understanding, the same applies to someone who's simply hopped over to the next state to visit friends or whatever.
Indeed that does seem to be what is being said:
Original AdvisoryThe whole thing is merely pointing out that some IP stacks don't give anything like enough randomness between connections (attempts) from different hosts.
The vendors of such stacks need to get a clue[tm].
There's not just hijacking the connect. If you know the necessary sequence number you could spoof a RST packet to make the connect drop, voila DoS.
Yes, drop prediction _for deciding on hits/kills_. You still use prediction to try and give each player a better view of things at that instant, but they know not to 100% rely on it being reality. It's better than seeing players freeze up anyway.
I may have a lovely 15ms ping to Barrysworld but I often play on upto 250ms ping and 10-20% packet loss for my Q3F clan, so I do experience it both ways.
Why should a modemer need slightly less _skill_ to kill me just because he gets greater compensation due to his higher ping?
I recently took up playing CounterStrike so checked lots of sites for tips on tweaking HL performance. Whilst doing this I came across some info about the netcode in 1.1.0.1. Apparently Valve did in fact make it easier for modems, at the expense of broadband connections. What they did was allow some of the client-side prediction from higher-ping players to over-ride what the server would have thought happened. Thus a broadband player could have gotten a kill denied that he SHOULD have gotten 'cos the prediction on the modemer's client said he got out of the way in time. Likewise a broadband user can get killed by a modemer due to this. IMO this plain sucks. Leave it with things decided purely at the server and if someone wants 'fairer' get a better connection. -Ath
To quote the article:
Attorney Walter Gordon routinely exchanges hefty word processing documents with other lawyers over the Internet, with each side making changes.
I wonder what juicy snippets are in those documents by the end, given how Word likes to retain bits of old revisions...