But 50ms maximum latency to where? And is it one way or return?
Getting latency down is very hard especially when modems are involved. Often modems need to keep a moving "sample window" of the signal before they can decide what bits were sent. That "window" = latency.
FWIW the distance between the east and west coast of the USA is about 13 light-milliseconds (following the surface of the earth) - assuming speed of light in vacuum.
But light travels slower in optical fibres. A naive calculation just using the index of refraction gives me about 20 milliseconds. Round trip time then becomes 40ms.
The fibre isn't taking a "great circle route" and there's some modulation and demodulation involved, so round trip time is likely to be higher than 50ms.
It's strange that your neighbours and in-laws call you up so frequently for tech help. How often do they really call you?
My friends and relatives don't bother me with IT problems every day. I think they're not that unfortunate to experience IT problems that often.
Heck my dad only checks his email once a week (or even less often), so even if he has email problems, he'd only bother me once a week at most;).
Anyway, I think my friends and relatives would help me too if I had problems they could help with. But I don't ask for free help every day, nor do they.
As the golden rule goes: Do unto others as you would have them do unto you. And apply the golden rule to others as you would have others apply the golden rule to you - e.g. not a stupid literal way;).
But I want a fast but smaller and cheaper SSD in the hard _disk_ drive itself.
Let the drive figure out and manage the permanent storage thing. I don't want to have to figure it out.
With stuff like NCQ the drive can probably have a good idea of what to cache and what not to cache.
If the drive is smart, it will not waste too much of the SSD cache on sequential writes and reads.
e.g. don't cache "read blocks: 1,2,3,4,5,6,7,8" on a high priority basis (leave them in a smaller cache for sequential reads).
If you get "read blocks:1,2,3,100,5,6,7,200,8" in one transaction, cache blocks 100 and 200 on a higher priority.
If you get sequential writes, don't cache. If you get random writes, cache them then slowly write them to the platters (when a block is written to the platters- you can readjust its priority in the cache).
Speaking of cache, for a desktop what might be nice is a smart 500GB/1TB HD that uses a high performance 1GB/4GB SSD as read+write cache (smaller and thus cheaper than a 32GB SSD).
Done intelligently it'll perform almost as well as a fast SSD for many real world tasks. Random burst writes can be fast then the drive slowly writes the stuff to the platters - it only slows down when the cache becomes full. Random reads can be fast too if they are already cached - which could be true since the cache persists even after you shut the computer down.
Shouldn't be too difficult to have special heuristics to speed boots for desktop drives (may not be so important for server drives[1]) - since the drive will know when power has just come back on.
The drive could also cache whatever that makes it seek a lot. Disks are quite fast for sequential reads/writes, it's the random access that hurts.
Wonder if it'll ever happen though - might be too expensive to build? But with this approach seagate and WD might be able to squeeze a lot more from their drives.
When you have a drive like that, there are a lot of nifty tricks you can do to speed up real world stuff.
Of course, there are also a lot of things that can go wrong - make sure your smart HDD code is bug free enough to not corrupt or lose data;).
[1] I guess one could allow the drive to be configured to prioritize boots (so it'll try to cache most of what is read during boot in the SSD), or not - e.g. boot up speeds might not be so important for systems where the computer is hardly ever rebooted (the drive can detect that as well, but some people might still want fast boots for such environments).
He has lost enough computer equipment from lightning that now he unplugs everything when he hears thunder when there's a storm approaching (which is frequent here).
That's assuming you have enough of the population educated to a level where they can count accurately (and are willing to do so).
"Counting is hard, let's go shopping!";).
Don't get me wrong, I'm all for paper ballots - after all they are simple, and you can even do the counting in front of the various observers and party representatives. That makes it a lot harder to cheat[1]. Then more people will understand what is going on and how hard it is to cheat.
As I've said before, elections don't just have to be fair, they have to be _seen_ as fair.
Otherwise even if it actually is "fair enough" you could get rioting in the streets.
That's why electronic voting is a bad idea.
[1] The easiest way to cheat then is via postal votes - but that affects the other systems as well.
If you're talking about companies going from SPARC to x86, Linux is far more responsible for that than Microsoft.
Linux+x86 is Solaris+SPARC's main competitor. Not Microsoft Windows+x86.
The stuff people would want to run on SPARC machines, can usually be run on Linux+x86 with decent performance (and often better price/performance).
And if they really wanted they could also do Solaris+x86. So Sun's also responsible for that...
If people like vmware manage to provide _seamless_ high availability features that are less buggy than good hardware, the x86 stuff will crush the high end HA server market too.
Would blood tests be conclusive though? I'm no doctor, just wondering whether some infections might not be detected by blood tests.
Anyway many body builders seem to be able to grow stuff. Their joints and bones definitely do grow[1]. Takes a while, but they do.
Glucosamine combined with chondroitin sulfate works for some people in making cartilage grow faster (or wear out slower).
[1] Which is why I think it might be better to treat some "carpal tunnel syndrome" cases by just keeping the affected nerves alive and well (with stuff like methylcobalamin) till the body grows to accommodate stuff properly.
I tend to skip those sites. Maybe I'm a small minority.
But usually the content (or better) is available elsewhere.
For example there were some PC review sites which require about 20-30 "next page" clicks per article, I gave up on them. Nowadays I just stick to Anandtech and a few other saner ones that at least allow you to put everything on one page.
I prefer it all on one page (even if it that single page includes many ads). I know how to use the scroll bar and the mouse scroll wheel/button.
> but which has a bigger screen, more memory, more storage or a faster processor
I thought that was typically called a laptop;).
The netbook spec just allows Microsoft to sell a cheaper Windows O/S for netbooks without affecting their pricing for the laptop market. I don't see how that netbook spec would keep Intel out of a premium netbook market. Linux runs fine on netbook/laptops with Intel CPUs. OSX runs fine too.
Even your link itself and this article show that Intel is going into more markets, whether Microsoft joins them or not.
Unless there's some clause which allows Via or AMD to have cheaper windows O/S than Intel for the same class netbook/laptop, I don't see how it hurts Intel at all.
Maybe they might get in trouble with some regulators or get sued, but how is this a dumb move by Intel?
Thanks. Wasn't aware that Symbian did that (or I forgot).
Yes it would be something like that, but more oriented to a desktop/server environment e.g. https://bugs.launchpad.net/ubuntu/+bug/156693
I was thinking more of there being a hopefully manageable number of "predefined sandbox templates" that an application could request to be run within e.g. "guest applet", "screensaver".
Specifying one of those predefined templates, would imply a list of access items (network, user data etc). These do not necessarily have to be signed.
Custom sandboxes must be possible as well, but they should preferably be signed (and you could set up a computer to only accept those signed by certain Vendors).
In fact it could be possible to have the application signed by one party (supplier), and have custom sandbox template for it that's signed by another party (IT security auditor).
After all you might trust your Supplier to provide you an application with the features you want, but you're not sure about the security, so you get some 3rd party to create/modify/check the sandbox template for that application.
But perhaps that adds too much complexity - as you said nothing is unbreakable, so it might be "diminishing returns".
There definitely has to be a way to revoke stuff. After all I hear there's symbian signed malware floating about.
Dunno. It happened to me too. I'm using firefox + noscript.
Seems random though. Probably a bug in Slashdot.
I dunno why Slashdot is doing all these changes. They're making it slower and less usable IMO. If I turn on javascript, the slashdot mainpage takes ages to load up.
Slashdot Mirror
Those were rather crap though...
Expensive and slow.
But 50ms maximum latency to where? And is it one way or return?
Getting latency down is very hard especially when modems are involved. Often modems need to keep a moving "sample window" of the signal before they can decide what bits were sent. That "window" = latency.
FWIW the distance between the east and west coast of the USA is about 13 light-milliseconds (following the surface of the earth) - assuming speed of light in vacuum.
But light travels slower in optical fibres. A naive calculation just using the index of refraction gives me about 20 milliseconds. Round trip time then becomes 40ms.
The fibre isn't taking a "great circle route" and there's some modulation and demodulation involved, so round trip time is likely to be higher than 50ms.
It's strange that your neighbours and in-laws call you up so frequently for tech help. How often do they really call you?
;).
;).
My friends and relatives don't bother me with IT problems every day. I think they're not that unfortunate to experience IT problems that often.
Heck my dad only checks his email once a week (or even less often), so even if he has email problems, he'd only bother me once a week at most
Anyway, I think my friends and relatives would help me too if I had problems they could help with. But I don't ask for free help every day, nor do they.
As the golden rule goes: Do unto others as you would have them do unto you. And apply the golden rule to others as you would have others apply the golden rule to you - e.g. not a stupid literal way
Is this really an idea from Apple? I skimmed through the article and saw no mention of "Opera as webserver" being Apple's idea.
Citation please?
But I want a fast but smaller and cheaper SSD in the hard _disk_ drive itself.
Let the drive figure out and manage the permanent storage thing. I don't want to have to figure it out.
With stuff like NCQ the drive can probably have a good idea of what to cache and what not to cache.
If the drive is smart, it will not waste too much of the SSD cache on sequential writes and reads.
e.g. don't cache "read blocks: 1,2,3,4,5,6,7,8" on a high priority basis (leave them in a smaller cache for sequential reads).
If you get "read blocks:1,2,3,100,5,6,7,200,8" in one transaction, cache blocks 100 and 200 on a higher priority.
If you get sequential writes, don't cache.
If you get random writes, cache them then slowly write them to the platters (when a block is written to the platters- you can readjust its priority in the cache).
Speaking of cache, for a desktop what might be nice is a smart 500GB/1TB HD that uses a high performance 1GB/4GB SSD as read+write cache (smaller and thus cheaper than a 32GB SSD).
;).
Done intelligently it'll perform almost as well as a fast SSD for many real world tasks. Random burst writes can be fast then the drive slowly writes the stuff to the platters - it only slows down when the cache becomes full. Random reads can be fast too if they are already cached - which could be true since the cache persists even after you shut the computer down.
Shouldn't be too difficult to have special heuristics to speed boots for desktop drives (may not be so important for server drives[1]) - since the drive will know when power has just come back on.
The drive could also cache whatever that makes it seek a lot. Disks are quite fast for sequential reads/writes, it's the random access that hurts.
Wonder if it'll ever happen though - might be too expensive to build? But with this approach seagate and WD might be able to squeeze a lot more from their drives.
When you have a drive like that, there are a lot of nifty tricks you can do to speed up real world stuff.
Of course, there are also a lot of things that can go wrong - make sure your smart HDD code is bug free enough to not corrupt or lose data
[1] I guess one could allow the drive to be configured to prioritize boots (so it'll try to cache most of what is read during boot in the SSD), or not - e.g. boot up speeds might not be so important for systems where the computer is hardly ever rebooted (the drive can detect that as well, but some people might still want fast boots for such environments).
So isn't what the company did fraud?
They said something was homeopathic (but it isn't) and tried to profit from it.
And they have caused rather permanent damage.
Go tell my friend that.
He has lost enough computer equipment from lightning that now he unplugs everything when he hears thunder when there's a storm approaching (which is frequent here).
That's assuming you have enough of the population educated to a level where they can count accurately (and are willing to do so).
;).
"Counting is hard, let's go shopping!"
Don't get me wrong, I'm all for paper ballots - after all they are simple, and you can even do the counting in front of the various observers and party representatives. That makes it a lot harder to cheat[1]. Then more people will understand what is going on and how hard it is to cheat.
As I've said before, elections don't just have to be fair, they have to be _seen_ as fair.
Otherwise even if it actually is "fair enough" you could get rioting in the streets.
That's why electronic voting is a bad idea.
[1] The easiest way to cheat then is via postal votes - but that affects the other systems as well.
But why write an April Fool's joke in February?
http://www.nytimes.com/2004/02/02/opinion/02SAFI.html
If you're talking about companies going from SPARC to x86, Linux is far more responsible for that than Microsoft.
Linux+x86 is Solaris+SPARC's main competitor. Not Microsoft Windows+x86.
The stuff people would want to run on SPARC machines, can usually be run on Linux+x86 with decent performance (and often better price/performance).
And if they really wanted they could also do Solaris+x86. So Sun's also responsible for that...
If people like vmware manage to provide _seamless_ high availability features that are less buggy than good hardware, the x86 stuff will crush the high end HA server market too.
How sure are we that a Culture that takes the risk and cost to heal the sick, would do worse than Cultures that don't?
A Culture might even outlive its original host species.
That would be a rather evolutionarily fit Culture right?
> Natural selection will always favor someone that spreads their genes
:).
Not all ants and bees in a colony can reproduce. So it seems as long as you help spread similar genes it works.
How similar? Who knows
Because humans have reached a stage where they can support Cultures.
The various cultures/memes/religions compete for survival.
Culture does also help a lot in the survival of the hosts.
Would blood tests be conclusive though? I'm no doctor, just wondering whether some infections might not be detected by blood tests.
Anyway many body builders seem to be able to grow stuff. Their joints and bones definitely do grow[1]. Takes a while, but they do.
Glucosamine combined with chondroitin sulfate works for some people in making cartilage grow faster (or wear out slower).
[1] Which is why I think it might be better to treat some "carpal tunnel syndrome" cases by just keeping the affected nerves alive and well (with stuff like methylcobalamin) till the body grows to accommodate stuff properly.
I am not a doctor.
But I'm curious - do you have bad gum disease?
http://www.webmd.com/rheumatoid-arthritis/news/20090612/fixing-gums-rheumatoid-arthritis
Maybe your immune system is fighting a stalemate battle with something and you are getting arthritis as collateral damage.
If it's not the gums, see if there's some other persistent infection, and try to get that dealt with.
I knew there was something about that post. Couldn't put my finger on it though.
Thanks for the tip!
Seriously though, wouldn't disabled be a far more negative term for geeks than handicapped?
After all in geek terms, if something is disabled it no longer functions.
Whereas if someone is handicapped he/she still can do stuff, just needs a bit "extra" to equalize things.
I found it rather peculiar when people switched from using the term "handicapped" to "disabled".
I tend to skip those sites. Maybe I'm a small minority.
But usually the content (or better) is available elsewhere.
For example there were some PC review sites which require about 20-30 "next page" clicks per article, I gave up on them. Nowadays I just stick to Anandtech and a few other saner ones that at least allow you to put everything on one page.
I prefer it all on one page (even if it that single page includes many ads). I know how to use the scroll bar and the mouse scroll wheel/button.
> but which has a bigger screen, more memory, more storage or a faster processor
;).
I thought that was typically called a laptop
The netbook spec just allows Microsoft to sell a cheaper Windows O/S for netbooks without affecting their pricing for the laptop market. I don't see how that netbook spec would keep Intel out of a premium netbook market. Linux runs fine on netbook/laptops with Intel CPUs. OSX runs fine too.
Even your link itself and this article show that Intel is going into more markets, whether Microsoft joins them or not.
Unless there's some clause which allows Via or AMD to have cheaper windows O/S than Intel for the same class netbook/laptop, I don't see how it hurts Intel at all.
Maybe they might get in trouble with some regulators or get sued, but how is this a dumb move by Intel?
Thanks. Wasn't aware that Symbian did that (or I forgot).
Yes it would be something like that, but more oriented to a desktop/server environment e.g. https://bugs.launchpad.net/ubuntu/+bug/156693
I was thinking more of there being a hopefully manageable number of "predefined sandbox templates" that an application could request to be run within e.g. "guest applet", "screensaver".
Specifying one of those predefined templates, would imply a list of access items (network, user data etc). These do not necessarily have to be signed.
Custom sandboxes must be possible as well, but they should preferably be signed (and you could set up a computer to only accept those signed by certain Vendors).
In fact it could be possible to have the application signed by one party (supplier), and have custom sandbox template for it that's signed by another party (IT security auditor).
After all you might trust your Supplier to provide you an application with the features you want, but you're not sure about the security, so you get some 3rd party to create/modify/check the sandbox template for that application.
But perhaps that adds too much complexity - as you said nothing is unbreakable, so it might be "diminishing returns".
There definitely has to be a way to revoke stuff. After all I hear there's symbian signed malware floating about.
Well I think it's more of a bug in their caching or load balancing stuff.
:) ) if I disable scripts.
Because it normally works fine (better
And for me that weird artifact only showed up for one article on it.slashdot.org. The other articles were fine.
Anyway, seems like someone else has scripts enabled and still experiences that issue.
What next? You want mug shots?
p.s. I rarely drink coffee- though it tastes good, it disagrees with my stomach.
> The problem is commercial usage.
But how's that creepy?
Trying to use the Wheels of Justice to grind Nescafe...
I wonder what will be the grounds for their decision.
Anyone going to spill the beans?
Dunno. It happened to me too. I'm using firefox + noscript.
Seems random though. Probably a bug in Slashdot.
I dunno why Slashdot is doing all these changes. They're making it slower and less usable IMO. If I turn on javascript, the slashdot mainpage takes ages to load up.