I just fired up ethereal and refreshed my RSS reader. Out of the dozen or so feeds I monitor, a few of them are using Etags and sensible cache-control headers, so I just get 304 Not Modified. Of the rest, not a single one is compressing even though my client is specifying gzip and deflate in its Accept-Encoding header.
HTTP compression will work even better here than it does for regular pages - RSS is basically all text so every response is going to be compressible. Looking at a handful of my feeds, some quick messing about with wget & gzip gives me an average compression ratio of 3:1. That's a 66% reduction in bandwidth utilization. If just half of your clients support HTTP compression, it's still a significant savings.
Now, the article is talking about poorly designed aggregators that don't check whether the content's changed (I'm assuming he's talking about Etags). There's not much you can do about that, but using compression for capable clients sure seems like it would be a good thing.
Your presumption is incorrect. The caller ID info is sent by your local switch to your phone between the first and second rings. This is before the circuit is completed so there is no chance for the calling party to "overwhelm" the telco's signal.
They're doing it by altering the number sent by the originator. PBXs have the capability to report the calling number for outbound calls (so that the correct extension shows up on Caller ID) and that may be what's used here.
What ILM is doing: Distributing render tasks to idle computers that reside in their offices, are connected at 100Mbps minimum, and are known quantities under their control.
What ILM is not doing: Distributing render tasks to idle computers that are connected at 1Mbps, running Windows ME on a P3 with 128MB RAM, and could reboot at any moment due to the cat walking on the keyboard.
I'm using Intel Pro/1000MT desktop NICs and the Netgear GS105 5-port copper switch. My desktop machine is an Athlon XP 2500+ running XP and my file server is a VIA C3 at 800MHz running Linux. Iperf gives me TCP throughput of about 650-700Mbps. I'm probably not getting quite that much on actual file transfers since there's a disk involved, but it is quite a bit faster than running at 100Mbps.
For those who claim gigabit isn't needed for home use... sorry, it is. I was tired of waiting for things to transfer at 100Mbps. There's my need right there. At $30 for a NIC and $85 for a switch, it's not like I'm throwing tons of money at it, and I'm seeing a significant improvement over 100BaseT.
Netgear (and a lot of the other consumer switch manufacturers) claim wire-speed performance on their gigabit switches, even the little 5-port versions. I don't doubt it since they're using Marvell or Broadcom switching silicon. I haven't had a chance to put the Smartbits on my Netgear to prove/disprove the claim though.
You don't need a special crossover cable (or any crossover cable, for that matter) for GigE. The 1000BaseTX standard requires that copper GigE ports do Auto MDIX, which means you can plug in a standard straight cable and the port will figure out if it needs to cross. I've used several machines with gig ports (IBM T40, G4 powerbook, Intel & Netgear NICs in desktops) and they have all supported this.
Not having to worry about crossover cables is reason alone to go GigE.
I don't know where you got 400Mbps - there's nothing in the standards that supports operation at 400Mbps.
I realize that this has nothing to do with administrator privileges, but Post-Its on the bottoms of all the optical mice in the lab worked well for me on April Fool's....
No, I'm not exaggerating. To be fair this only happened once in a while, but it was damned annoying when it did. I don't recall seeing the bug you mention. Excel was worse - I could often move around with the arrow keys and the system wouldn't keep up. Copy and paste operations in Excel were also very slow.
I don't know if it was relevant, but I usually had a ton of stuff running. Then again, it generally wasn't enough to exhaust the RAM and the machine wasn't swapping.
The other problem I had was clicking on the menu bar, waiting, thinking I had mis-clicked and clicking again just as the menu popped down. This had the effect of making the menu disappear again. Extremely annoying.
Whatever the gigahertz numbers say, Macintoshes are comparable in performance to Windows or Linux machines.
I'm tired of hearing this. Last year I decided to try the Mac, with a 866MHz G4 Powerbook. Even with 768MB of RAM, the thing was sooo sloooow. I got to watch Chimera render web pages. Mac Mail was like molasses in January. Word couldn't keep up with my typing, and moving between cells in Excel was an exercise in patience.
I now have an IBM T40 (1.3GHz P-M, 512MB RAM) and this thing just smokes the Powerbook in everyday usage.
There is nothing I cannot do with a Mac that I want to do, nor am I prevented from interacting with Windows boxes or Linux boxes. It just works. Transparently.
Yeah? How about browsing a routed Windows network with WINS? What about printing to a Sharp AR-810 printer/copier? These are things I found I could not do with a Mac.
Every security update from Microsoft means the Windows guys are running around updating. The Mac guys just sit there and keep working.
I dunno, I seem to recall getting an update about once a week on my Powerbook - and more often than not, having to reboot because of it.
They're talking about collisions within the 802.11 wireless network, not in the 802.3 wired network behind it. What you propose isn't feasible since you're essentially talking about dedicating a wireless channel to every user.
To transmit data (except SMS messages) it's necessary to open an end to end virtual circuit. So you can't trickly information back and forward to the phone all the time, at a very low bandwidth (and consequently very low cost). And there's no multicast, so software download to each phone has to be done one at a time.
Wait, I'm confused. Are you saying that this is an interesting feature? From your description it almost sounds like a disadvantage.
Bad numbers in the article.
on
802.11g Slows Down
·
· Score: 5, Informative
Everybody's going on and on about how it's hardly faster than.11b. Read the freakin article:
"Li estimated that that in mixed 802.11b and 802.11g networks running standard TCP/IP Internet protocols, this will reduce actual throughput to 10Mbit/sec. -- while pure 802.11g networks will have actual data rates of around 20Mbit/sec. Li pointed out that even at these data rates the 802.11g devices still outperform 802.11b devices, which have a raw data rate of 11Mbit/sec. but an actual throughput of about half that speed. "
See that? He's saying.11b is about 5Mbps true throughput..11g will be twice that in "safe mode" and four times that in pure.11g mode.
The article would have been much clearer if he had said ".11g is being reduced from 54Mbps raw data rate to X Mbps raw data rate, and from Y Mbps true throughput to 10 or 20Mbps true throughput." Instead he says it's getting reduced from 54Mbps raw data rate to 10 or 20Mbps true throughput. Way to mismatch your units to get the biggest reduction possible.
I've got an older Laserjet. I bought it because they run forever, and when they do break down they're easily serviced.
I got my start in the computer business working in a small computer shop. One of the things we did was repair Laserjets. This was in the days of the LJ II, III, and 4. Those old IIs and IIIs would come in broken with 100,000 pages on them, we'd replace a part (often the fuser), and they'd go back out again - total repair cost usually a hundred bucks or so. The HP service manuals for those things were the bomb!
When it came time to get something more stout than my Epson 750 inkjet for my home office, I picked up a LJ 4M on eBay for $200. It had something like 10,000 pages on it, and came with the JetDirect card and PostScript. About a month after I got it, it started throwing a service error. I ended up ordering a fuser bulb and replaced it in about half an hour. One year and about 4000 pages later, it's still going strong. The 3-year-old Epson inkjet is on its last legs after less than a thousand pages.
The only thing I wish I had done is gotten a 4+ because they have a low-power sleep mode. I have to remember to shut off my 4 or else it sucks up a bunch of power.
You can tell it to start and/or end on time, early, or late. I've used it for shows that ran a minute or two longer than expected. I don't know how far you can shift, though... I've only had occasion to use one or two minutes.
What if someone in another room wants to listen to something other than what's being broadcast from the server?
There was a comment similar to this one yesterday. I think you are missing one of the advantages of a setup like this - each receiver can play its own audio stream. If you don't want to do that, then an FM transmitter is fine. But if you want to listen to different things in different areas simultanously, then it doesn't help.
Headphones - Aiwa vs. Sennheiser vs. Bose
on
Controlling tha Noise?
·
· Score: 3, Informative
I've owned all three headphones, and currently use the Bose. Here are my impressions:
Aiwa (I don't remember the mode), ~$50:
- OK sound quality
- OK noise reduction
- Not-so-good comfort
Sennheiser HDC-451, ~$150:
- Good sound quality
- Good noise reduction
- Not-so-good comfort
The problem I had with the Aiwas and the Senns is that the earpieces sit against your ear lobes like regular open-air headphones, and after a while it becomes uncomfortable. They have thin earpads that don't offer much padding - I assume to keep noise from leaking through. The Bose set has earcups that surround your ears and sit against your head, and have a soft padding material on them. I've worn these on 5-hour flights and they don't bother me at all. The Bose also have the best sound quality, though the noise reduction doesn't seem to be any better than the Senns.
You might also want to look into Etymotic ER-4 or ER-6 headphones. These are little earbuds that seal out ambient noise (Etymotic advertises 20-25 dB passive noise reduction on the ER-4s). The ER-4s are spendy ($250-300) but the ER-6s are cheaper ($125-150). I tried a similar thing from Koss called "The Plug" but they sucked. They used a foam earplug-type material surrounding the driver on each side. The foam wasn't dense enough to seal well and it didn't hold its shape. Complete waste of time & money IMHO.
Slashdot Mirror
I just fired up ethereal and refreshed my RSS reader. Out of the dozen or so feeds I monitor, a few of them are using Etags and sensible cache-control headers, so I just get 304 Not Modified. Of the rest, not a single one is compressing even though my client is specifying gzip and deflate in its Accept-Encoding header.
HTTP compression will work even better here than it does for regular pages - RSS is basically all text so every response is going to be compressible. Looking at a handful of my feeds, some quick messing about with wget & gzip gives me an average compression ratio of 3:1. That's a 66% reduction in bandwidth utilization. If just half of your clients support HTTP compression, it's still a significant savings.
Now, the article is talking about poorly designed aggregators that don't check whether the content's changed (I'm assuming he's talking about Etags). There's not much you can do about that, but using compression for capable clients sure seems like it would be a good thing.
Your presumption is incorrect. The caller ID info is sent by your local switch to your phone between the first and second rings. This is before the circuit is completed so there is no chance for the calling party to "overwhelm" the telco's signal.
They're doing it by altering the number sent by the originator. PBXs have the capability to report the calling number for outbound calls (so that the correct extension shows up on Caller ID) and that may be what's used here.
What ILM is doing: Distributing render tasks to idle computers that reside in their offices, are connected at 100Mbps minimum, and are known quantities under their control.
What ILM is not doing: Distributing render tasks to idle computers that are connected at 1Mbps, running Windows ME on a P3 with 128MB RAM, and could reboot at any moment due to the cat walking on the keyboard.
Try connecting San Jose, California, to Reston, Virginia, with 100Mbps Ethernet and let us know how that goes.
I'm using Intel Pro/1000MT desktop NICs and the Netgear GS105 5-port copper switch. My desktop machine is an Athlon XP 2500+ running XP and my file server is a VIA C3 at 800MHz running Linux. Iperf gives me TCP throughput of about 650-700Mbps. I'm probably not getting quite that much on actual file transfers since there's a disk involved, but it is quite a bit faster than running at 100Mbps.
For those who claim gigabit isn't needed for home use... sorry, it is. I was tired of waiting for things to transfer at 100Mbps. There's my need right there. At $30 for a NIC and $85 for a switch, it's not like I'm throwing tons of money at it, and I'm seeing a significant improvement over 100BaseT.
Netgear (and a lot of the other consumer switch manufacturers) claim wire-speed performance on their gigabit switches, even the little 5-port versions. I don't doubt it since they're using Marvell or Broadcom switching silicon. I haven't had a chance to put the Smartbits on my Netgear to prove/disprove the claim though.
You don't need a special crossover cable (or any crossover cable, for that matter) for GigE. The 1000BaseTX standard requires that copper GigE ports do Auto MDIX, which means you can plug in a standard straight cable and the port will figure out if it needs to cross. I've used several machines with gig ports (IBM T40, G4 powerbook, Intel & Netgear NICs in desktops) and they have all supported this.
Not having to worry about crossover cables is reason alone to go GigE.
I don't know where you got 400Mbps - there's nothing in the standards that supports operation at 400Mbps.
I realize that this has nothing to do with administrator privileges, but Post-Its on the bottoms of all the optical mice in the lab worked well for me on April Fool's....
No, I'm not exaggerating. To be fair this only happened once in a while, but it was damned annoying when it did. I don't recall seeing the bug you mention. Excel was worse - I could often move around with the arrow keys and the system wouldn't keep up. Copy and paste operations in Excel were also very slow.
I don't know if it was relevant, but I usually had a ton of stuff running. Then again, it generally wasn't enough to exhaust the RAM and the machine wasn't swapping.
The other problem I had was clicking on the menu bar, waiting, thinking I had mis-clicked and clicking again just as the menu popped down. This had the effect of making the menu disappear again. Extremely annoying.
Yes, I'm aware of DAVE, but I'd rather not have to buy extra software in order to make the computer Just Work.
I'm tired of hearing this. Last year I decided to try the Mac, with a 866MHz G4 Powerbook. Even with 768MB of RAM, the thing was sooo sloooow. I got to watch Chimera render web pages. Mac Mail was like molasses in January. Word couldn't keep up with my typing, and moving between cells in Excel was an exercise in patience.
I now have an IBM T40 (1.3GHz P-M, 512MB RAM) and this thing just smokes the Powerbook in everyday usage.
Yeah? How about browsing a routed Windows network with WINS? What about printing to a Sharp AR-810 printer/copier? These are things I found I could not do with a Mac.
I dunno, I seem to recall getting an update about once a week on my Powerbook - and more often than not, having to reboot because of it.
They're talking about collisions within the 802.11 wireless network, not in the 802.3 wired network behind it. What you propose isn't feasible since you're essentially talking about dedicating a wireless channel to every user.
Wait, I'm confused. Are you saying that this is an interesting feature? From your description it almost sounds like a disadvantage.
What about KMFDM?
Everybody's going on and on about how it's hardly faster than .11b. Read the freakin article:
.11b is about 5Mbps true throughput. .11g will be twice that in "safe mode" and four times that in pure .11g mode.
"Li estimated that that in mixed 802.11b and 802.11g networks running standard TCP/IP Internet protocols, this will reduce actual throughput to 10Mbit/sec. -- while pure 802.11g networks will have actual data rates of around 20Mbit/sec. Li pointed out that even at these data rates the 802.11g devices still outperform 802.11b devices, which have a raw data rate of 11Mbit/sec. but an actual throughput of about half that speed. "
See that? He's saying
The article would have been much clearer if he had said ".11g is being reduced from 54Mbps raw data rate to X Mbps raw data rate, and from Y Mbps true throughput to 10 or 20Mbps true throughput." Instead he says it's getting reduced from 54Mbps raw data rate to 10 or 20Mbps true throughput. Way to mismatch your units to get the biggest reduction possible.
I've got an older Laserjet. I bought it because they run forever, and when they do break down they're easily serviced.
I got my start in the computer business working in a small computer shop. One of the things we did was repair Laserjets. This was in the days of the LJ II, III, and 4. Those old IIs and IIIs would come in broken with 100,000 pages on them, we'd replace a part (often the fuser), and they'd go back out again - total repair cost usually a hundred bucks or so. The HP service manuals for those things were the bomb!
When it came time to get something more stout than my Epson 750 inkjet for my home office, I picked up a LJ 4M on eBay for $200. It had something like 10,000 pages on it, and came with the JetDirect card and PostScript. About a month after I got it, it started throwing a service error. I ended up ordering a fuser bulb and replaced it in about half an hour. One year and about 4000 pages later, it's still going strong. The 3-year-old Epson inkjet is on its last legs after less than a thousand pages.
The only thing I wish I had done is gotten a 4+ because they have a low-power sleep mode. I have to remember to shut off my 4 or else it sucks up a bunch of power.
You can tell it to start and/or end on time, early, or late. I've used it for shows that ran a minute or two longer than expected. I don't know how far you can shift, though... I've only had occasion to use one or two minutes.
Putting this on your beer keg would be like putting a huge heat sink on it. You'd pull heat into the beer.
And I don't know where you got the polarity idea... it's not a Peltier.
What if someone in another room wants to listen to something other than what's being broadcast from the server?
There was a comment similar to this one yesterday. I think you are missing one of the advantages of a setup like this - each receiver can play its own audio stream. If you don't want to do that, then an FM transmitter is fine. But if you want to listen to different things in different areas simultanously, then it doesn't help.
I've owned all three headphones, and currently use the Bose. Here are my impressions:
Aiwa (I don't remember the mode), ~$50:
- OK sound quality
- OK noise reduction
- Not-so-good comfort
Sennheiser HDC-451, ~$150:
- Good sound quality
- Good noise reduction
- Not-so-good comfort
Bose QuietComfort, $300:
- Excellent sound quality
- Good noise reduction
- Excellent comfort
The problem I had with the Aiwas and the Senns is that the earpieces sit against your ear lobes like regular open-air headphones, and after a while it becomes uncomfortable. They have thin earpads that don't offer much padding - I assume to keep noise from leaking through. The Bose set has earcups that surround your ears and sit against your head, and have a soft padding material on them. I've worn these on 5-hour flights and they don't bother me at all. The Bose also have the best sound quality, though the noise reduction doesn't seem to be any better than the Senns.
You might also want to look into Etymotic ER-4 or ER-6 headphones. These are little earbuds that seal out ambient noise (Etymotic advertises 20-25 dB passive noise reduction on the ER-4s). The ER-4s are spendy ($250-300) but the ER-6s are cheaper ($125-150). I tried a similar thing from Koss called "The Plug" but they sucked. They used a foam earplug-type material surrounding the driver on each side. The foam wasn't dense enough to seal well and it didn't hold its shape. Complete waste of time & money IMHO.