I've been commuting on a singlespeed road bike (Specialized Langster) here in London virtually every day for the last 9 months.
Ongoing costs so far have been a set of brake pads and some chain lube. The singlespeed drivetrain is almost bulletproof - just wipe and re-lube every so often.
The bike "paid for itself" (i.e. offset what I would have paid on the Tube) in less than 6 months, and I've lost 10kg of blubber at the same time. Now I'm "in profit" in both financial and health terms. I've never felt better, and arrive at work feeling sharp and don't need (as much) coffee.
I understand that bicycling isn't/can't be for everyone, especially not on a singlespeed (my London commute is astonishingly flat - Keira Knightley-esque!) but it's made a huge difference to my life. I wish more people would try it!
I've used GPRS and 3G mobile broadband in both Australia and the UK. The problems they have to solve are very different, which is why you get a very different experience.
Australia is very large, and has low population density. The UK is small, with very high population density. This works against the "shared bandwidth" nature of wireless comms. I've even noticed it over the last year - my 3.5G connection in Central London was noticeably faster in January 2008 than it was in December 08, just because more people got on the mobile broadband bandwagon.
The providers in the UK are just desperately trying to maintain the "it's really really fast!!!111" illusion for as long as they can, by doing image compression, proxying et al, all the while knowing that every new customer is actually slowing the whole thing down for everybody else.
Professional studio equipment will usually offer A-D/D-A conversion or interfacing at:
16-bit, 44.1kHz aka CD-quality
16-bit, 48kHz (original DAT maximum)
24-bit versions at the above rates
24-bit versions at twice these rates e.g. 88.2/96kHz
24-bit, 192kHz
In a studio situation, you absolutely want to record at "quite a bit higher than even an uncompressed CD". Particularly when multiple streams of such audio are mixed in the digital domain, you want the noise floor as low as possible. The final master will be 24-bit/192k and only at the last possible minute will it be rendered down to 16/44.1k for the CD press.
The hundreds of people I've just seen inside the Apple Store on Regent Street in London would beg to differ.
If you look around a typical coffee shop in Central London, you will see just as many Apple machines as PCs, if not more. This is probably a function of necessity (lots of visual/graphic design people), appreciation of the Macbook hardware design, general affluence and a very effective localised "I'm a Mac" ad campaign. I would estimate at least 95% of the laptop PCs you see inside a Starbucks are running Windows.
I don't think you can linearly-extrapolate your German experience - In *my* experience, Germany is very pro-open-source. But I'm not going to make a blanket statement about the rest of Europe;-)
This box definitely needs some firmware/FPGA tuning if it's not able to saturate its SATA connections.
Add to that the fact that the two-channel "RAID" configuration is *nowhere near* twice as quick as the single bus - this thing urgently needs some performance tweaks to fully realise the potential of the concept.
Pretty sure you've got your definition of "factoid" messed up there champ. Like most people, you have assumed it means "little fact" or perhaps "little-known fact". Possibly due to abuse by CNN using the word in this sense.
A factoid is a spurious - unverified, incorrect, or fabricated - statement formed and asserted as a fact, but with no veracity. The word appears in the Oxford English Dictionary as "something which becomes accepted as fact, although it may not be true.
It muddies the intention of the sentence when you use this word, because its meaning has been overloaded like this. I would have gone with:
"which contains an argument I have not seen mentioned before"
I was on the Queensland - NSW border over the weekend and it most certainly is _not_ fine there!
Every single appointment time etc has to be qualified with "Queensland Time" or "New South Wales Time" after it. Stores are having to stay open an hour longer per day to handle people expecting them to be open. VCR programming becomes even more challenging for some people (is my program on a NSW-based channel or Qld channel?) Most people on the border are using two sets of clocks.
Apparently there are still lots of not-too-bright people who say no to daylight saving "because it fades the curtains"... The sooner Qld joins civilisation the better...
IAAC as well, and it seems to me that the reason I'm at my current site is due to their bizarre HR policies for permanent staff, leading to:
Staff who can bullshit their way through an interview, but that's all
Lack of suitably-vicious firing policy for above
Uneven hiring of different skillsets resulting in large portions of staff being totally idle
Thus, low motivation to actually learn new things and get things done
Promotion of staff to managerial positions based on time served rather than merit
Zero repercussions when projects fail/slip horribly (e.g. an 18-month slip on a "6-month" project?!)
Then when the company decides they need a shiny new system there's no-one interested/able/willing to do it. Cue consulting company who brings in a team of bright-eyed and clued-up consultants. Sure we cost a lot but we'll get it done, done right, and on time, and afterwards we are much easier to get rid of than the permanent dross.
Look at it from the company above's point-of-view - where they know they have (at least some of) the above issues with their permanent staff.
I used to work at a company that made the *test equipment* for big-iron routers. In other words, generated the packets at rates that made them scream for mercy. We used ASICs to serialize/deserialize onto big fat data buses and FPGAs to do stuff like CAMs, massively-parallel matching and byte sweeping. We used the biggest, baddest FPGAs Xilinx et al had to offer and had them bulging at the seams.
You won't get even close to saturating/dealing with a "serious" link (i.e. OC-12 or more) without being able to distribute the effort around the place - i.e. L2 framing and L3 inspection on a line card and massive backplane capacity. THEN you can use your general-purpose CPU to make routing decisions etc. If your Xeon or whatever has to get interrupted on every packet, or you have packets trying to make their way at line rates across a general-purpose bus like PCI etc, you're gonna hit a problem shortly after adding your second line card to your 4U Dell.
By the way, a few years ago we used our gear to do some tests on PC-based 155mbps ATM cards. On a "state of the art" Pentium III we got packet loss rates of 5-10%. That's with ONE card.
While I agree completely with all of your points (make Friday spectacular again!), I suspect what the FIA are trying to do is simplify some of their own procedures and reduce some of their costs. Currently all code on board an F1 car goes through pre- and post-race scrutineering just like every other part. Looking over the (compiled!) code for illegalities is a true nightmare.
Now imagine everyone is running ECU-OS 1.0 (ignoring all MS jokes for the time being)... The "OS" is exactly the same (i.e. it checksums to the same value) for all, only the various configuration parameters (held in RAM?) vary. Now the scrutineering effort becomes: hook up to ECU ROM, download code (or do a boundary scan), perform checksum. If the checksum isn't right - bang - you're disqualified.
To a lesser extent it will also save the teams some money. Rich teams might currently have 200 engine parameters that they can tweak. Poorer teams might have only been able to afford to develop 50. If everyone gets 100 parameters, it comes down to engineering quality rather than quantity to work out what works best.
I hope she has some other form of (non-government-provided) income! A couple hundred a month is nowhere near enough to provide for five children (Assumption 1: She lives in the US or an equivalent country, not a third-world country where a few hundred a month is insanely rich. Assumption 2: When you say "a few hundred", I assume that means "a few hundred dollars". Since you probably really mean "a few hundred pounds", given your usage of the word "mum", adjust the following numbers appropriately for the currency.)
Actually, given the poster's use of the term op-shop plus mum I suspect they are Australian, like me. So adjusting for the relatively low cost-of-living here, plus the Australian manner of speaking, the poster probably meant:
"Considering she is at home looking after 5 kids and hence doesn't have a lot of time on her hands, she's doing a lot better than she would if she did nothing at all with that time. A few hundred really helps. "
What is surprising is that Ballmer and his boys actually tried...knowing what they know, they must have known they were on a fool's errand...
IF this is actually true and Ballmer really did bring the box in for his "top engineers" to fix (and not just some random support guys), can you not envision the battle for L337-ness that would ensue?:
Ballmer OK boys, I've tried to fix this POS and it's not responding to my advanced techniques. It's all yours MS Uber-nerds Can we remove this chair from the monitor? B OK... MSU Can we remove the chair wedged in the floppy slot? B I suppose, although I thought that might help Fucking Kill the spyware??? No??? MSU We'll work around it. But we're really busy you know, there's Vista to de-feature and we have to copy-and-paste a whole lot of code from Word into Excel, then Excel into Access and finally Access into IIS so Frontpage-published websites look good in IE. That's gonna take some time! B I'll give $20M to whoever gets it going again. MS Uber-nerd 1 I'd like to try first, sir. MS Uber-nerd 2 I've been a loyal Microsoftie for longer sir, I'd like to try first etc,etc,etc.
The introduction of Ubuntu 6.06 LTS (Long Term Support) would seem like a perfect time for a printed manual to be released. The two will go hand-in-hand to help PHBs accept this as a real product with real support.
The issue of the manuals falling out-of-date is probably less relevant in a highly-controlled corporate environment - most large corps have Windows Update turned off on their desktops, I would imagine they would do the same with Ubuntu. Thus the manuals will stay relevant far longer.
On a non-corporate note, this is a big step towards grandma-compatibility too -
- That generation prefers information in dead-tree format
- Avoids problems with help windows obscuring program windows
- Grandma is more likely to feel comfortable poking around a book
You cannot directly connect from the internet to a specific mobile phhone's IP address, regardless of the existence of a mobile web server there.
You can iff the phone initiated the communication. Here in Australia, all 3 major mobile networks use time-based firewall rules to allow a "push" *BACK* to a GPRS-enabled mobile device. This allows the device to get bidirectional communications with a server. Depending on the network you get between 15 seconds and 2 minutes to form up your reply packet and fire it back. After that, the device is "no longer there". You've also got to make sure that your reply packet is coming from the exact same IP:port combination that the phone sent to, or the firewall won't let it through. I imagine the networks in other countries do very similar things.
So if your server got hacked your mobile devices could get pingflooded - at AUS$10 (US$7.50) per megabyte received. Ouch.
Agree completely. The reason why someone would log into your phone's "website" would be because it has some "cool" data/real-time values on it that they want to see. So why not get the phone to send nice, snappy, minimal-overhead UDP packets to the back end of some *proper* HTTP server that can then wrap the values in pretty HTML/XML/CSS/what-have-you. Let it take any slashdotting for you, rather than your phone burning a hole in your pocket...
Slashdot Mirror
"Failed" implies the scheme has been abandoned. That is most certainly not the case - it's still operating and has been expanding.
You might have been hearing the propaganda being spread around by JCDecaux (the commercial arm) that was being used as a bargaining tool.
I've been commuting on a singlespeed road bike (Specialized Langster) here in London virtually every day for the last 9 months.
Ongoing costs so far have been a set of brake pads and some chain lube. The singlespeed drivetrain is almost bulletproof - just wipe and re-lube every so often.
The bike "paid for itself" (i.e. offset what I would have paid on the Tube) in less than 6 months, and I've lost 10kg of blubber at the same time. Now I'm "in profit" in both financial and health terms. I've never felt better, and arrive at work feeling sharp and don't need (as much) coffee.
I understand that bicycling isn't/can't be for everyone, especially not on a singlespeed (my London commute is astonishingly flat - Keira Knightley-esque!) but it's made a huge difference to my life. I wish more people would try it!
I've used GPRS and 3G mobile broadband in both Australia and the UK. The problems they have to solve are very different, which is why you get a very different experience.
Australia is very large, and has low population density. The UK is small, with very high population density. This works against the "shared bandwidth" nature of wireless comms. I've even noticed it over the last year - my 3.5G connection in Central London was noticeably faster in January 2008 than it was in December 08, just because more people got on the mobile broadband bandwagon.
The providers in the UK are just desperately trying to maintain the "it's really really fast!!!111" illusion for as long as they can, by doing image compression, proxying et al, all the while knowing that every new customer is actually slowing the whole thing down for everybody else.
In a studio situation, you absolutely want to record at "quite a bit higher than even an uncompressed CD". Particularly when multiple streams of such audio are mixed in the digital domain, you want the noise floor as low as possible. The final master will be 24-bit/192k and only at the last possible minute will it be rendered down to 16/44.1k for the CD press.
Outside of the US, Apple doesn't sell.
The hundreds of people I've just seen inside the Apple Store on Regent Street in London would beg to differ.
If you look around a typical coffee shop in Central London, you will see just as many Apple machines as PCs, if not more. This is probably a function of necessity (lots of visual/graphic design people), appreciation of the Macbook hardware design, general affluence and a very effective localised "I'm a Mac" ad campaign. I would estimate at least 95% of the laptop PCs you see inside a Starbucks are running Windows.
I don't think you can linearly-extrapolate your German experience - In *my* experience, Germany is very pro-open-source. But I'm not going to make a blanket statement about the rest of Europe ;-)
This box definitely needs some firmware/FPGA tuning if it's not able to saturate its SATA connections.
Add to that the fact that the two-channel "RAID" configuration is *nowhere near* twice as quick as the single bus - this thing urgently needs some performance tweaks to fully realise the potential of the concept.
Pretty sure you've got your definition of "factoid" messed up there champ. Like most people, you have assumed it means "little fact" or perhaps "little-known fact". Possibly due to abuse by CNN using the word in this sense.
From Wikipedia:
A factoid is a spurious - unverified, incorrect, or fabricated - statement formed and asserted as a fact, but with no veracity. The word appears in the Oxford English Dictionary as "something which becomes accepted as fact, although it may not be true.
It muddies the intention of the sentence when you use this word, because its meaning has been overloaded like this. I would have gone with:
"which contains an argument I have not seen mentioned before"
[dons flame-retardant suit]
I was on the Queensland - NSW border over the weekend and it most certainly is _not_ fine there!
Every single appointment time etc has to be qualified with "Queensland Time" or "New South Wales Time" after it.
Stores are having to stay open an hour longer per day to handle people expecting them to be open.
VCR programming becomes even more challenging for some people (is my program on a NSW-based channel or Qld channel?)
Most people on the border are using two sets of clocks.
Apparently there are still lots of not-too-bright people who say no to daylight saving "because it fades the curtains"...
The sooner Qld joins civilisation the better...
then create and use a TTF font that is "upside-down", and set your text direction as right-to-left.
!em rof skroW
- Staff who can bullshit their way through an interview, but that's all
- Lack of suitably-vicious firing policy for above
- Uneven hiring of different skillsets resulting in large portions of staff being totally idle
- Thus, low motivation to actually learn new things and get things done
- Promotion of staff to managerial positions based on time served rather than merit
- Zero repercussions when projects fail/slip horribly (e.g. an 18-month slip on a "6-month" project?!)
Then when the company decides they need a shiny new system there's no-one interested/able/willing to do it. Cue consulting company who brings in a team of bright-eyed and clued-up consultants. Sure we cost a lot but we'll get it done, done right, and on time, and afterwards we are much easier to get rid of than the permanent dross.Look at it from the company above's point-of-view - where they know they have (at least some of) the above issues with their permanent staff.
Damn right.
I used to work at a company that made the *test equipment* for big-iron routers. In other words, generated the packets at rates that made them scream for mercy. We used ASICs to serialize/deserialize onto big fat data buses and FPGAs to do stuff like CAMs, massively-parallel matching and byte sweeping. We used the biggest, baddest FPGAs Xilinx et al had to offer and had them bulging at the seams.
You won't get even close to saturating/dealing with a "serious" link (i.e. OC-12 or more) without being able to distribute the effort around the place - i.e. L2 framing and L3 inspection on a line card and massive backplane capacity. THEN you can use your general-purpose CPU to make routing decisions etc. If your Xeon or whatever has to get interrupted on every packet, or you have packets trying to make their way at line rates across a general-purpose bus like PCI etc, you're gonna hit a problem shortly after adding your second line card to your 4U Dell.
By the way, a few years ago we used our gear to do some tests on PC-based 155mbps ATM cards. On a "state of the art" Pentium III we got packet loss rates of 5-10%. That's with ONE card.
While I agree completely with all of your points (make Friday spectacular again!), I suspect what the FIA are trying to do is simplify some of their own procedures and reduce some of their costs. Currently all code on board an F1 car goes through pre- and post-race scrutineering just like every other part. Looking over the (compiled!) code for illegalities is a true nightmare.
:-)
Now imagine everyone is running ECU-OS 1.0 (ignoring all MS jokes for the time being)... The "OS" is exactly the same (i.e. it checksums to the same value) for all, only the various configuration parameters (held in RAM?) vary. Now the scrutineering effort becomes: hook up to ECU ROM, download code (or do a boundary scan), perform checksum. If the checksum isn't right - bang - you're disqualified.
To a lesser extent it will also save the teams some money. Rich teams might currently have 200 engine parameters that they can tweak. Poorer teams might have only been able to afford to develop 50. If everyone gets 100 parameters, it comes down to engineering quality rather than quantity to work out what works best.
Maybe...
"Considering she is at home looking after 5 kids and hence doesn't have a lot of time on her hands, she's doing a lot better than she would if she did nothing at all with that time. A few hundred really helps. "
A new Slashdot synonym for Vista?
What is surprising is that Ballmer and his boys actually tried...knowing what they know, they must have known they were on a fool's errand...
IF this is actually true and Ballmer really did bring the box in for his "top engineers" to fix (and not just some random support guys), can you not envision the battle for L337-ness that would ensue?:
Ballmer OK boys, I've tried to fix this POS and it's not responding to my advanced techniques. It's all yours
MS Uber-nerds Can we remove this chair from the monitor?
B OK...
MSU Can we remove the chair wedged in the floppy slot?
B I suppose, although I thought that might help Fucking Kill the spyware??? No???
MSU We'll work around it. But we're really busy you know, there's Vista to de-feature and we have to copy-and-paste a whole lot of code from Word into Excel, then Excel into Access and finally Access into IIS so Frontpage-published websites look good in IE. That's gonna take some time!
B I'll give $20M to whoever gets it going again.
MS Uber-nerd 1 I'd like to try first, sir.
MS Uber-nerd 2 I've been a loyal Microsoftie for longer sir, I'd like to try first
etc,etc,etc.
The issue of the manuals falling out-of-date is probably less relevant in a highly-controlled corporate environment - most large corps have Windows Update turned off on their desktops, I would imagine they would do the same with Ubuntu. Thus the manuals will stay relevant far longer.
On a non-corporate note, this is a big step towards grandma-compatibility too -
You cannot directly connect from the internet to a specific mobile phhone's IP address, regardless of the existence of a mobile web server there.
You can iff the phone initiated the communication. Here in Australia, all 3 major mobile networks use time-based firewall rules to allow a "push" *BACK* to a GPRS-enabled mobile device. This allows the device to get bidirectional communications with a server. Depending on the network you get between 15 seconds and 2 minutes to form up your reply packet and fire it back. After that, the device is "no longer there". You've also got to make sure that your reply packet is coming from the exact same IP:port combination that the phone sent to, or the firewall won't let it through. I imagine the networks in other countries do very similar things.
So if your server got hacked your mobile devices could get pingflooded - at AUS$10 (US$7.50) per megabyte received. Ouch.
Agree completely. The reason why someone would log into your phone's "website" would be because it has some "cool" data/real-time values on it that they want to see. So why not get the phone to send nice, snappy, minimal-overhead UDP packets to the back end of some *proper* HTTP server that can then wrap the values in pretty HTML/XML/CSS/what-have-you. Let it take any slashdotting for you, rather than your phone burning a hole in your pocket...