Ok, who's going to fork over the MILLIONS necessary to build this "new" company? People, offices, servers, metric butt loads of bandwidth... this things add up fast. And at the end of the day, very few people will be willing to pay what it will take to make such a service profitable. I see enough people bitching about paying Tivo, Inc. $6.95 - $12.95 per month for guide data, software development, and customer support. And Tivo still isn't making a profit. So, wtf makes you think there are enough MythTV users out there willing to pay anything for guide data? I bet you 100$ more people will resort to scraping web pages before they'll *gulp* P. A. Y. for guide listings for their complicated homebrew "tivo". (btw, that's exactly what they were doing before Zap2It Labs.)
I'm thinking this sounds like a good job for Google, Inc. They already have the infrastructure. (and a wad of cash the size of a small moon.)
Sorry. WRONG. http://www.opencable.com/primer/cablecard_primer.h tml From the very early specifications and draft standards, the CableCARD module has been a two-way device.... The media has frequently reported that first-generation CableCARD 1.0 modules are one-way devices1. This is simply not true. CableLabs had always intended to develop the CableCARD module and host receiver standards with two-way capability. However the manufacturers of digital TVs requested that a host standard be developed that only had one-way capability.
All UDCP certified devices support all three means of headend signaling... even if it's only one way. That's the two protocols for the QPSK OOB channel ("motorola" and "scientific atlanta" protocols), and the modern DOCSIS Digital Set-top Gateway. In the case of the QPSK modem, the cablecard handles the protocol. For DOCSIS DSG, the cablecard communicates with the headend via the host's embeded cable modem. In both cases, it's still up to the host to provide the RF interface... a QPSK (de)modulator AND a DOCSIS cablemodem.
The CC is a conditional access mechanism. The CC has very little to do with VOD or iPPV -- it provides a channel map and optionally a QPSK modem for talking to the headend. The actual protocol for selecting or controlling a channel is not part of the cablecard spec. (that would be the mound of poo called OCAP.) The CC does not control the tuner(s)... if the host retunes the channel without telling the CC, it's not going to get a valid mpeg stream because it won't know to reset the decryptor.
The franchises are contractually obligated by the terms of their agreement with cable labs to insure that the CableCARD is not installed in an unauthorized device. Otherwise the cards' serial numbers could be entered at the head-end before they were even sent to the customer, and they could be activated via push-data broadcast over the entire network the same way non-CableCARD STBs are currently activated. The receiver ID wouldn't even be needed.
As I understand the process, the CC's serial number alone is not enough. Until it's paired with it's intended host, the necessary information for the headend isn't available.
It is important to note, the integrated stb's provided by cable providers don't have to be Cable Labs certified (btw, they aren't) as they are being disgned by/on behalf of the cable provider -- usually by the same people making the headend gear. Ultimately, the MSO can put what ever they want into their network. However, they are required by the FCC to support any Cable Labs certified device.
Also, stb's have been "addressable" for decades -- much longer than this digital tv BS has been going around.
As for Dish/DTV PPV, for most content, ordering by remote enables access immediately. That is, without any confirmation. The purchase is recorded on the access card and later reported to Dish/DTV sometime during the month. For other content -- say a boxing match, access is not granted until the purchase has been "confirmed" (read: reported and billed) Disconnecting the phone line for a long time, at least for DTV, will kill ordering via remote -- and supposedly revoke access to some sports subscriptions. In the case of cable, the channel may not even be available until someone has ordered it -- switched digital video; and it's no longer a static channel assignment so ordering by phone or web becomes problematic.
Each channel is encrypted with one key and broadcast across the network. That key is then encrypted individually for each cablecard entitled to decode it and broadcast across the network. The CC decodes the channel using the "master key" and then encodes it for the host using the key negotiated when the card was paired to the host. (That's why the card doesn't work when moved to another device, and why you cannot swap slots on a tivo.) This mess is intended to prevent interception of the raw data stream at any point. (of course, it's a trivial matter to "steal" it once it's on the host...)
The broadcast encryption key rotates. Quickly. Anywhere from several times per minute to several times per second. This is also supposed to discourage signal theft. And to be honest, without knowing what's inside the CC, it's an effective system. I'm betting it'll be game over in seconds once someone cracks open a CC... the industry doesn't have a good track record for these sorts of things.
Possible? Yes. Legally? No. There's currently 4 sources for digital programming... OTA, cable, DISH, and DirecTV.
OTA is trivial to deal with, but of course, there's very little content.
Cable is an "open system", but without a certified device, you're left with essentially the same as OTA, plus whatever unencrypted channels you can find -- which is increasingly ZERO. The FCC mandate only requires cableco's to support certified devices; they don't have to provide service to uncertified devices (and they won't, btw.)
DISH uses the world standard DVB. But, I really doubt they'll give you an access card for a system you built yourself. So, good luck decoding any of it. (assuming they still use a standard conditional access system.)
DirecTV uses their proprietary ("top secret") DSS protocol. Nobody makes receivers without a license. Their CAM is also classified. (unless you're an intern working the copy room:-))
64QAM vs. 256QAM makes no difference. Both transmit, bit-for-bit, exactly what they received from the content provider. It's DISH and DirecTV that re-encode the signal to MPEG-4. As for cable rebroadcast of OTA digital TV, they are required to broadcast them exactly as they receive them, without any modification -- that also means they cannot be encrypted, thus they are available without a cablecard.
Actually, the cablecard interface is very well documented, and PUBLIC. The only thing prevent any random person from creating their own PVR is the availability of hardware -- that nifty ATI card is only available to OEM's -- coupled with cableco's refusing to put a cablecard in non-certified hardware. Look at how that played out for linux and DVD playback...
First off, all cablecards are bidirectional. They always have been. Second, the only "on-site activation" is communication of the necessary serial numbers to the headend so the card will receive the correct decryption codes. It works just about the same as the smartcards in satellite receivers. The fact that 99% of cable operators require an expensive on-site installation is true bullshit... I can read the numbers on the screen just as well as a $50/hr tech. Let's look at DTV for a way to do it right... I can buy a new receiver and have it up and running in under 5 minutes without even touching the phone; adding a receiver is as easy as typing a number into a web form. (ok, 2 numbers... receiver id, and card number.) But they want their money; and they want to make sure the CC is placed in a certified device.
That said, there are some cableco's that will let you "self install" CC's all you want. In fact, they prefer it because it's faster and ultimately cheaper -- from the FCC's deployment reports, those self installs have far fewer issues.
If you've read the CC2.0 spec, that's exactly how it works... the preferred communications channel is via a host based embeded cable modem ("DOCSIS Digital Set-top Gateway") -- if they cannot get the CC's numbers entered correctly, what makes you think they'll get the cablemodem setup correctly? The CC controlled out-of-band QPSK upstream channel is still supported, but depreciated. VOD and impulse PPV cannot be deployed in a "unidirectional device" -- Cable Labs certification for UDCP's states that explicitly. There is no certification for bidirectional devices - yet. Switched Digital Video (SDV) is the current "screw you" scheme loophole cable co's are using to make 3rd party CC devices useless, and pressure the FCC to push the integration ban back even further.
The "warehouse" puts them in a box and sends them back to whomever made them to be reset. According to FCC reports (granted, those are old reports), the majority of CC failures are due to firmware updates not completing correctly. Moving a card from one host to another is a trivial action; it's only complicated because of untrained operators and overly complex headend systems requiring elaborate proceedures to unassign/reassign a card.
Cable companies hate cablecards because it means they no longer have a monopoly on set-top hardware. Cable operators have done everything they can to make CC deployment as big a pain in the ass as possible -- none (or very few) actively advertise cablecards; TW doesn't apply any bundle discounts to people who don't have their PoS DVR. And Switched Digital Video... don't believe the BS that it's all about capacity -- they have plenty of capacity to begin with (even more when they dump the analog tier); in the case of TW, it's pretty obvious they're doing it to make cablecards useless. (there's no active certification path for bi-directional CC devices -- only one has ever been demo'd and it's not yet certified by Cable Labs.) SDV is a proprietary process designed by Scientific Atlanta and Time Warner; it requires an "SDV client" application on the STB. It has nothing at all to do with any standards or certifications from Cable Labs.
They make all their GPL code available to anyone (www.tivo.com/linux), not just those to whom they have sold binaries as the GPL clearly states. And you are free to make as many modifications as you wish -- thousands of people already do. You'll have to replace the boot rom ("BIOS") -- that isn't even remotely GPL'd -- so it'll boot without a valid signature.
PLUS, Tivo's Linux kernel is based off 2.1. It cannot be retroactively re-licensed under GPLv3 - PERIOD. The other various tools are based on pretty old pre-GPLv3 source as well. I don't think the developers at Tivo, Inc. are very concerned with what's going on w.r.t. modern kernel development.
So how does its procedural code, trigger capability, and user defined functions compare to Oracle?
They don't. Oracle has a real programming language... PL/SQL. However, many will say writing library code for mysql is easier -- I'm one of them because I prefer C/C++.
Is it possible to port between them without jumping through too many hoops?
NO. Applications written for MySQL tend to have bits of MySQL "hackery" wired all through them for dealing with the way MySQL isn't a real database -- like truncating data without error, etc. Oracle applications will almost always have complex queries that MySQL *cannot* process AT ALL. (and if it does understand the query, it'll take days to give the correct result(s).) MySQL has gotten better in this respect since v3, but it's still a joke for any serious use of SQL.
Does MySQL support partitioning?
No. Or at least, not the last time I looked. I think v5 introduced the ability to have multiple files for one table, but writes land in the last file; and an index is still one file.
How about performance with large dbs?
MySQL performance drops measurablly the larger the database tables become. Oracle scales very well to HUGE tables -- due in large part to partitioning and efficient index usage.
For small "junk" sites, MySQL gets the job done nicely... for free; and "free" is a very powerful motivator. (Btw, one can use Sybase ASE for free, too, but very few people do.) If you're serious and care about your data, as a rule, you don't use MySQL. Wikipedia does, but they're insane. (read above re: free)
Much like the early days of IPv4... IPv6 is a classful system. This simply will never work. Even with the network part of your address being asigned by Magic(tm), it's still a pain in the ass. Does your DNS records also update by the same Magic(tm)? (NO) By forcing IPv6 into a classful world, they've actually significantly reduced the address space. This, btw, is the exact same mistake from the first days of IPv4.
If it were, in fact, a "tax" then that $2.33 would be handed over to the FCC. It isn't. The fact is, 99% of what you pay for your phone goes right in the telco's pockets. The state and federal taxes are the only thing they hand over. (ala sales tax) All those "fees" are bullshit padding of your bill. In every case I've ever checked it was a way to get around tariffs -- they cannot legally change the service price, but they can add "cost recovery" fees and other crap to increase the total monthly bill.
All Fluorescent Lamps and Tubes Should Be Recycled or Disposed as Hazardous Waste
All fluorescent lamps and tubes are considered hazardous waste in California when they are discarded because they contain mercury. (Title 22, division 4.5, chapter 11, section 66261.50)
... All fluorescent lamps and tubes must be recycled, or taken to a household hazardous waste disposal facility, a universal waste handler (e.g., storage facility or broker), or an authorized recycling facility. (Title 22, division 4.5, chapter 23, section 66273.8)
Indeed. Honeywell makes consumer thermostats that can do just that. It's designed for selecting an efficient, cost effective heating/cooling source. (i.e. gas, heat pump, etc.) You can make one of the sources a vent fan. (I don't think any homeowner does that 'tho.)
It's sort of ironic that they turn the AC (cooling) off in the evening because that's when the system will be most efficient and not be need as often with most/all of the people out of the building. (The non-people heat load will be nearly constant. Letting it build up overnight and pumping it out in the morning after the sun comes up and starts heating the outside air is a little bit stupid.)
Save money from the lights maybe, but the AC? Doubtful. Office complexes strive to be energy efficient -- keep the cold side cold and hot side hot. In many instances this gives them a tax and/or energy break. That also translates to less savings from turning off the AC. The heat that builds up in the office isn't going anywhere unless the HVAC systems move it there. So, all the savings you think you make up in turning the AC off all night are consumed in the morning when the AC cuts back on. There will be marginal savings from pumping out the heat when it's cooler outside (making the transfer more efficient) and a tiny savings from running the compressor(s) less, but I've never seen an office building leak enough heat to really be worth it. And ultimately, the varying temperature significantly reduces the lifespan of office equipment. (your hard drive and monitor will last a lot longer if it's turned on an left on.)
I'm not asking you to prove YOUR experience, I'm demanding you prove MY experience as you seem to know more about what I've seen than I do.
Again you put words in my mouth... I'm not saying one way or the other what poeple might do if they had the chance. I'm am stating my experience. Even though it's been illegal to throw aluminum cans in the trash in Raleigh (NC) for several years now, PEOPLE STILL DO. And everyone with municiple trash pickup gets *2* green recycling bins. So, obviously they have a choice and choose not to. One might extrapolate that some people might not recycle light bulbs either. As there's no obvious places to recycle them around here, it's not much of an option.
Maybe CA has those bins because state and/or local laws require it?
"unsupported claim"? Ok. YouPROVE that I've actually seen someone recycle a light bulb. "I've never seen"... I'm not in Oakland, CA. And never have been.
Supposedly Home Depot and the like accept batteries for recycling too. However, the only place I've ever seen with the "put your batteries here" boxes was an actual battery store (Batteries Plus.)
Like that HEPA filter is going to trap mercury vapor? I'm not even sure activated charcoal will capture it.
The problem here is not the vapor in the bulb, but the bead of mercury that was spilled and now irreversiblly mixed into the carpet, padding, and floor.
Unless it's sealed inside an air tight container, it will gas and enter the atmosphere. I've never seen anyone, anywhere recycle a light bulb. Any light bulb.
All fluorescent lamps contain mercury. They always have. Why is this suddenly a revelation to people?
Lost source isn't the only problem... Source that no one (alive) can understand is also a growing problem. There are alot of old, obscure programming systems out there. Nobody teaches COBOL anymore:-)
heh. you may wish to check the back of a smart jack and see how many pairs support a single t1:) because you are dead wrong. and a T1 ***IS*** dsl.
And now...
It's clear that I was speaking of how the service gets from the CO to the prem.
I repeat: A T1 is NOT DSL. A T1's electrical interface and signal are defined by ANSI; that spec is a 4 wire (2 pair) electrical interface.
the fact still remains, T1 ***IS*** just another flavor of DSL.
No. It isn't. It isn't any flavor of DSL. A T1 is a T1. HDSL is HDSL. ADSL is ADSL. When you order a T1, you order a specific format and interface; 2 pairs exactly as ANSI T1.403 specifies. Because one thing can be repackaged as or transported via some other thing does not change what it is. The digital information within a T1 can be carried by just about anything, however, it has to be a T1 when it gets to the end.
A T1 is no more DSL than it is any of the other multitudes of technologies used to transport them today. When you order a "T1", the telco doesn't show up, hand you a single pair of wires and leave. That single pair of wires is NOT your T1. It cannot be plugged into a PBX, channel bank, CSU/DSU, or ANYTHING with a T1 interface -- either a DB15 or RJ45. Because it's not a fucking T1; it's not the same electrical interface or signal. The only thing that can make sense of what's on that pair of wires is a matching converter to give you a real, ANSI compliant T1.
Is my ethernet "ISDN" because it's carried across a pair of Combinet ISDN bridges? NO. It's not the same interface, and it's not the same signaling. (it's not even remotely the same speed.)
I've been to many of the CO's in my area, there is alot more than 300ft of copper in them.
I wasn't talking about the 30years of spaghetti in a CO. I said my T1 from TW. In that entire voice/data T1 setup, the only copper is between the DMX (fibre) in the 1st floor phone room to my server room on the 3rd floor. Unless the DMX on the other end(s) (voice and data could drop out at different places) feed the DMS (voice) and router (data) via DS3, that's the only copper in the entire loop. (The tech's circuit design doesn't show how the far ends connect other than channel numbers.)
To get back to the original FA: T1's are expensive because they've always been expensive. Tariffs agreed upon by the FCC and state PUCs that set the prices have been around longer than I've been alive. Back then, a T1 really was a complicated, costly thing; consuming both manpower to setup and maintain, and quite limited resources (there's only so many wires.) None of that's true today, but the tariffs still remain -- it's pure profit for telcos, so they literally fight to the death to keep them.
You aren't listening. A T1 interface ("A T1", ANSI T1.403) is two balanced pairs. HOW THE TELCO TRANSPORTS IT DOESN'T MATTER. What the telco delivers to you is 2 pairs. What you plug into your CSU or router is 2 pairs.
For the record, T1's have been 100% digitally switched for decades. As such, the end-to-end copper loop T1 passed into oblivion long ago; and with it almost all of the provisioning "overhead". However, all the tariffs are still in place from the era where it was Real Work to hook up a T1 -- i.e. it took real people connecting wires instead of today's mouse click digital switching. The T1 feeding the office here is maybe 300ft of copper. Total. And that's the run through the walls to the fibre node in the 1st floor wiring closet. (where TimeWarner has a DMX shelf)
I've been around them for 15 years.... TWO PAIRS How your cheap, lame telco carries it through their network is immaterial; a T1 is a TX pair and an RX pair. T1's are NOT DSL; the technology predates DSL by several decades. Go buy the spec and read it for your self... a T1 network interface is two f'ing pairs.
Negative. A T1 is two pairs. ("balanced pairs") One TX (+/-) and one RX (+/-). Anything else... Is. Not. A. T1. That said, there are a lot of toys out there to convert a T1 to a "single pair". Everyone I've even seen or heard of was, in fact, a DSL bridge; usually VHDSL, but not always. There's one at both ends in order to provide a true T1 interface. (via either an RJ45 or DB15 connector.)
Such technology has been popular for about a decade. Copper is expensive realestate. And only so many T1's can be on the same trunk before crosstalk starts screwing up POTS and the other T1's.
Slashdot Mirror
Ok, who's going to fork over the MILLIONS necessary to build this "new" company? People, offices, servers, metric butt loads of bandwidth... this things add up fast. And at the end of the day, very few people will be willing to pay what it will take to make such a service profitable. I see enough people bitching about paying Tivo, Inc. $6.95 - $12.95 per month for guide data, software development, and customer support. And Tivo still isn't making a profit. So, wtf makes you think there are enough MythTV users out there willing to pay anything for guide data? I bet you 100$ more people will resort to scraping web pages before they'll *gulp* P. A. Y. for guide listings for their complicated homebrew "tivo". (btw, that's exactly what they were doing before Zap2It Labs.)
I'm thinking this sounds like a good job for Google, Inc. They already have the infrastructure. (and a wad of cash the size of a small moon.)
http://www.opencable.com/primer/cablecard_primer.
From the very early specifications and draft standards, the CableCARD module has been a two-way device.
The media has frequently reported that first-generation CableCARD 1.0 modules are one-way devices1. This is simply not true. CableLabs had always intended to develop the CableCARD module and host receiver standards with two-way capability. However the manufacturers of digital TVs requested that a host standard be developed that only had one-way capability.
All UDCP certified devices support all three means of headend signaling... even if it's only one way. That's the two protocols for the QPSK OOB channel ("motorola" and "scientific atlanta" protocols), and the modern DOCSIS Digital Set-top Gateway. In the case of the QPSK modem, the cablecard handles the protocol. For DOCSIS DSG, the cablecard communicates with the headend via the host's embeded cable modem. In both cases, it's still up to the host to provide the RF interface... a QPSK (de)modulator AND a DOCSIS cablemodem.
The CC is a conditional access mechanism. The CC has very little to do with VOD or iPPV -- it provides a channel map and optionally a QPSK modem for talking to the headend. The actual protocol for selecting or controlling a channel is not part of the cablecard spec. (that would be the mound of poo called OCAP.) The CC does not control the tuner(s)... if the host retunes the channel without telling the CC, it's not going to get a valid mpeg stream because it won't know to reset the decryptor.As I understand the process, the CC's serial number alone is not enough. Until it's paired with it's intended host, the necessary information for the headend isn't available.
It is important to note, the integrated stb's provided by cable providers don't have to be Cable Labs certified (btw, they aren't) as they are being disgned by/on behalf of the cable provider -- usually by the same people making the headend gear. Ultimately, the MSO can put what ever they want into their network. However, they are required by the FCC to support any Cable Labs certified device.
Also, stb's have been "addressable" for decades -- much longer than this digital tv BS has been going around.
As for Dish/DTV PPV, for most content, ordering by remote enables access immediately. That is, without any confirmation. The purchase is recorded on the access card and later reported to Dish/DTV sometime during the month. For other content -- say a boxing match, access is not granted until the purchase has been "confirmed" (read: reported and billed) Disconnecting the phone line for a long time, at least for DTV, will kill ordering via remote -- and supposedly revoke access to some sports subscriptions. In the case of cable, the channel may not even be available until someone has ordered it -- switched digital video; and it's no longer a static channel assignment so ordering by phone or web becomes problematic.
Each channel is encrypted with one key and broadcast across the network. That key is then encrypted individually for each cablecard entitled to decode it and broadcast across the network. The CC decodes the channel using the "master key" and then encodes it for the host using the key negotiated when the card was paired to the host. (That's why the card doesn't work when moved to another device, and why you cannot swap slots on a tivo.) This mess is intended to prevent interception of the raw data stream at any point. (of course, it's a trivial matter to "steal" it once it's on the host...)
The broadcast encryption key rotates. Quickly. Anywhere from several times per minute to several times per second. This is also supposed to discourage signal theft. And to be honest, without knowing what's inside the CC, it's an effective system. I'm betting it'll be game over in seconds once someone cracks open a CC... the industry doesn't have a good track record for these sorts of things.
Possible? Yes. Legally? No. There's currently 4 sources for digital programming... OTA, cable, DISH, and DirecTV.
:-))
OTA is trivial to deal with, but of course, there's very little content.
Cable is an "open system", but without a certified device, you're left with essentially the same as OTA, plus whatever unencrypted channels you can find -- which is increasingly ZERO. The FCC mandate only requires cableco's to support certified devices; they don't have to provide service to uncertified devices (and they won't, btw.)
DISH uses the world standard DVB. But, I really doubt they'll give you an access card for a system you built yourself. So, good luck decoding any of it. (assuming they still use a standard conditional access system.)
DirecTV uses their proprietary ("top secret") DSS protocol. Nobody makes receivers without a license. Their CAM is also classified. (unless you're an intern working the copy room
64QAM vs. 256QAM makes no difference. Both transmit, bit-for-bit, exactly what they received from the content provider. It's DISH and DirecTV that re-encode the signal to MPEG-4. As for cable rebroadcast of OTA digital TV, they are required to broadcast them exactly as they receive them, without any modification -- that also means they cannot be encrypted, thus they are available without a cablecard.
Actually, the cablecard interface is very well documented, and PUBLIC. The only thing prevent any random person from creating their own PVR is the availability of hardware -- that nifty ATI card is only available to OEM's -- coupled with cableco's refusing to put a cablecard in non-certified hardware. Look at how that played out for linux and DVD playback...
First off, all cablecards are bidirectional. They always have been. Second, the only "on-site activation" is communication of the necessary serial numbers to the headend so the card will receive the correct decryption codes. It works just about the same as the smartcards in satellite receivers. The fact that 99% of cable operators require an expensive on-site installation is true bullshit... I can read the numbers on the screen just as well as a $50/hr tech. Let's look at DTV for a way to do it right... I can buy a new receiver and have it up and running in under 5 minutes without even touching the phone; adding a receiver is as easy as typing a number into a web form. (ok, 2 numbers... receiver id, and card number.) But they want their money; and they want to make sure the CC is placed in a certified device.
That said, there are some cableco's that will let you "self install" CC's all you want. In fact, they prefer it because it's faster and ultimately cheaper -- from the FCC's deployment reports, those self installs have far fewer issues.
If you've read the CC2.0 spec, that's exactly how it works... the preferred communications channel is via a host based embeded cable modem ("DOCSIS Digital Set-top Gateway") -- if they cannot get the CC's numbers entered correctly, what makes you think they'll get the cablemodem setup correctly? The CC controlled out-of-band QPSK upstream channel is still supported, but depreciated. VOD and impulse PPV cannot be deployed in a "unidirectional device" -- Cable Labs certification for UDCP's states that explicitly. There is no certification for bidirectional devices - yet. Switched Digital Video (SDV) is the current "screw you" scheme loophole cable co's are using to make 3rd party CC devices useless, and pressure the FCC to push the integration ban back even further.
The "warehouse" puts them in a box and sends them back to whomever made them to be reset. According to FCC reports (granted, those are old reports), the majority of CC failures are due to firmware updates not completing correctly. Moving a card from one host to another is a trivial action; it's only complicated because of untrained operators and overly complex headend systems requiring elaborate proceedures to unassign/reassign a card.
Cable companies hate cablecards because it means they no longer have a monopoly on set-top hardware. Cable operators have done everything they can to make CC deployment as big a pain in the ass as possible -- none (or very few) actively advertise cablecards; TW doesn't apply any bundle discounts to people who don't have their PoS DVR. And Switched Digital Video... don't believe the BS that it's all about capacity -- they have plenty of capacity to begin with (even more when they dump the analog tier); in the case of TW, it's pretty obvious they're doing it to make cablecards useless. (there's no active certification path for bi-directional CC devices -- only one has ever been demo'd and it's not yet certified by Cable Labs.) SDV is a proprietary process designed by Scientific Atlanta and Time Warner; it requires an "SDV client" application on the STB. It has nothing at all to do with any standards or certifications from Cable Labs.
They make all their GPL code available to anyone (www.tivo.com/linux), not just those to whom they have sold binaries as the GPL clearly states. And you are free to make as many modifications as you wish -- thousands of people already do. You'll have to replace the boot rom ("BIOS") -- that isn't even remotely GPL'd -- so it'll boot without a valid signature.
PLUS, Tivo's Linux kernel is based off 2.1. It cannot be retroactively re-licensed under GPLv3 - PERIOD. The other various tools are based on pretty old pre-GPLv3 source as well. I don't think the developers at Tivo, Inc. are very concerned with what's going on w.r.t. modern kernel development.
For small "junk" sites, MySQL gets the job done nicely... for free; and "free" is a very powerful motivator. (Btw, one can use Sybase ASE for free, too, but very few people do.) If you're serious and care about your data, as a rule, you don't use MySQL. Wikipedia does, but they're insane. (read above re: free)
Much like the early days of IPv4... IPv6 is a classful system. This simply will never work. Even with the network part of your address being asigned by Magic(tm), it's still a pain in the ass. Does your DNS records also update by the same Magic(tm)? (NO) By forcing IPv6 into a classful world, they've actually significantly reduced the address space. This, btw, is the exact same mistake from the first days of IPv4.
You could also draw the conclusion that pr0n was what was crashing the thing. After all, they rarely crashed when not full of porn.
If it were, in fact, a "tax" then that $2.33 would be handed over to the FCC. It isn't. The fact is, 99% of what you pay for your phone goes right in the telco's pockets. The state and federal taxes are the only thing they hand over. (ala sales tax) All those "fees" are bullshit padding of your bill. In every case I've ever checked it was a way to get around tariffs -- they cannot legally change the service price, but they can add "cost recovery" fees and other crap to increase the total monthly bill.
http://www.ciwmb.ca.gov/WPIE/FluoresLamps/
Indeed. Honeywell makes consumer thermostats that can do just that. It's designed for selecting an efficient, cost effective heating/cooling source. (i.e. gas, heat pump, etc.) You can make one of the sources a vent fan. (I don't think any homeowner does that 'tho.)
It's sort of ironic that they turn the AC (cooling) off in the evening because that's when the system will be most efficient and not be need as often with most/all of the people out of the building. (The non-people heat load will be nearly constant. Letting it build up overnight and pumping it out in the morning after the sun comes up and starts heating the outside air is a little bit stupid.)
Save money from the lights maybe, but the AC? Doubtful. Office complexes strive to be energy efficient -- keep the cold side cold and hot side hot. In many instances this gives them a tax and/or energy break. That also translates to less savings from turning off the AC. The heat that builds up in the office isn't going anywhere unless the HVAC systems move it there. So, all the savings you think you make up in turning the AC off all night are consumed in the morning when the AC cuts back on. There will be marginal savings from pumping out the heat when it's cooler outside (making the transfer more efficient) and a tiny savings from running the compressor(s) less, but I've never seen an office building leak enough heat to really be worth it. And ultimately, the varying temperature significantly reduces the lifespan of office equipment. (your hard drive and monitor will last a lot longer if it's turned on an left on.)
I'm not asking you to prove YOUR experience, I'm demanding you prove MY experience as you seem to know more about what I've seen than I do.
Again you put words in my mouth... I'm not saying one way or the other what poeple might do if they had the chance. I'm am stating my experience. Even though it's been illegal to throw aluminum cans in the trash in Raleigh (NC) for several years now, PEOPLE STILL DO. And everyone with municiple trash pickup gets *2* green recycling bins. So, obviously they have a choice and choose not to. One might extrapolate that some people might not recycle light bulbs either. As there's no obvious places to recycle them around here, it's not much of an option.
Maybe CA has those bins because state and/or local laws require it?
"unsupported claim"? Ok. You PROVE that I've actually seen someone recycle a light bulb. "I've never seen"... I'm not in Oakland, CA. And never have been.
Supposedly Home Depot and the like accept batteries for recycling too. However, the only place I've ever seen with the "put your batteries here" boxes was an actual battery store (Batteries Plus.)
Like that HEPA filter is going to trap mercury vapor? I'm not even sure activated charcoal will capture it.
The problem here is not the vapor in the bulb, but the bead of mercury that was spilled and now irreversiblly mixed into the carpet, padding, and floor.
Unless it's sealed inside an air tight container, it will gas and enter the atmosphere. I've never seen anyone, anywhere recycle a light bulb. Any light bulb.
All fluorescent lamps contain mercury. They always have. Why is this suddenly a revelation to people?
Lost source isn't the only problem... Source that no one (alive) can understand is also a growing problem. There are alot of old, obscure programming systems out there. Nobody teaches COBOL anymore :-)
Your original post...And now...I repeat: A T1 is NOT DSL. A T1's electrical interface and signal are defined by ANSI; that spec is a 4 wire (2 pair) electrical interface.No. It isn't. It isn't any flavor of DSL. A T1 is a T1. HDSL is HDSL. ADSL is ADSL. When you order a T1, you order a specific format and interface; 2 pairs exactly as ANSI T1.403 specifies. Because one thing can be repackaged as or transported via some other thing does not change what it is. The digital information within a T1 can be carried by just about anything, however, it has to be a T1 when it gets to the end.
A T1 is no more DSL than it is any of the other multitudes of technologies used to transport them today. When you order a "T1", the telco doesn't show up, hand you a single pair of wires and leave. That single pair of wires is NOT your T1. It cannot be plugged into a PBX, channel bank, CSU/DSU, or ANYTHING with a T1 interface -- either a DB15 or RJ45. Because it's not a fucking T1; it's not the same electrical interface or signal. The only thing that can make sense of what's on that pair of wires is a matching converter to give you a real, ANSI compliant T1.
Is my ethernet "ISDN" because it's carried across a pair of Combinet ISDN bridges? NO. It's not the same interface, and it's not the same signaling. (it's not even remotely the same speed.)I wasn't talking about the 30years of spaghetti in a CO. I said my T1 from TW. In that entire voice/data T1 setup, the only copper is between the DMX (fibre) in the 1st floor phone room to my server room on the 3rd floor. Unless the DMX on the other end(s) (voice and data could drop out at different places) feed the DMS (voice) and router (data) via DS3, that's the only copper in the entire loop. (The tech's circuit design doesn't show how the far ends connect other than channel numbers.)
To get back to the original FA: T1's are expensive because they've always been expensive. Tariffs agreed upon by the FCC and state PUCs that set the prices have been around longer than I've been alive. Back then, a T1 really was a complicated, costly thing; consuming both manpower to setup and maintain, and quite limited resources (there's only so many wires.) None of that's true today, but the tariffs still remain -- it's pure profit for telcos, so they literally fight to the death to keep them.
You aren't listening. A T1 interface ("A T1", ANSI T1.403) is two balanced pairs. HOW THE TELCO TRANSPORTS IT DOESN'T MATTER. What the telco delivers to you is 2 pairs. What you plug into your CSU or router is 2 pairs.
For the record, T1's have been 100% digitally switched for decades. As such, the end-to-end copper loop T1 passed into oblivion long ago; and with it almost all of the provisioning "overhead". However, all the tariffs are still in place from the era where it was Real Work to hook up a T1 -- i.e. it took real people connecting wires instead of today's mouse click digital switching. The T1 feeding the office here is maybe 300ft of copper. Total. And that's the run through the walls to the fibre node in the 1st floor wiring closet. (where TimeWarner has a DMX shelf)
I've been around them for 15 years.... TWO PAIRS How your cheap, lame telco carries it through their network is immaterial; a T1 is a TX pair and an RX pair. T1's are NOT DSL; the technology predates DSL by several decades. Go buy the spec and read it for your self... a T1 network interface is two f'ing pairs.
Negative. A T1 is two pairs. ("balanced pairs") One TX (+/-) and one RX (+/-). Anything else... Is. Not. A. T1. That said, there are a lot of toys out there to convert a T1 to a "single pair". Everyone I've even seen or heard of was, in fact, a DSL bridge; usually VHDSL, but not always. There's one at both ends in order to provide a true T1 interface. (via either an RJ45 or DB15 connector.)
Such technology has been popular for about a decade. Copper is expensive realestate. And only so many T1's can be on the same trunk before crosstalk starts screwing up POTS and the other T1's.