If even the dev abandons it, give me one rationalization or justification why the users shouldn't follow their lead and also abandon it... The dev probably quit for a good, well informed reason, and if the user knew it, they'd probably run for the hills too.
Yes, and "the hills" would be "some other browser", because I suspect that for many add-on developers the "good, well informed reason" is that they now have to spend a lot more time on their add-on because Mozilla decided to roll out a lot more releases per year.
Despite any automated system for maintaining compatibility, an add-on developer will have to spend time verifying that Mozilla has not introduced any new bugs that affect the add-on. With massive changes like the status bar disappearing, menus moving, and many other UI changes, it's quite likely that an add-on that does some UI will stop working correctly. Then, the developer will have to figure out how to make the add-on work with this week's version of the new, "improved" Firefox, while still working for users that use older versions (not old versions, just versions from 6 months ago).
If you're going to whine, at least use real world values. I get a 25% increase in clock speed on my Core i7-920 with a $40 aftermarket heatsink. With that same heatsink, you can get a 40% increase on Sandy Bridge CPUs.
Intel is massively underrating their current consumer CPUs, and it's pretty easy to purchase a cheaper CPU and spend the difference on an aftermarket heatsink and end up with a much better value.
Maybe if you're going the liquid route. That is a significant improvement in performance(not sure about 50% and 100%, though).
With the current Sandy Bridge -K chips (unlocked multiplier), it's fairly easy to get a 40% boost in speed with just a $50 aftermarket HSF. With water cooling, a boost of about 70% is possible.
There has always been the option of Retail (CPU Boxed with Cooler) or OEM (CPU and non-branded packaging).
I haven't seen any of the recent Intel CPUs offered as OEM. I'm sure big companies can get them, but resellers don't seem to have them.
Also, all Intel boxed CPUs (regardless of whether they come with a cooler, since Xeons never have) have a three-year warranty, while OEM is 90 days or less. Although it's rare for a CPU to fail down the road, the price difference usually worked out to about $10/year for peace of mind. I can't speak for CPUs, but I've never had an issue with Intel's service for other hardware RMAs.
What's interesting here is that Intel is saying that all chips of these types are capable of running at a faster speed.
Intel is just finally admitting what everybody already knows.
Since the first Core i7, every CPU Intel has sold can be trivially overclocked by 10%, by 20% with any decent motherboard and aftermarket cooler, and 40% with some hard work. Meanwhile, the new unlocked Sandy Bridge CPUs can be overclocked by 35% with any aftermarket cooler and no effort.
ISTR that there is a benchmark where "notepad.exe" is in the Startup folder, autologin is enabled, and the time is measured from BIOS handoff to the Notepad window appearing.
The parent is spot on... work and home machines are different beasts entirely. What it means to boot in the home setting is a fractional subset of what needs to be accomplished on boot for a work machine.
The difference isn't what needs to be done, but rather what management and IT believe must be done.
My normal login at home does pretty much the same as at work for real need...domain login, re-connect shares, anti-virus, etc. The difference is that I don't force the install of new programs to everyone at login when only 20% of the users need that program. Nor do I set the AV to do a full system scan at user login. Most IT departments are just clueless about how to correctly set up a system so that it is secure and provides users with what they need.
Also, it's not uncommon for power users to have much slower systems at work than at home, and when you combine that with the extra useless tasks that work machine has to do, it makes it an even worse comparison.
Nobody likes to sit and wait for a few minutes while their computer boots (well, for anything, really, who likes waiting?) but is a few minutes waiting for boot in the morning really worth the energy cost in the thing being on all night long consuming energy for no reason at all?
Absolutely, since it's not unusual to walk in to an "emergency" (as defined by somebody who could fire me) that needs to be handled in less time than it takes to bring the computer up to a usable level.
Although the time to a Windows desktop is only about 2 minutes on my machine, it's another 10-15 minutes to get logged in to all the resources I need to be able to get anything done. Since many of these are remote support for clients who aren't part of our network, they can't be handled by single-sign-on.
This is just a reporter and not a thorough one at that. But the original poster also pointed out FF needs good UI for letting old addons run.
Despite it's name, Add-on Compatibility Reporter also allows you to disable add-on version checking using a reasonable UI.
If you think about it for a minute, disabling version checking is a requirement, as it would be impossible for you to check if an add-on still works with the new version of Firefox if the add-on was disabled because of version checking.
Google Toolbar still doesn't officially support Firefox 5, but works just fine with version checking turned off. Since Google applications only support 3 versions of the browser, when Firefox 7 releases in about 3 months, it will then be impossible to find any version of Firefox that supports every Google product.
Yes Mom, I blocked you because you kept spamming me with Farmville crap so I didn't get your post that Grandma died.
But, since you didn't block Grandma, you should have gotten the news directly from her last status update...you know, the one that ended with "NO CARRIER".
They're talking about blacking out communications. And let's be honest - they're not going to just turn off the cell network; to do this right you'll have to kill landlines as well.
Sounds like professional criminals would like this sort of thing to happen, as it would make things like alarms less useful, and would prevent people from calling the police.
I had months of arguing over emails that did not arrive, because the other server was not setup to do a retry(as postgrey asks for a 5 min "cooldown")
If the other server did not retry, then it would return a failure to the client, and there would be no doubt the e-mail was never sent. But, I suspect that a server that violates one part of the RFC by not retrying would violate other parts, too. That really isn't your problem, and it was your job to explain to management why it wasn't worth the hassle to deal with such a company (hint: think about how convenient it is to have a "broken" e-mail server when you want to deny that you received an e-mail that would cost you money to follow the instructions).
Also, I don't know of any greylist implementation that can't be configured with permanent whitelisted IPs to deal with such problems, if management wants to work with a company that can't even set up a proper e-mail server (which speaks pretty poorly about their tech skills given that every Linux distro will out of the box give you a outgoing e-mail server that will follow the RFC).
The statements "Verizon does not oversell FiOS" and "you won't all be able to get it right now [due to everyone in the neighborhood using more than the concentrator will provide]" are mutually exclusive, just in case you didn't realize that when you were typing it.
The point is that Verizon won't oversell. You can't get higher than 80Mbps in areas where they have not upgraded their hardware because they won't sell it to you. They won't sell it to you because it could cause contention, and they will not do that. In areas where the hardware is upgraded, they will sell higher speeds, but still only up to the speed that will not cause contention.
In other words, FiOS bandwith is guaranteed because Verizon won't sell you a higher speed than they can guarantee, and they base their guarantee on the assumption that they will have 100% uptake. This means they have a lot of spare bandwidth, since they will never get 100% uptake. This is why current customers often get more than their rated speed (sometimes much more...I was getting 30Mbps while paying for 25Mbps).
This is what I meant by "you won't all be able to get it right now". You won't be able to buy a 100Mbps plan in a neighborhood where the concentrator doesn't support 100Mbps x (# of homes in neighborhood), regardless of the number of subscribers.
In fact, I don't know that JRE 5 users were ever auto-updated to 6, and if they were, it was after JRE 5 was EOL'd (roughly three years after Java 6 was released, and long after most people had moved on of their own accord).
No, Java 6 was pushed out in auto-update starting in 2007...that's the first time the Java app that is part of our site stopped working.
The whole point of Java 6 was that it would not let you choose which Java version would be used for browser apps, because Sun could not fix the bug that allowed malware to change that setting with no notice to the user.
That's as may be, but it's not illegal. In addition, pretty much every device that your packets traverse does some sort of "in-place modification". Whether that modification is "bad" is always a point of contention, and depends on who you ask.
Akamai: good or bad? Google cache pages: good or bad? Malware stripping web proxy? "Friendly" HTTP error messages? NAT? Captive-page HTTP to give you free Wi-Fi? Take some time and think about both sides of these examples, and you'll see that it's really not black and white.
That said, business connections seem to be untouched by almost every ISP, so if you really don't want to worry about modifications to your connections, spend the money.
If they were to alter your requests to, say, 8.8.8.8, then they would be deliberately violating their common-carrier status and exposing themselves to all kinds of lawyer-bait.
ISPs are not common carriers, and this sort of level of proxying happens all the time. In particular, many ISPs re-direct all outgoing connections to port 25 to their own mail server, and similarly all connections to port 53 (DNS) are sent to their own DNS server. It's not that they are "altering requests to 8.8.8.8", but rather they are altering requests to particular ports.
Also, almost every ISP blocks incoming requests to well-known "server" ports for their non-business customers. If "altering requests" was a problem, then every ISP would be in trouble for this.
No enterprises go and install the newest version of Java the day it comes out on their production apps.
For applications it may not be a big deal, but it is for desktops, for several reasons.
First, trying to turn off the Java auto-update is like killing cockroaches. Every time you do update manually (after any testing), it seems like the auto-update is re-enabled.
Second, although Oracle is not pushing Java 7 as the default download right now, it will likely happen soon, and then it becomes very difficult to locate older versions on the Oracle site (Sun was bad about this, too).
With all the incompatibilities introduced even with patches, this makes browser-based Java apps tough to keep working, which is why you see a lot of "must use IE 6 or 7 and Java 1.5.xx" instructions.
So even with our generous 10 watts saved, and assuming you can get away with this budget drives, You are still looking at over 6 years to break even.
Three to four years should break you even, since pretty much 100% of the power input for a spinning disk drive is turned into heat which must be removed by costly air conditioning.
Blocking results in Google search is built into the search...the Chrome extension is just a better UI.
For known domains, Google has a page that lets you quickly add sites to block. Note that you must be logged in to your Google account for this to work.
For arbitrary domains, do a search, click on the link in the Google results, then click the "back" in your browser. The Google result will now have a "block" option.
who are you getting a symmetric 5mbps for $250/month from? We are paying almost twice that. Is there any sort of SLA at that price?
He's getting it from somewhere that FiOS isn't available.
Seriously, if it wasn't for Verizon's decision to get into TV delivery (which has caused them no end of regulatory issues with local governments who have been bribed by cable companies), FiOS would now be so widespread that the days of paying more than $200/month for a business Internet connection would almost be at an end (unless you want something faster than 150/35).
And thus Captain Obvious answers his own question. If you're on a shared circuit and nobody else is using it then it's effectively acting dedicated.
OK, I know it's hard to understand, but all bandwidth is shared at some point.
For example, if ISP A peers to ISP B at 10Gbps, then even if a line is "dedicated", the fastest a customer of ISP A can get data from ISP B is 10Gbps. If there are 200 ISP A users with "dedicated" 100Mbps lines and all of them try to pull content from servers on ISP B, then they will not be able to do it at their "dedicated" speed. But, ISP A never said their "dedicated" 100Mbps line would get you that speed to every site...just to the inside of their network.
FiOS is exactly like that. Although the sharing is closer to you, Verizon still makes sure you have full bandwidth to their internal network. You share bandwidth with other users in your neighborhood, and if every home in your neighborhood subscribes to FiOS at 100Mbps, you won't all be able to get it right now, as not all of their network is upgraded...the concentrators would limit you to 80Mbps. In the future, that limit will be about 300Mbps.
Since Verizon does not oversell FiOS in any way (i.e., they limit the speed they sell in an area to what would be available if they had 100% uptake), you don't have a "dedicated" line, but you have exactly the same result.
And, if you pay for business FiOS, you do get guaranteed bandwidth in the agreement. It does cost more, but at $85/month for 25/25, it's down in the same range with "standard" connections from everybody else, and certainly not the several thousand dollars a month I suspect you think is required to get "dedicated" bandwidth.
Once FIOS expands, I'm sure the network will get bogged down and speeds will drop to below advertising speed.
Even with 100% uptake, FiOS can provide 80/80 service to every customer. The limiting factor is the neighborhood concentrators, which are slowly being upgraded to move the limit to somewhere near 300Mbps per customer.
Slashdot Mirror
the Titanic WILL sink in the end. There's no way around it
Unless somebody who really hates Celine Dion goes back in time and prevents the ship from sinking.
If even the dev abandons it, give me one rationalization or justification why the users shouldn't follow their lead and also abandon it... The dev probably quit for a good, well informed reason, and if the user knew it, they'd probably run for the hills too.
Yes, and "the hills" would be "some other browser", because I suspect that for many add-on developers the "good, well informed reason" is that they now have to spend a lot more time on their add-on because Mozilla decided to roll out a lot more releases per year.
Despite any automated system for maintaining compatibility, an add-on developer will have to spend time verifying that Mozilla has not introduced any new bugs that affect the add-on. With massive changes like the status bar disappearing, menus moving, and many other UI changes, it's quite likely that an add-on that does some UI will stop working correctly. Then, the developer will have to figure out how to make the add-on work with this week's version of the new, "improved" Firefox, while still working for users that use older versions (not old versions, just versions from 6 months ago).
If you're going to whine, at least use real world values. I get a 25% increase in clock speed on my Core i7-920 with a $40 aftermarket heatsink. With that same heatsink, you can get a 40% increase on Sandy Bridge CPUs.
Intel is massively underrating their current consumer CPUs, and it's pretty easy to purchase a cheaper CPU and spend the difference on an aftermarket heatsink and end up with a much better value.
Maybe if you're going the liquid route. That is a significant improvement in performance(not sure about 50% and 100%, though).
With the current Sandy Bridge -K chips (unlocked multiplier), it's fairly easy to get a 40% boost in speed with just a $50 aftermarket HSF. With water cooling, a boost of about 70% is possible.
There has always been the option of Retail (CPU Boxed with Cooler) or OEM (CPU and non-branded packaging).
I haven't seen any of the recent Intel CPUs offered as OEM. I'm sure big companies can get them, but resellers don't seem to have them.
Also, all Intel boxed CPUs (regardless of whether they come with a cooler, since Xeons never have) have a three-year warranty, while OEM is 90 days or less. Although it's rare for a CPU to fail down the road, the price difference usually worked out to about $10/year for peace of mind. I can't speak for CPUs, but I've never had an issue with Intel's service for other hardware RMAs.
What's interesting here is that Intel is saying that all chips of these types are capable of running at a faster speed.
Intel is just finally admitting what everybody already knows.
Since the first Core i7, every CPU Intel has sold can be trivially overclocked by 10%, by 20% with any decent motherboard and aftermarket cooler, and 40% with some hard work. Meanwhile, the new unlocked Sandy Bridge CPUs can be overclocked by 35% with any aftermarket cooler and no effort.
ISTR that there is a benchmark where "notepad.exe" is in the Startup folder, autologin is enabled, and the time is measured from BIOS handoff to the Notepad window appearing.
The parent is spot on... work and home machines are different beasts entirely. What it means to boot in the home setting is a fractional subset of what needs to be accomplished on boot for a work machine.
The difference isn't what needs to be done, but rather what management and IT believe must be done.
My normal login at home does pretty much the same as at work for real need...domain login, re-connect shares, anti-virus, etc. The difference is that I don't force the install of new programs to everyone at login when only 20% of the users need that program. Nor do I set the AV to do a full system scan at user login. Most IT departments are just clueless about how to correctly set up a system so that it is secure and provides users with what they need.
Also, it's not uncommon for power users to have much slower systems at work than at home, and when you combine that with the extra useless tasks that work machine has to do, it makes it an even worse comparison.
Nobody likes to sit and wait for a few minutes while their computer boots (well, for anything, really, who likes waiting?) but is a few minutes waiting for boot in the morning really worth the energy cost in the thing being on all night long consuming energy for no reason at all?
Absolutely, since it's not unusual to walk in to an "emergency" (as defined by somebody who could fire me) that needs to be handled in less time than it takes to bring the computer up to a usable level.
Although the time to a Windows desktop is only about 2 minutes on my machine, it's another 10-15 minutes to get logged in to all the resources I need to be able to get anything done. Since many of these are remote support for clients who aren't part of our network, they can't be handled by single-sign-on.
This is just a reporter and not a thorough one at that. But the original poster also pointed out FF needs good UI for letting old addons run.
Despite it's name, Add-on Compatibility Reporter also allows you to disable add-on version checking using a reasonable UI.
If you think about it for a minute, disabling version checking is a requirement, as it would be impossible for you to check if an add-on still works with the new version of Firefox if the add-on was disabled because of version checking.
Google Toolbar still doesn't officially support Firefox 5, but works just fine with version checking turned off. Since Google applications only support 3 versions of the browser, when Firefox 7 releases in about 3 months, it will then be impossible to find any version of Firefox that supports every Google product.
Yes Mom, I blocked you because you kept spamming me with Farmville crap so I didn't get your post that Grandma died.
But, since you didn't block Grandma, you should have gotten the news directly from her last status update...you know, the one that ended with "NO CARRIER".
They're talking about blacking out communications. And let's be honest - they're not going to just turn off the cell network; to do this right you'll have to kill landlines as well.
Sounds like professional criminals would like this sort of thing to happen, as it would make things like alarms less useful, and would prevent people from calling the police.
The theme is the very appropriately named Firefox 3 theme for Firefox 4+.
I had months of arguing over emails that did not arrive, because the other server was not setup to do a retry(as postgrey asks for a 5 min "cooldown")
If the other server did not retry, then it would return a failure to the client, and there would be no doubt the e-mail was never sent. But, I suspect that a server that violates one part of the RFC by not retrying would violate other parts, too. That really isn't your problem, and it was your job to explain to management why it wasn't worth the hassle to deal with such a company (hint: think about how convenient it is to have a "broken" e-mail server when you want to deny that you received an e-mail that would cost you money to follow the instructions).
Also, I don't know of any greylist implementation that can't be configured with permanent whitelisted IPs to deal with such problems, if management wants to work with a company that can't even set up a proper e-mail server (which speaks pretty poorly about their tech skills given that every Linux distro will out of the box give you a outgoing e-mail server that will follow the RFC).
The statements "Verizon does not oversell FiOS" and "you won't all be able to get it right now [due to everyone in the neighborhood using more than the concentrator will provide]" are mutually exclusive, just in case you didn't realize that when you were typing it.
The point is that Verizon won't oversell. You can't get higher than 80Mbps in areas where they have not upgraded their hardware because they won't sell it to you. They won't sell it to you because it could cause contention, and they will not do that. In areas where the hardware is upgraded, they will sell higher speeds, but still only up to the speed that will not cause contention.
In other words, FiOS bandwith is guaranteed because Verizon won't sell you a higher speed than they can guarantee, and they base their guarantee on the assumption that they will have 100% uptake. This means they have a lot of spare bandwidth, since they will never get 100% uptake. This is why current customers often get more than their rated speed (sometimes much more...I was getting 30Mbps while paying for 25Mbps).
This is what I meant by "you won't all be able to get it right now". You won't be able to buy a 100Mbps plan in a neighborhood where the concentrator doesn't support 100Mbps x (# of homes in neighborhood), regardless of the number of subscribers.
In fact, I don't know that JRE 5 users were ever auto-updated to 6, and if they were, it was after JRE 5 was EOL'd (roughly three years after Java 6 was released, and long after most people had moved on of their own accord).
No, Java 6 was pushed out in auto-update starting in 2007...that's the first time the Java app that is part of our site stopped working.
The whole point of Java 6 was that it would not let you choose which Java version would be used for browser apps, because Sun could not fix the bug that allowed malware to change that setting with no notice to the user.
That's as may be, but it's not illegal. In addition, pretty much every device that your packets traverse does some sort of "in-place modification". Whether that modification is "bad" is always a point of contention, and depends on who you ask.
Akamai: good or bad? Google cache pages: good or bad? Malware stripping web proxy? "Friendly" HTTP error messages? NAT? Captive-page HTTP to give you free Wi-Fi? Take some time and think about both sides of these examples, and you'll see that it's really not black and white.
That said, business connections seem to be untouched by almost every ISP, so if you really don't want to worry about modifications to your connections, spend the money.
If they were to alter your requests to, say, 8.8.8.8, then they would be deliberately violating their common-carrier status and exposing themselves to all kinds of lawyer-bait.
ISPs are not common carriers, and this sort of level of proxying happens all the time. In particular, many ISPs re-direct all outgoing connections to port 25 to their own mail server, and similarly all connections to port 53 (DNS) are sent to their own DNS server. It's not that they are "altering requests to 8.8.8.8", but rather they are altering requests to particular ports.
Also, almost every ISP blocks incoming requests to well-known "server" ports for their non-business customers. If "altering requests" was a problem, then every ISP would be in trouble for this.
No enterprises go and install the newest version of Java the day it comes out on their production apps.
For applications it may not be a big deal, but it is for desktops, for several reasons.
First, trying to turn off the Java auto-update is like killing cockroaches. Every time you do update manually (after any testing), it seems like the auto-update is re-enabled.
Second, although Oracle is not pushing Java 7 as the default download right now, it will likely happen soon, and then it becomes very difficult to locate older versions on the Oracle site (Sun was bad about this, too).
With all the incompatibilities introduced even with patches, this makes browser-based Java apps tough to keep working, which is why you see a lot of "must use IE 6 or 7 and Java 1.5.xx" instructions.
So even with our generous 10 watts saved, and assuming you can get away with this budget drives, You are still looking at over 6 years to break even.
Three to four years should break you even, since pretty much 100% of the power input for a spinning disk drive is turned into heat which must be removed by costly air conditioning.
Blocking results in Google search is built into the search...the Chrome extension is just a better UI.
For known domains, Google has a page that lets you quickly add sites to block. Note that you must be logged in to your Google account for this to work.
For arbitrary domains, do a search, click on the link in the Google results, then click the "back" in your browser. The Google result will now have a "block" option.
Business lines are far more expensive, but worth it if you actually want to get what is advertised.
Verizon generally only charges about 15-20% more for business service, but for the highest speeds it's even less (about 7%).
who are you getting a symmetric 5mbps for $250/month from? We are paying almost twice that. Is there any sort of SLA at that price?
He's getting it from somewhere that FiOS isn't available.
Seriously, if it wasn't for Verizon's decision to get into TV delivery (which has caused them no end of regulatory issues with local governments who have been bribed by cable companies), FiOS would now be so widespread that the days of paying more than $200/month for a business Internet connection would almost be at an end (unless you want something faster than 150/35).
And thus Captain Obvious answers his own question. If you're on a shared circuit and nobody else is using it then it's effectively acting dedicated.
OK, I know it's hard to understand, but all bandwidth is shared at some point.
For example, if ISP A peers to ISP B at 10Gbps, then even if a line is "dedicated", the fastest a customer of ISP A can get data from ISP B is 10Gbps. If there are 200 ISP A users with "dedicated" 100Mbps lines and all of them try to pull content from servers on ISP B, then they will not be able to do it at their "dedicated" speed. But, ISP A never said their "dedicated" 100Mbps line would get you that speed to every site...just to the inside of their network.
FiOS is exactly like that. Although the sharing is closer to you, Verizon still makes sure you have full bandwidth to their internal network. You share bandwidth with other users in your neighborhood, and if every home in your neighborhood subscribes to FiOS at 100Mbps, you won't all be able to get it right now, as not all of their network is upgraded...the concentrators would limit you to 80Mbps. In the future, that limit will be about 300Mbps.
Since Verizon does not oversell FiOS in any way (i.e., they limit the speed they sell in an area to what would be available if they had 100% uptake), you don't have a "dedicated" line, but you have exactly the same result.
And, if you pay for business FiOS, you do get guaranteed bandwidth in the agreement. It does cost more, but at $85/month for 25/25, it's down in the same range with "standard" connections from everybody else, and certainly not the several thousand dollars a month I suspect you think is required to get "dedicated" bandwidth.
Once FIOS expands, I'm sure the network will get bogged down and speeds will drop to below advertising speed.
Even with 100% uptake, FiOS can provide 80/80 service to every customer. The limiting factor is the neighborhood concentrators, which are slowly being upgraded to move the limit to somewhere near 300Mbps per customer.