While what you say is true, it's not just the WRT54GL that's likely affected. It's highly likely that WRT54G/WRT54GS v1-v4 are also affected.
The WRT54GL router uses nearly identical firmware to the v1, v2, v3, and v4 models of the WRT54G and WRT54GS router. In fact, the specs on a WRT54G/WRT54GS v4 are nearly identical to a WRT54GL--same chipset, RAM, and flash.
A lot of these were manufactured and sold, and I bet the vast majority are still running stock.
I love Tomato too--in fact, I use it at my house. However, Tomato was originally based off Stock Linksys, and might also be affected. Until full disclosure occurs, we'll not know for sure.
Wow, after reading the recall, this is SUPER stupid.
Personally, I don't think the government needs to be our nanny. But I at least understand the motivation of trying to keep things like these away from 2-3 year old because they like, yanno, eat them and it causes problems (even though there's only 2 documented cases).
The recall is this:
Between March 2009 through March 2010, buckballs were sold labeled 13+. The federal law requires powerful magnets to be 14+. This was corrected in March 2010 when the magnets were labeled "keep away from all children".
So this recall protects all of those THIRTEEN year old children out there who might accidentally eat a buckyball because they didn't read the packaging. I'm sure that's, like, super common. Must. happen. a. lot.
I use RAID 1 for backups all the time. I shutdown the system, pull a drive, swap it out, and then restart the system. While system is up, replace second drive with a new drive and reimage over.
Total downtime: 5 minutes or so. Degraded performance for a few hours during rebuild.
Basically, if you use RAID 1 like tapes, where each HD is a tape, it can be an extremely economical and reliable way to backup data. You know your data's being backed up because you're actually running off that data. You can't have a situation where the tape drive fails to write data and then suddenly when you have a disaster, you find out that you have 6 months of blank tapes (I've had that happen).
At $100 per tape (actually HD+case), it's very cheap per 2TB of backups.
This is absolutely true, but only on pressurized aircraft. It's the stress/release of the metal which causes metal fatigue and eventual failure of the airframe.
I expect (but don't know) that the 727 they used for the show had exceeded its "safe" pressurization cycles and was destined for the scrapyard anyway--so it only had scrap value anyway.
However, the DC3 that dblll mentioned is NOT a pressurized aircraft--so it need not worry about cycles at all. It's only about flight hours and wear and tear. As such, there's still many DC3's in operation today--some even commercially.
I'm sorry. He signed the bill into law. Had he truly been opposed to it, he could have vetoed it. Sure, his veto could have been overruled, but at least he'd be officially on the record that he didn't support it.
Also, Obama specifically requested the removal of language that said the NDAA would not apply to US citizens or lawful residents. See:
http://www.youtube.com/watch?v=4DNDHbT44cY
I'm sorry--I think there's a big difference between "swiftboat veterans" for "truth" and "Obama supported the NDAA and specifically requested some of the draconian language in the bill"
So how exactly is a processor running a program to implement another instruction set architecture, with the main memory used by the implemented ISA being accessed by special operations, and with the program and its internal data existing in a separate block of memory, different from, say, a (vertical) microcode engine, running microcode to implement another instruction set architecture, with the main memory used by the implemented ISA being accessed by special microcode operations, and with the microprogram and its internal data existing in a separate block of memory?
Thanks for actually posting facts. It's refreshing, for a change. You must be new here....
Anecdotal: I no longer fly due to TSA regulations. I either drive or take a train. Yes, it's tremendously inconvenient. Yes, I hate it. No, I don't think there's a lot of people like me...
...okay, I don't even know what to say to that. I have no idea what it's like on your planet, but around here we're only human. No wonder developers aren't up to your standards....
Totally agree. I was initially inclined to say (s)he's trolling, but (s)he's clearly quite learned in computers. Maybe (s)he expects that all people are just that smart... Expecting that people get parallel programs right on the first try, given their complexity is not reasonable, at least where I work (myself included). In fact, I was just working with a developer today to fix a reader/writer issue triggered by parallelism both in code and in writing to the DB. We had to sit down and think out the use cases for about an hour before we had a good working solution.
Parallel programming is hard. It's necessary, but hard, at least to us "normal" people.:)
The same ones that invented the internet, lasers, microchips, GPSs, and microwave ovens, went to the moon, built the largest highway system in the world, created our clean water infrastructure, and electrified most of a continent.
How about the state governments do those things? Pretty sure the state governments can figure out how to build roads, clean water, and manage electrical grids. The only item on the list that I see as problematic would be "went to the moon".
I'm very supportive of minimizing the federal government. That doesn't mean I don't want government services at all--we have state and local governments, after all...
(I've used Accordance RAID systems in the past, before my current Raidon).
Then I buy an additional HD tray and rotate the 3 drives occasionally to an offsite backup. Basically the cost is about $345 total, $75*3 drives+$120 for the unit, and I get 3 copies of the data. Additional offsite backups run $105 ($75 drive+$30 tray).
I've found this the easiest to understand (for end users) and a very cheap and reliable storage method. I've had the nightmare of backing up to 8 different tapes, only to go to restore and find out that 6 tapes are empty, 1 is bad, and 1 has data from 6 months back. Because you're actually running your OS off of the drives, you don't get into that situation. The main key is to make sure that you pull drives when the system is off (so that all files are written to disk), and that you insert drives when the system is ON (so that the raid array knows the drive is replaced and doesn't corrupt your data). NEVER INSERT A DRIVE WHILE THE SYSTEM IS OFF.
I have files from over 15 years ago on my main desktop computer, and the only files I'm missing are from my original 286 before I started this backup method.
This solution can be retrofitted on any desktop computer. All you need to do is install an eSata card (if you don't have an eSata port) and then purchase additional drives (SAME model number OR larger capacity as the current drive in the computer). Simply stick the current HD into the raid array FIRST, then put the new empty drives in SECOND, and it'll automatically mirror. Set the computer to boot from eSata and you're all set.
If you ever want to upgrade capacity, simply buy 3 larger drives and it will automatically mirror those drives to a larger capacity for you.
Or how about this:
I'm a developer on an enterprise application suite that exercises critical bugs in Java 6 Update 18-24 (we haven't yet tested 25 and 26). Oracle introduced a regression in 6u18 that they fixed in 6u21, but in 6u20 or 6u21 they introduced yet another regression. Both regressions cause a complete crash of the JDK that, on busy production systems, causes a complete crash, usually 1+ times a day.
Therefore, we can only recommend 6u17 as the stable version of the software, because 6u18+ isn't. I would strongly prefer that you not "beat the ever living shit out of me" for Oracle not being able to create a stable JVM for an enterprise product.
We'd get an Oracle support agreement to get these problems resolved, but you wouldn't BELIEVE how much money they want for such support. And even if we did pay Oracle a BOATLOAD of money, there's no guarantee that they'd even fix our issues--just that they would listen.
on busy production systems, causes a complete crash, usually 1+ times a day.
Java7 doesn't appear to be much better for stability. I just got this email today:
Hello Apache Lucene & Apache Solr users,
Hello users of other Java-based Apache projects,
Oracle released Java 7 today. Unfortunately it contains hotspot compiler
optimizations, which miscompile some loops. This can affect code of several
Apache projects. Sometimes JVMs only crash, but in several cases, results
calculated can be incorrect, leading to bugs in applications (see Hotspot
bugs 7070134 [1], 7044738 [2], 7068051 [3]).
Actually, it's IE2 that gets 93. IE1 fails to render the Acid tests and actually crashes on the 3rd acid test.
Of course, the editors couldn't bother to fact check the submission...
UPnP and NAT-PMP both workaround this limitation by automatically setting up the static port mappings. This has significant security implications, but these were created to solve problems just like these.
Skype (Version 5, at least) ships with UPnP support enabled, and it will automatically create port forwards for any home router that has UPnP enabled. Since most home routers have this feature enabled, many people will be able to route directly, without an intermediate server.
My $10 FM radio on my desk is not able to call Taiwan, and neither is my desk phone (no international service). However, my personal Skype account allowed me to call a prospective vendor for $.50 (15-30 minutes), which significantly increased my productivity at work. This is a single example of how my company's permissive (to its employees) security policy has increased my productivity.
Giving employees the flexibility to do their job with minimal or no red tape improves both employee satisfaction and productivity. Needing to use Skype is not something that is normal at my office, and if I had to file an IT support ticket just to use Skype, it would have taken multiple manhours to do what took a few minutes to do. Also, if I had to wait for that IT ticket to get completed, I could not have called the vendor when the account manager was at my desk. All of these things had tangible benefits to my company.
I understand security is a major concern. Allowing outbound access on all ports is VERY permissive. However, there are also costs to allowing employees to run other applications on your network. What many IT folk and upper managers don't consider is there are hidden costs in firewalling everything off. I deal with IT support and the most restricted and secured customer environments take FOREVER to get anything done, because it requires multiple people to sign off on and make any change.
At my office, I have full access to the internet (behind a packet inspection firewall) and full admin access to my box. This directly increases my productivity by allowing me to use the best tool for the job without getting approval from another person. When I need a tool to call Taiwan to ask a few quick questions about a prospective product that we may want to purchase, I can do that. I don't have to ask anyone, get permission from anyone, or wait for an IT guy to walk over and put something on my box.
You have to incrementally update through each version. If you have 8.04, you have to go to 8.10, then 9.04, then finally 9.10.
The incremental updates can be done through the install updates on your desktop. If you wait too long, you'll have to change your apt sources, so I'd upgrade sooner rather than later.
Yeah, I suppose I meant KiB, not technically KB. In general, which one KB refers to is unambiguous, and so I don't go through the trouble of looking up which specific term I need.:) If I ever write a hardware spec, though, I'll be sure to specify which one I mean!
Good questions, unfortunately I'm afraid I may not have most of the information you're looking for...
1. The answer to this would be that each manufacturer may or may not document this, at their whim. It would certainly be easy for them to do so, if they wanted. I don't know where any such documents are--I suspect that if Google doesn't know, you're kinda left in the dark. Also note that manufacturers may change these "implementation details" at any time, while not changing the model number (see also, WRT54G or almost any consumer-grade network product today). Many of these products have MAJOR internal details changed (flash memory, CPU speed, RAM, OS) without any major change in the model number. For instance a WRT54Gv1 contains twice the ram and flash and half the CPU of a WRT54Gv5, and it runs Linux instead of VxWorks. To test, I quickly googled the block size of the Intel SSD's and didn't see anything obvious in the first 30 seconds or so...
2. None that I know of. Doesn't mean one doesn't exist. However, the vendor specific firmware tools may give such info. For instance, I got some information when I flashed my Intel SSD with the later firmware. I don't remember if block size was included.
3. Absolutely. If you broke open the SSD and looked at the flash chips themselves, they would likely be printed with a model number. I'm 99% sure that Googling such a number would give you the exact flash manufacturer, block size, and a slew of other items about the chips. However, this would likely void warranty, so that may not recommended.
In short, unless you're a geek and want to physically play with the device, the implementation details (block size, etc) often don't matter as much as one would think. For instance, Intel combines writes together, so even small writes go much faster than they may otherwise go. This means that the block size on these devices matters MUCH less than other SSD's. That being said, the block size is special--if we knew what the size actually was, it would allow us to set the block size in our OS filesystem to match, which would theoretically give improved performance.
Also, if you're looking at a SSD purchase in the near future, I would avoid any product other than an Indilinx or Intel at the current time. Samsung SSD's are also decent, but I haven't found as much information about their design, specifically if their performance will degrade over time, and how much. I'm currently using a Generation 1 Intel SSD and it works very very well.
Slashdot Mirror
Tomato may also be affected. See my post here:
http://hardware.slashdot.org/comments.pl?sid=3381865&cid=42588997
We won't know until full disclosure.
While what you say is true, it's not just the WRT54GL that's likely affected. It's highly likely that WRT54G/WRT54GS v1-v4 are also affected.
The WRT54GL router uses nearly identical firmware to the v1, v2, v3, and v4 models of the WRT54G and WRT54GS router. In fact, the specs on a WRT54G/WRT54GS v4 are nearly identical to a WRT54GL--same chipset, RAM, and flash.
A lot of these were manufactured and sold, and I bet the vast majority are still running stock.
I love Tomato too--in fact, I use it at my house. However, Tomato was originally based off Stock Linksys, and might also be affected. Until full disclosure occurs, we'll not know for sure.
Wow, after reading the recall, this is SUPER stupid.
Personally, I don't think the government needs to be our nanny. But I at least understand the motivation of trying to keep things like these away from 2-3 year old because they like, yanno, eat them and it causes problems (even though there's only 2 documented cases).
The recall is this: Between March 2009 through March 2010, buckballs were sold labeled 13+. The federal law requires powerful magnets to be 14+. This was corrected in March 2010 when the magnets were labeled "keep away from all children".
So this recall protects all of those THIRTEEN year old children out there who might accidentally eat a buckyball because they didn't read the packaging. I'm sure that's, like, super common. Must. happen. a. lot.
How about an airave or a cellular repeater? Sounds like you'd be an ideal candidate for one...
http://www.wilsonelectronics.com/
Three times?
http://www.thewebsiteisdown.com/
Which vendors are those? I'd love to buy one of those for my next phone, but a quick Google didn't turn up anything...
Also, I know running a rooted phone is against my carrier's TOS (Sprint).
I use RAID 1 for backups all the time. I shutdown the system, pull a drive, swap it out, and then restart the system. While system is up, replace second drive with a new drive and reimage over.
Total downtime: 5 minutes or so. Degraded performance for a few hours during rebuild.
Basically, if you use RAID 1 like tapes, where each HD is a tape, it can be an extremely economical and reliable way to backup data. You know your data's being backed up because you're actually running off that data. You can't have a situation where the tape drive fails to write data and then suddenly when you have a disaster, you find out that you have 6 months of blank tapes (I've had that happen).
At $100 per tape (actually HD+case), it's very cheap per 2TB of backups.
This is absolutely true, but only on pressurized aircraft. It's the stress/release of the metal which causes metal fatigue and eventual failure of the airframe.
I expect (but don't know) that the 727 they used for the show had exceeded its "safe" pressurization cycles and was destined for the scrapyard anyway--so it only had scrap value anyway.
However, the DC3 that dblll mentioned is NOT a pressurized aircraft--so it need not worry about cycles at all. It's only about flight hours and wear and tear. As such, there's still many DC3's in operation today--some even commercially.
I'm sorry. He signed the bill into law. Had he truly been opposed to it, he could have vetoed it. Sure, his veto could have been overruled, but at least he'd be officially on the record that he didn't support it.
Also, Obama specifically requested the removal of language that said the NDAA would not apply to US citizens or lawful residents. See: http://www.youtube.com/watch?v=4DNDHbT44cY
I'm sorry--I think there's a big difference between "swiftboat veterans" for "truth" and "Obama supported the NDAA and specifically requested some of the draconian language in the bill"
NHTSA FARS data, 2002-2006: 27 legally blind pedestrians were killed by automobiles. 27/5 == 5.4 per year.
Who cares about facts, won't someone please think about the blind children?
So how exactly is a processor running a program to implement another instruction set architecture, with the main memory used by the implemented ISA being accessed by special operations, and with the program and its internal data existing in a separate block of memory, different from, say, a (vertical) microcode engine, running microcode to implement another instruction set architecture, with the main memory used by the implemented ISA being accessed by special microcode operations, and with the microprogram and its internal data existing in a separate block of memory?
Each would be granted a separate patent?
Thanks for actually posting facts. It's refreshing, for a change. You must be new here....
Anecdotal: I no longer fly due to TSA regulations. I either drive or take a train. Yes, it's tremendously inconvenient. Yes, I hate it. No, I don't think there's a lot of people like me...
...okay, I don't even know what to say to that. I have no idea what it's like on your planet, but around here we're only human. No wonder developers aren't up to your standards....
Totally agree. I was initially inclined to say (s)he's trolling, but (s)he's clearly quite learned in computers. Maybe (s)he expects that all people are just that smart... Expecting that people get parallel programs right on the first try, given their complexity is not reasonable, at least where I work (myself included). In fact, I was just working with a developer today to fix a reader/writer issue triggered by parallelism both in code and in writing to the DB. We had to sit down and think out the use cases for about an hour before we had a good working solution.
:)
Parallel programming is hard. It's necessary, but hard, at least to us "normal" people.
The same ones that invented the internet, lasers, microchips, GPSs, and microwave ovens, went to the moon, built the largest highway system in the world, created our clean water infrastructure, and electrified most of a continent.
How about the state governments do those things? Pretty sure the state governments can figure out how to build roads, clean water, and manage electrical grids. The only item on the list that I see as problematic would be "went to the moon".
I'm very supportive of minimizing the federal government. That doesn't mean I don't want government services at all--we have state and local governments, after all...
Perhaps if they get enough negative feedback, they'll drop the threatening postures and lawsuits...
http://www.firststatesuper.com.au/EmailEnquiries
I personally use a RAID enclosure from Raidon with two drives in a mirrored configuration. Something like the GR3630-2S-SB2:
http://www.newegg.com/Product/Product.aspx?Item=N82E16816142002
(I've used Accordance RAID systems in the past, before my current Raidon).
Then I buy an additional HD tray and rotate the 3 drives occasionally to an offsite backup. Basically the cost is about $345 total, $75*3 drives+$120 for the unit, and I get 3 copies of the data. Additional offsite backups run $105 ($75 drive+$30 tray).
I've found this the easiest to understand (for end users) and a very cheap and reliable storage method. I've had the nightmare of backing up to 8 different tapes, only to go to restore and find out that 6 tapes are empty, 1 is bad, and 1 has data from 6 months back. Because you're actually running your OS off of the drives, you don't get into that situation. The main key is to make sure that you pull drives when the system is off (so that all files are written to disk), and that you insert drives when the system is ON (so that the raid array knows the drive is replaced and doesn't corrupt your data). NEVER INSERT A DRIVE WHILE THE SYSTEM IS OFF.
I have files from over 15 years ago on my main desktop computer, and the only files I'm missing are from my original 286 before I started this backup method.
This solution can be retrofitted on any desktop computer. All you need to do is install an eSata card (if you don't have an eSata port) and then purchase additional drives (SAME model number OR larger capacity as the current drive in the computer). Simply stick the current HD into the raid array FIRST, then put the new empty drives in SECOND, and it'll automatically mirror. Set the computer to boot from eSata and you're all set.
If you ever want to upgrade capacity, simply buy 3 larger drives and it will automatically mirror those drives to a larger capacity for you.
Or how about this: I'm a developer on an enterprise application suite that exercises critical bugs in Java 6 Update 18-24 (we haven't yet tested 25 and 26). Oracle introduced a regression in 6u18 that they fixed in 6u21, but in 6u20 or 6u21 they introduced yet another regression. Both regressions cause a complete crash of the JDK that, on busy production systems, causes a complete crash, usually 1+ times a day.
Therefore, we can only recommend 6u17 as the stable version of the software, because 6u18+ isn't. I would strongly prefer that you not "beat the ever living shit out of me" for Oracle not being able to create a stable JVM for an enterprise product.
We'd get an Oracle support agreement to get these problems resolved, but you wouldn't BELIEVE how much money they want for such support. And even if we did pay Oracle a BOATLOAD of money, there's no guarantee that they'd even fix our issues--just that they would listen. on busy production systems, causes a complete crash, usually 1+ times a day.
Java7 doesn't appear to be much better for stability. I just got this email today:
Hello Apache Lucene & Apache Solr users, Hello users of other Java-based Apache projects,
Oracle released Java 7 today. Unfortunately it contains hotspot compiler optimizations, which miscompile some loops. This can affect code of several Apache projects. Sometimes JVMs only crash, but in several cases, results calculated can be incorrect, leading to bugs in applications (see Hotspot bugs 7070134 [1], 7044738 [2], 7068051 [3]).
Actually, it's IE2 that gets 93. IE1 fails to render the Acid tests and actually crashes on the 3rd acid test. Of course, the editors couldn't bother to fact check the submission...
UPnP and NAT-PMP both workaround this limitation by automatically setting up the static port mappings. This has significant security implications, but these were created to solve problems just like these.
Skype (Version 5, at least) ships with UPnP support enabled, and it will automatically create port forwards for any home router that has UPnP enabled. Since most home routers have this feature enabled, many people will be able to route directly, without an intermediate server.
My $10 FM radio on my desk is not able to call Taiwan, and neither is my desk phone (no international service). However, my personal Skype account allowed me to call a prospective vendor for $.50 (15-30 minutes), which significantly increased my productivity at work. This is a single example of how my company's permissive (to its employees) security policy has increased my productivity.
Giving employees the flexibility to do their job with minimal or no red tape improves both employee satisfaction and productivity. Needing to use Skype is not something that is normal at my office, and if I had to file an IT support ticket just to use Skype, it would have taken multiple manhours to do what took a few minutes to do. Also, if I had to wait for that IT ticket to get completed, I could not have called the vendor when the account manager was at my desk. All of these things had tangible benefits to my company.
I understand security is a major concern. Allowing outbound access on all ports is VERY permissive. However, there are also costs to allowing employees to run other applications on your network. What many IT folk and upper managers don't consider is there are hidden costs in firewalling everything off. I deal with IT support and the most restricted and secured customer environments take FOREVER to get anything done, because it requires multiple people to sign off on and make any change.
At my office, I have full access to the internet (behind a packet inspection firewall) and full admin access to my box. This directly increases my productivity by allowing me to use the best tool for the job without getting approval from another person. When I need a tool to call Taiwan to ask a few quick questions about a prospective product that we may want to purchase, I can do that. I don't have to ask anyone, get permission from anyone, or wait for an IT guy to walk over and put something on my box.
Lawyers start out as children. Kill them when they're 4...
You have to incrementally update through each version. If you have 8.04, you have to go to 8.10, then 9.04, then finally 9.10.
The incremental updates can be done through the install updates on your desktop. If you wait too long, you'll have to change your apt sources, so I'd upgrade sooner rather than later.
This is a 2*40 mph (80 mph) collision. This is roadway speeds, not highway.
Highway would have been WAY worse.
Shic:
Yeah, I suppose I meant KiB, not technically KB. In general, which one KB refers to is unambiguous, and so I don't go through the trouble of looking up which specific term I need. :) If I ever write a hardware spec, though, I'll be sure to specify which one I mean!
Good questions, unfortunately I'm afraid I may not have most of the information you're looking for...
1. The answer to this would be that each manufacturer may or may not document this, at their whim. It would certainly be easy for them to do so, if they wanted. I don't know where any such documents are--I suspect that if Google doesn't know, you're kinda left in the dark. Also note that manufacturers may change these "implementation details" at any time, while not changing the model number (see also, WRT54G or almost any consumer-grade network product today). Many of these products have MAJOR internal details changed (flash memory, CPU speed, RAM, OS) without any major change in the model number. For instance a WRT54Gv1 contains twice the ram and flash and half the CPU of a WRT54Gv5, and it runs Linux instead of VxWorks. To test, I quickly googled the block size of the Intel SSD's and didn't see anything obvious in the first 30 seconds or so...
2. None that I know of. Doesn't mean one doesn't exist. However, the vendor specific firmware tools may give such info. For instance, I got some information when I flashed my Intel SSD with the later firmware. I don't remember if block size was included.
3. Absolutely. If you broke open the SSD and looked at the flash chips themselves, they would likely be printed with a model number. I'm 99% sure that Googling such a number would give you the exact flash manufacturer, block size, and a slew of other items about the chips. However, this would likely void warranty, so that may not recommended.
In short, unless you're a geek and want to physically play with the device, the implementation details (block size, etc) often don't matter as much as one would think. For instance, Intel combines writes together, so even small writes go much faster than they may otherwise go. This means that the block size on these devices matters MUCH less than other SSD's. That being said, the block size is special--if we knew what the size actually was, it would allow us to set the block size in our OS filesystem to match, which would theoretically give improved performance.
Also, if you're looking at a SSD purchase in the near future, I would avoid any product other than an Indilinx or Intel at the current time. Samsung SSD's are also decent, but I haven't found as much information about their design, specifically if their performance will degrade over time, and how much. I'm currently using a Generation 1 Intel SSD and it works very very well.
HTH