little b is bits, big B is bytes - the bigger one is the bigger unit (100Mbps is about 12MB/s). To get your 700MB file, you'll be pulling down 5600 bits plus overhead, and it'll take a little over 1 minute on a fully saturated link. Real world, you'll be waiting a little longer.
Someone else already mentioned that Firefox bugs actually get *fixed*, and often don't have exploits available until after they're disclosed.
This bug is with a required piece of system software that you can't turn off, *and* it's not fixed yet, *and* there is a working exploit available. If you can think of other similar situations that aren't reported, please, feel free to submit them. Otherwise, your apples don't belong in this orange tree.
Random people who post ads on Craigs list aren't "friends and family". If one recognizes that "most of society" thinks their practices are weird, one should realize that people would be likely to post prank ads on Craigslist, and therefore take *a litte* precaution before sharing that information. Simply telling everyone who asks (and providing unsolicited photographs, etc) implies to me that it's not such a big secret...
I recently stopped bouncing phishing eBay messages to eBay, because their "spoof" address now auto-responds, to which you have to reply in order for the report to count. I don't want *more* spam for pointing out known scam attempts, and it's not worth the hassle to help them out. I guess someone else will forward them along.
On top of that, your balance is reported as the average daily balance. Even if you pay it off in full each month, so you have no "old" balance, if you're spending close to the card's max (because you have a reward card, say, and buy everything with the credit card instead of carrying around loads of cash - because your card gives you an automatic warranty as well as 1% back), your average balance looks like you're "maxed out" all the time. After that realization earlier this year, I've started paying my card off every 3-5 days, since it's easy to just click a couple of times to pay electronically out of a checking account set up just for paying off the credit card. With no other credit-affecting changes, that's bumped my score up several points.
I hate credit reporting algorithms, but more than that, I hate people who look at those reports as anything more than a rough estimate tool.
Presumably most people know )and trust) their wife a little better than some random person who posted an ad on cragislist?
The point is that these people sent their information to someone they don't know *at all*, which is roughly similar to just posting it on their web page for all to see.
Personally, I wonder why the're doing things that they're emberrased by to begin with. Seems like it might be useful to seek phsychiatric help such that they can either accept themselves for what they are, or change to reflect what they want to be. It can't be fun living a life where you're perpetually ashamed of yourself...
Well, I'm not sure why it matters to the company who provides the support. That $40K/year in RHEL licensing would get the same thing if they were using, say, CentOS. Why not do that and give *me* the $40K on top of my normal salary? It's my ass on the line either way, RedHat isn't gonna be out of a job when RHEL fails to work at my employer...
Essentially, any money the company pays out on things they don't need is money that could either be going to the employees or going to making the business more stable (savings to weather hard times, or investments in better hardware, etc). It's like anyting else - stupid to spend money in places where it's not useful.
Why do so many people say that Ubuntu's not acceptable to enterprise because it doesn't have support, there's no one to blame, etc? Has no one ever gone to ubuntu.com and seen that big friggin' link at the top of the front page, which says "support"?
I almost bought an '01 Impala before I bought my last car - a Pontiac Grand Am GT SC/T with the "ram air" 3800. After less than six months when I decided to get rid of the G/A (for various dissatisfaction-related reasons), I almost bought a black '05 Impala SS (the supercharged 3800 is less than boring, and they look especially cool in black when the windows are tinted) before I decided to go with the Grand Marquis. I decided against the first Impala because of the boring factor, and the second time I decided against it because the speedometer broke (started reading 45MPH too high) while testdriving. And the Grand Marquis just has a better build quality "feel" to it, though I know the W-body Impalas actually are quite reliable cars as you have found. I can't understand how the cops can actually use them for "pursuit" vehicles, though, after I test drove a few. The Ford panther platform cars aren't rockets, but they sure *feel* a lot stronger...
The Grand Marquis/Crown Victoria don't hold their value very well for the first year or two, so even the top-of-the-line "Ultimate Edition" like the one I got is relatively affordable when a couple years old. I think I gave like $17K with 27,000 miles on the clock, which I think is pretty good for a car that was about $38K new two years ago. After adding a Dynomax UltraFlow muffler, it sounds like a big V8 RWD car should (nearly silent under normal usage, adequately angry when pressed). And I don't expect to trade this one real soon, so I'm not real concerned about it losing even more value. Getting a "certified used" Mercury puts it under warranty until just about 100K miles and something like 7 years from now, too. I still really like both of my LT1 Caprices, but this Ford product really is growing on me.:)
The, IMHO, cool part about the LT1s is that they're built like tanks. With little more than regular fluid changes and general care, they'll run for 300K miles. The 4L60-E transmission, however, will usually need rebuilt somewhere in the 150K mile range if the fluid wasn't changed every 50K or so. And there are quire a few options, from retired cop cars that run around $1.5K up to the Buick Roadmaster and Cadillac Fleetwood (same platform - all '94-'96 Fleetwoods and wagons (caprice + roadmaster) are LT1s) which will run somewhere in the $5-$7K range for a fairly luxurious car - and you can still spend $10K or more on an impala of the same vintage, which boils down to a cop car with leather seats. Make sure it's an LT-1 and not the smaller version, though. The emissions sticker under the hood should say 5.7 liters, and Caprices (as well as the other sedans, IIRC) will have an option code sticker on the bottom of the trunk lid. It should have LT1 in the list. And note that the civilian Caprice didn't get all of the same goodies as the other models - specifically, the dual electric fans make a little difference in mileage potential, and the aux transmission cooler seems to really help reliability.
I'm partial to the '95s, since there were some minor reliability changes between '94 and '95, and then in '96 they switched to OBD-II. So you can get a '96 scanned for check engine light at Auto Zone for free, but it's a lot cheaper to get software to scan and program an OBD-I '94-'95 yourself than it is for the '96.
My '95 Chevy Caprice wagon averaged 22 MPG with 100K miles all the way up to when the transmission burned up completely at 120K (original owner apparently never changed the fluid). Several owners I know with Roadmasters/Caprices get 25+ MPG. You've gotta get the '94-'96 models with the LT-1, though, not the '91-'93 models with the TBI 350.
Just FYI. I also know people with wagons that run 13s in the 1/4 and still get nearly 20 MPG...
Did I mention that one of the places I worked with RHEL was Intel, where we used it on hardware that wasn't on the market yet? I personally dislike RedHat lagely *because* of the problems getting new drivers working - not because they're great and responsive.
That said, sure, RedHat's been good to the community in some places. I'm just saying that the lage shops I've worked in who use RedHat are using it because management wants something with the illusion of support, not because anyone has any intention of actually *using* the support. We used it at Intel (in my group, anyway) because it's what "everyone else" uses and we wanted to sell hardware, not because it's better. The two responses to my post are the only people who I've ever heard imply that they might have actually used the support. Well, I tried to use it once, but no one was able to help me figure out what happened to the remainder of the money for the full year subscriptions I/we paid for the desktop version, back when RedHat killed support part way through the year in favor of keeping people's money and not fulfilling their promised level of support.
The only reason RedHat sells anything to "corporate" America is because they *offer* support. It's not because anyone actually uses it. The people making decisions at large companies (such as the Fortune 25 company I work for now, or the Fortune 50 semiconductor company I worked for before) want another company that says they'll support the product. Perhaps even one that will have meetings with them. They don't care if the support is ever actually used, and the admins actually working with the software know that nearly anything they try to *do* with the software will invalidate the support, but no one cares. It's just that empty promise and a big ad in a trade magazine. Somehow I doubt that Oracle will make things any better, except for supporting a couple more specific configurations/situations that probably no one will actually find themselves in.:)
1) I just looked up -n, not -l, and a real geek doesn't need line numbers most of the time... 2) timing those commands, both took about.08 seconds of real execution time (with plain grep getting.076 once), so I saved more time by not typing the extra char in "fgrep".;)
The lesson isn't that GnuCash (and anything else) has problems, it's that you should not be running critical applications on Mandrake. Your finances are critical applications. Mandrake is for browing the web and telling people that you "run Linux". Redhat, SuSE, Debian, maybe Ubuntu - things that a company might run a server on. Those are for important things.
Disclaimer: I've used GnuCash for many years, and have never had an upgrade problem. I also don't use Mandrake. Coincidence? Perhaps...
You should contact your city engineer person and ask why they placed stoplights exactly 1/4 mile apart on a country road (near a grumpy old man, apparently). At that point they may as well install timing devices and a bleach box, because it's a drag strip.
I run slackware. I have, in fact, since 1994. It was my first distro, and it still runs my firewall + mail relay. But thanks for posting the tips - someone might benefit from them.:)
Anyway, the point was not that it's impossible to figure the scripts out - it was that the scripts are non intuitive. Look at "rc.M" and tell me that it's obvious how that filename's connected to lpr. It's more difficult to do simple tasks with BSD-style init than it needs to be. Envision, instead, a SysV-style init setup:
Lpr (actually lpd) starts from the lpd (or possibly "cups") init script. chmod -x to disable it in all runlevels, or remove the symlink in the runlevel you want it disabled in. No dicking around with grep to find which file it's in, finding the appropriate line, and editing the script. And it's possible to easily disable it in just one runlevel without affecting it in others.
To determine when nfs starts relative to ssh, just run ls on the directory corresponding to the runlevel you care about - the one that's listed first starts first.
There's a reason that almost every other distro has switched to a System V style init setup. It's easier to maintain and configure (both for people and automated scripts, which I'd rather see running "chmod -x" than inserting/removing comment chars), and it's more intuitive in almost every way. That, and it's a lot easier to have different, easily readable setups for differnet runlevels. Those "rc.N" text files aren't as easy to work with as a directory...
I'll just ignore the semi-screwball setup Gentoo uses, where scripts have things like "depend" and "provides" - almost like a RPM.spec for init...:) It works well for scripted startup configuration, though.
PS, on my slackware box, which has been upgraded a few times, lpr *used* to be started in rc.inet2. So did nfs and ssh - inet2 was network services, and inet1 was interface setup.
I'm pretty sure that any time someone says that something tastes good or bad, it's implied that it's only their opinion. Other peole are welcome to drink whatever crap they like - but on-topic, most seem to agree that coffee and alcohol are both very often accuired tastes...:)
I especially like the BSD style init system, it just makes sense
Yeah, having a bunch of scripts that call each other in an unclear order (unless you read all of the scripts) is a whole lot easier to understand than one directory - named after the current runlevel - full of scripts, all of which are run in the same order that ls shows them to you. "Hmmm, does lpr launch from rc.net1, rc.net2, or something else? When does nfs start relative to ssh?" The only "simple" thing about BSD-style init is that there's no scary case statement in the init scripts to handle the start and stop arguments, and no worrisome symlinks. I'd posit that, if someone doesn't know enough shell scripting to read (and maybe write) a case statement or make a symlink, perhaps they shouldn't be messing with the machine's init scripts. Just edit/etc/rc.local - both SysV and BSD use that one.
Yeah, buying a car with satelite radio functionality or On Star, and then having to pay an extra fee to use it would be painful. No one would do that./looks around and notices that most of GM's vehicles come with OnStar enabled for just one year, but that the capability remains unusably in the car after the subscription runs out (for example)
Slashdot Mirror
little b is bits, big B is bytes - the bigger one is the bigger unit (100Mbps is about 12MB/s). To get your 700MB file, you'll be pulling down 5600 bits plus overhead, and it'll take a little over 1 minute on a fully saturated link. Real world, you'll be waiting a little longer.
And then they should have to change it every 90 days, because some dumbass read a document entitled "best practices" about that kind of thing.
Someone else already mentioned that Firefox bugs actually get *fixed*, and often don't have exploits available until after they're disclosed.
This bug is with a required piece of system software that you can't turn off, *and* it's not fixed yet, *and* there is a working exploit available. If you can think of other similar situations that aren't reported, please, feel free to submit them. Otherwise, your apples don't belong in this orange tree.
Random people who post ads on Craigs list aren't "friends and family". If one recognizes that "most of society" thinks their practices are weird, one should realize that people would be likely to post prank ads on Craigslist, and therefore take *a litte* precaution before sharing that information. Simply telling everyone who asks (and providing unsolicited photographs, etc) implies to me that it's not such a big secret...
I recently stopped bouncing phishing eBay messages to eBay, because their "spoof" address now auto-responds, to which you have to reply in order for the report to count. I don't want *more* spam for pointing out known scam attempts, and it's not worth the hassle to help them out. I guess someone else will forward them along.
On top of that, your balance is reported as the average daily balance. Even if you pay it off in full each month, so you have no "old" balance, if you're spending close to the card's max (because you have a reward card, say, and buy everything with the credit card instead of carrying around loads of cash - because your card gives you an automatic warranty as well as 1% back), your average balance looks like you're "maxed out" all the time. After that realization earlier this year, I've started paying my card off every 3-5 days, since it's easy to just click a couple of times to pay electronically out of a checking account set up just for paying off the credit card. With no other credit-affecting changes, that's bumped my score up several points.
I hate credit reporting algorithms, but more than that, I hate people who look at those reports as anything more than a rough estimate tool.
Presumably most people know )and trust) their wife a little better than some random person who posted an ad on cragislist?
The point is that these people sent their information to someone they don't know *at all*, which is roughly similar to just posting it on their web page for all to see.
Personally, I wonder why the're doing things that they're emberrased by to begin with. Seems like it might be useful to seek phsychiatric help such that they can either accept themselves for what they are, or change to reflect what they want to be. It can't be fun living a life where you're perpetually ashamed of yourself...
Stupid pointy-hairs and their desire for data visualization.
Maybe in 5 years, we will be able to buy a diesel powered Camero for $25k. The prototypes look pretty good...
Yeah, I'm looking forward to when Chavrolat brings back the Camero.
Well, I'm not sure why it matters to the company who provides the support. That $40K/year in RHEL licensing would get the same thing if they were using, say, CentOS. Why not do that and give *me* the $40K on top of my normal salary? It's my ass on the line either way, RedHat isn't gonna be out of a job when RHEL fails to work at my employer...
Essentially, any money the company pays out on things they don't need is money that could either be going to the employees or going to making the business more stable (savings to weather hard times, or investments in better hardware, etc). It's like anyting else - stupid to spend money in places where it's not useful.
Why do so many people say that Ubuntu's not acceptable to enterprise because it doesn't have support, there's no one to blame, etc? Has no one ever gone to ubuntu.com and seen that big friggin' link at the top of the front page, which says "support"?
http://www.ubuntu.com/support/paid
Alternatively, has anyone ever actually used RedHat support? *I* wasn't impressed...
I almost bought an '01 Impala before I bought my last car - a Pontiac Grand Am GT SC/T with the "ram air" 3800. After less than six months when I decided to get rid of the G/A (for various dissatisfaction-related reasons), I almost bought a black '05 Impala SS (the supercharged 3800 is less than boring, and they look especially cool in black when the windows are tinted) before I decided to go with the Grand Marquis. I decided against the first Impala because of the boring factor, and the second time I decided against it because the speedometer broke (started reading 45MPH too high) while testdriving. And the Grand Marquis just has a better build quality "feel" to it, though I know the W-body Impalas actually are quite reliable cars as you have found. I can't understand how the cops can actually use them for "pursuit" vehicles, though, after I test drove a few. The Ford panther platform cars aren't rockets, but they sure *feel* a lot stronger...
:)
The Grand Marquis/Crown Victoria don't hold their value very well for the first year or two, so even the top-of-the-line "Ultimate Edition" like the one I got is relatively affordable when a couple years old. I think I gave like $17K with 27,000 miles on the clock, which I think is pretty good for a car that was about $38K new two years ago. After adding a Dynomax UltraFlow muffler, it sounds like a big V8 RWD car should (nearly silent under normal usage, adequately angry when pressed). And I don't expect to trade this one real soon, so I'm not real concerned about it losing even more value. Getting a "certified used" Mercury puts it under warranty until just about 100K miles and something like 7 years from now, too. I still really like both of my LT1 Caprices, but this Ford product really is growing on me.
Indeed. The 3800 V6 was one of the best things GM ever did. :)
The, IMHO, cool part about the LT1s is that they're built like tanks. With little more than regular fluid changes and general care, they'll run for 300K miles. The 4L60-E transmission, however, will usually need rebuilt somewhere in the 150K mile range if the fluid wasn't changed every 50K or so. And there are quire a few options, from retired cop cars that run around $1.5K up to the Buick Roadmaster and Cadillac Fleetwood (same platform - all '94-'96 Fleetwoods and wagons (caprice + roadmaster) are LT1s) which will run somewhere in the $5-$7K range for a fairly luxurious car - and you can still spend $10K or more on an impala of the same vintage, which boils down to a cop car with leather seats. Make sure it's an LT-1 and not the smaller version, though. The emissions sticker under the hood should say 5.7 liters, and Caprices (as well as the other sedans, IIRC) will have an option code sticker on the bottom of the trunk lid. It should have LT1 in the list. And note that the civilian Caprice didn't get all of the same goodies as the other models - specifically, the dual electric fans make a little difference in mileage potential, and the aux transmission cooler seems to really help reliability.
I'm partial to the '95s, since there were some minor reliability changes between '94 and '95, and then in '96 they switched to OBD-II. So you can get a '96 scanned for check engine light at Auto Zone for free, but it's a lot cheaper to get software to scan and program an OBD-I '94-'95 yourself than it is for the '96.
My '95 Chevy Caprice wagon averaged 22 MPG with 100K miles all the way up to when the transmission burned up completely at 120K (original owner apparently never changed the fluid). Several owners I know with Roadmasters/Caprices get 25+ MPG. You've gotta get the '94-'96 models with the LT-1, though, not the '91-'93 models with the TBI 350.
Just FYI. I also know people with wagons that run 13s in the 1/4 and still get nearly 20 MPG...
Integrating kernel drivers for new hardware
Did I mention that one of the places I worked with RHEL was Intel, where we used it on hardware that wasn't on the market yet? I personally dislike RedHat lagely *because* of the problems getting new drivers working - not because they're great and responsive.
That said, sure, RedHat's been good to the community in some places. I'm just saying that the lage shops I've worked in who use RedHat are using it because management wants something with the illusion of support, not because anyone has any intention of actually *using* the support. We used it at Intel (in my group, anyway) because it's what "everyone else" uses and we wanted to sell hardware, not because it's better. The two responses to my post are the only people who I've ever heard imply that they might have actually used the support. Well, I tried to use it once, but no one was able to help me figure out what happened to the remainder of the money for the full year subscriptions I/we paid for the desktop version, back when RedHat killed support part way through the year in favor of keeping people's money and not fulfilling their promised level of support.
The only reason RedHat sells anything to "corporate" America is because they *offer* support. It's not because anyone actually uses it. The people making decisions at large companies (such as the Fortune 25 company I work for now, or the Fortune 50 semiconductor company I worked for before) want another company that says they'll support the product. Perhaps even one that will have meetings with them. They don't care if the support is ever actually used, and the admins actually working with the software know that nearly anything they try to *do* with the software will invalidate the support, but no one cares. It's just that empty promise and a big ad in a trade magazine. Somehow I doubt that Oracle will make things any better, except for supporting a couple more specific configurations/situations that probably no one will actually find themselves in. :)
1) I just looked up -n, not -l, and a real geek doesn't need line numbers most of the time... .08 seconds of real execution time (with plain grep getting .076 once), so I saved more time by not typing the extra char in "fgrep". ;)
2) timing those commands, both took about
The lesson isn't that GnuCash (and anything else) has problems, it's that you should not be running critical applications on Mandrake. Your finances are critical applications. Mandrake is for browing the web and telling people that you "run Linux". Redhat, SuSE, Debian, maybe Ubuntu - things that a company might run a server on. Those are for important things.
Disclaimer: I've used GnuCash for many years, and have never had an upgrade problem. I also don't use Mandrake. Coincidence? Perhaps...
You should contact your city engineer person and ask why they placed stoplights exactly 1/4 mile apart on a country road (near a grumpy old man, apparently). At that point they may as well install timing devices and a bleach box, because it's a drag strip.
Anyway, the point was not that it's impossible to figure the scripts out - it was that the scripts are non intuitive. Look at "rc.M" and tell me that it's obvious how that filename's connected to lpr. It's more difficult to do simple tasks with BSD-style init than it needs to be. Envision, instead, a SysV-style init setup:
Lpr (actually lpd) starts from the lpd (or possibly "cups") init script. chmod -x to disable it in all runlevels, or remove the symlink in the runlevel you want it disabled in. No dicking around with grep to find which file it's in, finding the appropriate line, and editing the script. And it's possible to easily disable it in just one runlevel without affecting it in others.
To determine when nfs starts relative to ssh, just run ls on the directory corresponding to the runlevel you care about - the one that's listed first starts first.
There's a reason that almost every other distro has switched to a System V style init setup. It's easier to maintain and configure (both for people and automated scripts, which I'd rather see running "chmod -x" than inserting/removing comment chars), and it's more intuitive in almost every way. That, and it's a lot easier to have different, easily readable setups for differnet runlevels. Those "rc.N" text files aren't as easy to work with as a directory...
I'll just ignore the semi-screwball setup Gentoo uses, where scripts have things like "depend" and "provides" - almost like a RPM
PS, on my slackware box, which has been upgraded a few times, lpr *used* to be started in rc.inet2. So did nfs and ssh - inet2 was network services, and inet1 was interface setup.PPS, admit it - you had to look up "-n" in the grep man page, too (or you cheated and used a text editor to get the line numbers)...
I'm pretty sure that any time someone says that something tastes good or bad, it's implied that it's only their opinion. Other peole are welcome to drink whatever crap they like - but on-topic, most seem to agree that coffee and alcohol are both very often accuired tastes... :)
I especially like the BSD style init system, it just makes sense
/etc/rc.local - both SysV and BSD use that one.
Yeah, having a bunch of scripts that call each other in an unclear order (unless you read all of the scripts) is a whole lot easier to understand than one directory - named after the current runlevel - full of scripts, all of which are run in the same order that ls shows them to you. "Hmmm, does lpr launch from rc.net1, rc.net2, or something else? When does nfs start relative to ssh?" The only "simple" thing about BSD-style init is that there's no scary case statement in the init scripts to handle the start and stop arguments, and no worrisome symlinks. I'd posit that, if someone doesn't know enough shell scripting to read (and maybe write) a case statement or make a symlink, perhaps they shouldn't be messing with the machine's init scripts. Just edit
It works just fine. /drinks neither alcohol nor coffee, because they both taste awful
Yeah, buying a car with satelite radio functionality or On Star, and then having to pay an extra fee to use it would be painful. No one would do that. /looks around and notices that most of GM's vehicles come with OnStar enabled for just one year, but that the capability remains unusably in the car after the subscription runs out (for example)