I think we need to call in Al Gore. Maybe one day the global warming alarmists and hoaxsters will realize that change is a *natural* thing in this universe whether caused by inanimate or animate forces.
Storms come and go. Icecaps expand and shrink. Glaciers advance and recede. Species thrive and decline.
Get over it.
Indeed, the one difference between animate and inanimate forces is that inanimate generated change is usually random in its effect while the net effect caused by animate generated change tends to be for the overall net better effect of humanity (not every aspect is positive, not every individual benefits equally and not every day is progress but the overall level of societal wealth, comfort, and knowledge tends to move upwards over time).
Well given that there have been reports of several hundred such crashes, I guess it can't be a coincidence unless there are a billion or so Linux servers;)
I only said that the last log was 20 seconds before midnight and that based on my clock display, it crashed no later than 00:00:00. I would bet any amount of money that the actual crash was at the 00:00:00 GMT changeover.
But there have now been reports (just adding up the comments posted on slashdot and emails to me) of hundreds of machines going down at precisely 00:00:00 GMT (across multiple timezones).
That combined set of data points plus the obvious potential issue of a leap second being introduces at that precise time would seem to make your coincidence theory astronomically unlikely.
As the OP and having read about several hundred other stable machines in multiple timezones crashing at that precise second, I do not believe that is a coincidence.
Several people have reported 10 or more machines crashing at precisely 00:00:00 GMT -- it is so unlikely that hundreds of stable servers crashed precisely at this moment, that I would bet anything that this is NOT a coincidence.
Plus, the last time my machine encountered a leap second, it was running FC1 with a very different kernel so that does not prove anything.
Except that based on the responders that there were hundreds of such crashes at exactly that time (GMT) -- including people who had 10 or more servers crash at exactly that moment.
Sorry - I was trying to keep the post short.
Not a hardware failure in that it booted right up when I noticed it the next morning.
No kernel panics or any other evidence of why it crashed.
I agree - but seems like an awfully unlikely coincidence.
By my math, chance of a crash in any given 20 second period is just under 1 in 5 million:
(2 crashes in 2237 days)/(2237*24*60*60 seconds) * (20 second interval).
Seems suspicious to me...
I rebooted it and it went just fine.
I looked at the logs and saw now errors. Last entry was in/var/log/maillog at 18:59:40 (not an error).
So, not sure how to figure it out - tempted to try to replicate though by setting time back to 18:59 on 12/31/08 (and shutting off ntpd)
Froze - couldn't ping or ssh or get console response.
I know the time cuz last maillog entry was 18:59:40 and the clock (on my emacs session) said 18:59 at time of crash.
Hardware is: ASUS P4P
Rebooted without ever but required me to manually poweroff
What are you smoking? ### Setting up the guide subscription, and renewing it every few months (I'm told),
---> [Setup] Go to url, register, select zipcode, cable/satellite/OTA provider, check stations you want [5minutes?]
[Renewal] Every 3 months, you get a renewal email, click link, click "Extend my subscription", done [30 seconds]
Seems to me a lot easier, cheaper and faster than paying $12.95 each month (even with automatic payments)
### Recompile the kernel for a new feature
---> Haven't had to recompile a kernel since Slackware 1.0 or maybe Redhat 3.0.0.1 - more than 10 years ago
### need to hang on while the SQL database cleans itself
---> WTF are you talking about? This is utter nonsense. Sort of like saying, well you have to wait for the random logic turbolyzer to frobnicate the kryptonite-shielded holographic disk.
About the only valid point is your initial comment that setting up remotes can require tinkering but even this has gotten much easier with lirc packages.
Now I admit that MythTV may not be for anybody, but it is downright funny to hear Slashdot readers making up stuff like you do about the complexity. Sounds like you should be spending your time reading "computers for dummy" books rather than hanging out here on Slashdot.
The ease and difficulty of setting up MythTV depends on what approach you take -- that is what open software is all about - the user gets to control how they install, modify, and use the software.
Here are some options ranging from hard to easy 1. [Hardest] Download bleeding edge development releases, compile from source, patch/modify as needed, download & compile dependency programs too as needed, install on multiple backends and frontends, customize every possible setting, etc [Install & customization time = never done:)]
2. Locate and dowload pre-compiled packages, install them along with dependencies, customize settings as needed [Install & customization time = 1/2 day]
3. Use a package manager such as yum under Fedora and point it to the atRPMS repository -- all necessary packages and dependencies will be installed automatically. Run mythtvsetup and supply a few basic answers, accept defaults on everything else. Do minimal if any customization using mythfrontend [Install & customization time = couple of hours max]
4. [Easiest] Install a myth-enabled version of Linux such as Mythdora with Myth pre-loaded and configured as a default app. No extra setup or customization. Just runs [Install & customization time = install length of the custom distro, maybe 1 hour]
Option #4 is the analog of installing MCE which is the only option for Windows Media Center. The only difference is that even with option #4, under Linux, you can still later come back and customize MythTV as much as you want (even to the extent of upgrading to bleeding edge sources and toggling all the customization switches). In contrast, with Windows all you get is the base Media Center and maybe the ability to change a little eye candy. And if the conspiracy guys at Micro$oft-Media companies use DRM and other "evil" things to hobble your PVR, well you are then stuck.
So before, you gripe about MythTV, please make sure you compare apples-to-apples -- and in that context, MythTV is just as easy while having more power, customization, and DRM-freedom.
> As for succinctness and structure, I have found, over three years of being the only person in charge of a system used by most of > management (i.e. above me), that the only way to get anything accross is to make an email a maximum of three paragraphs, each > containing a maximum of two sentences. And anything that is mildly complicated should be in a bulleted list. I think that's > ridiculous. Basically, most people (especially when reading stuff at work) have something like an 8th grade attention span and > literacy level. So you have to be like a newspaper, write for 13 year olds.
I think that the way you describe above is the CORRECT way to write a business email. It has nothing to do with attention span or education level. Rather, senior managers with dozens or even hundreds of indirect reports tend to get hundreds of emails a day. They are looking for the 'snapshot' picture to identify the key details or issues that they need to know and address. Anything more and you lose the wheat with the chaff and they end up spending most of the day on emails rather than on managing the business.
I find in general that email is not a good tool for very detailed communication -- for that either face-to-face or more formal presentations (like Powerpoint or even formatted text reports) tend to work better. Now not everyone will agree and some may prefer to communicate via long emails, but conversely, I don't think you can expect colleagues to read your every email like a Bible scholar parsing and probing every word.
>Typical slshdot arrogance. How about IE has functionality that your sacred cow doesn't?
Even if that were true, IE is not even an option for Linux users. It would be one thing if Micro$oft really had a superior product that it offered impartially across all major platforms. Then if some users persisted on using an inferior product, perhaps you could blame them.
The reality is that Micro$oft *intentionally* develops incompatible browser implementation with proprietary extensions in order to gain advantage in the browser wars.
Micro$oft also *refuses* to develop versions for Linux so as to gain (unfair) competitive advantage in the platform wars. Even MacOS is only supported grudgingly (and often belatedly or incompletely) so that Micro$oft can claim that they are not a 100% monopolist.
The thing that amazes me even more is that some of the biggest eCommerce sites are broken without IE.
Recently, I needed to open an account on eTRADE in order to access a stock grant given to me by one of my clients. Well, it turns out that it is impossible to open an account without IE.
I then called tech support to complain. Well, the rep said that I had no choice but to use IE. I then said that I don't use IE because of security issues and that I was surprised that the leading eCommerce financial services company requires users to exclusively use buggy and insecure Micro$oft software. He kept insisting that I couldn't open an account and access my stock grants without IE. In fact, he couldn't even access my account until I used IE to first open the account.
I then suggested that maybe he could open it for me but he said that too was impossible. Finally, I got him to go to a supervisor to ask whether there is any way for me to get access to my money (i.e. stocks) without being required to use IE. After a long time on hold, he said that if I was willing to wait for 3-4 weeks they could snail mail out a written form that I could then fill out and return by snail me -- he warned that even after I returned it, manual processing would delay opening of the account. He was not even able to fax or email the form.
Even I was not willing to take that much time and effort to stand on principle. So, after haranguing the poor rep a little longer and finally getting him to file a complaint and bug report, I had no choice but to break down and launch my dusty copy of IE on my laptop.
What makes you think that people need to read ever word of every email they receive? Most people are busy and receive far more email than they need to. So, the most effective and efficient approach is to learn to prioritize and scan email for the key points rather than reading linearly like a kindergarten child.
Maybe they were scanning the email for key messages to see whether it needed further attention. Seeing words like "beer" might have triggered their natural inter parser to skip over that section as irrelevant.
And if in scanning their email, they missed some hidden gem of important information, perhaps the problem was that you did not structure your email properly and keep it sufficiently succinct so that the key information could be identified. I for one, always spend time thinking before composing to make sure that the gist of my email is clear from the subject and or initial sentence or two. If someone misses the key point of my email or even some hidden gem, then I may be as much as fault for not communicating clearly.
Your case is very different from the OPs where judges are presumably required to read (or at least skim) the entire brief and per the OP it seems that more than half of the brief was never even physically opened.
I go one step further... For any config file that I change, I rename the orig with the suffix.xorig and I name the modified one.xmod. Then I create a link from the unsuffixed version to the.xmod version. If I create a new config file that wasn't included in the distro rpm, then I label it with the suffix.xnew and link from the unsuffixed version to the.xnew version. That way upgrades (of packages or distros) and deletions don't remove my changes and I can also easily use 'find' to find all the config files that I changed and use diff to find the specific changes. When I upgrade packages I just do a quick search for rpmnew/rpmsave to see if any of my config files need to be updated or re-linked.
Yes... it's "perfect" if you believe in the "command-and-control" method of computing and that everyone will obey the registry laws to the letter as they come down from Lord Gates.
But in reality -- i.e. real life -- companies don't obey it. Registry entries get orphaned, corrupted, misplaced, misused, etc. I can't tell you the number of programs that I have had trouble removing because of registry issues. Also, every week when my Symantec System Works program runs, it finds hundreds of registry issues that seem to reappear no matter how many times they are corrected or scrubbed. And who can forget the dreaded registry corruption. Change/delete one entry and you can corrupt and lock up your computer with no easy way to recover (unless you backed up your registry recently).
I don't know about you but I am reasonable computer competent and I find the registry very cumbersome and confusing with often hundreds of entries strewn all over the place.
Linux is just sooooo easy. Config files are almost always text files in either/etc (for system-wide) or/home/ for user specific. Each config file is separate and independent. If someone rights a messed-up config program, it doesn't screw up everything else. And with a decent package manager, it is easy to update or delete old config files when programs are updated or deleted respectively. And it doesn't take special programs to read the config files nor does it take minutes to 'search' them -- just a simple, almost instantaneous grep. And good config files are commented so they allow for in-line documentation of various options.
So, yes in theory, one grand, humongous, registry file is "elegant" in some perfect world where everybody rights perfectly Microsoft-compliant code. In reality, the registry is a huge PITA that is infinitely worth then the humbler Unix ascii config file method.
Slashdot Mirror
I think we need to call in Al Gore. Maybe one day the global warming alarmists and hoaxsters will realize that change is a *natural* thing in this universe whether caused by inanimate or animate forces. Storms come and go. Icecaps expand and shrink. Glaciers advance and recede. Species thrive and decline. Get over it. Indeed, the one difference between animate and inanimate forces is that inanimate generated change is usually random in its effect while the net effect caused by animate generated change tends to be for the overall net better effect of humanity (not every aspect is positive, not every individual benefits equally and not every day is progress but the overall level of societal wealth, comfort, and knowledge tends to move upwards over time).
Ahhh but in 2005, it was running FC1 with a 2.4 kernel (I believe). So, the issue may very well not have been present 3 years ago...
But there are several hundred other reported crashes at precisely that moment... still a coincidence? I would bet NOT.
Well given that there have been reports of several hundred such crashes, I guess it can't be a coincidence unless there are a billion or so Linux servers ;)
I only said that the last log was 20 seconds before midnight and that based on my clock display, it crashed no later than 00:00:00. I would bet any amount of money that the actual crash was at the 00:00:00 GMT changeover.
But there have now been reports (just adding up the comments posted on slashdot and emails to me) of hundreds of machines going down at precisely 00:00:00 GMT (across multiple timezones). That combined set of data points plus the obvious potential issue of a leap second being introduces at that precise time would seem to make your coincidence theory astronomically unlikely.
As the OP and having read about several hundred other stable machines in multiple timezones crashing at that precise second, I do not believe that is a coincidence. Several people have reported 10 or more machines crashing at precisely 00:00:00 GMT -- it is so unlikely that hundreds of stable servers crashed precisely at this moment, that I would bet anything that this is NOT a coincidence. Plus, the last time my machine encountered a leap second, it was running FC1 with a very different kernel so that does not prove anything.
Except that based on the responders that there were hundreds of such crashes at exactly that time (GMT) -- including people who had 10 or more servers crash at exactly that moment.
I am running: 2.6.26.6-49.fc8
Hardware is ASUS P4PE. 2 1-TB Seagate SATA hardrives, 1 200MB PATA drive. 2GB DDR. 1 pchdtv5500 card, 1 winfast 2000XP tv card, 1 nVidia 6200 graphics card. Checked syslog, cron, maillog, mythbackend.log, wtmp,
Sorry - I was trying to keep the post short. Not a hardware failure in that it booted right up when I noticed it the next morning. No kernel panics or any other evidence of why it crashed.
Interesting, I run mythtv too but wasn't recording at the time of the crash. Has been stable since I rebooted a couple of hours ago.
I agree - but seems like an awfully unlikely coincidence. By my math, chance of a crash in any given 20 second period is just under 1 in 5 million: (2 crashes in 2237 days)/(2237*24*60*60 seconds) * (20 second interval). Seems suspicious to me...
I rebooted it and it went just fine. I looked at the logs and saw now errors. Last entry was in /var/log/maillog at 18:59:40 (not an error).
So, not sure how to figure it out - tempted to try to replicate though by setting time back to 18:59 on 12/31/08 (and shutting off ntpd)
Froze - couldn't ping or ssh or get console response. I know the time cuz last maillog entry was 18:59:40 and the clock (on my emacs session) said 18:59 at time of crash. Hardware is: ASUS P4P Rebooted without ever but required me to manually poweroff
What are you smoking?
### Setting up the guide subscription, and renewing it every few months (I'm told),
---> [Setup] Go to url, register, select zipcode, cable/satellite/OTA provider, check stations you want [5minutes?]
[Renewal] Every 3 months, you get a renewal email, click link, click "Extend my subscription", done [30 seconds]
Seems to me a lot easier, cheaper and faster than paying $12.95 each month (even with automatic payments)
### Recompile the kernel for a new feature
---> Haven't had to recompile a kernel since Slackware 1.0 or maybe Redhat 3.0.0.1 - more than 10 years ago
### need to hang on while the SQL database cleans itself
---> WTF are you talking about? This is utter nonsense. Sort of like saying, well you have to wait for the random logic turbolyzer to frobnicate the kryptonite-shielded holographic disk.
About the only valid point is your initial comment that setting up remotes can require tinkering but even this has gotten much easier with lirc packages.
Now I admit that MythTV may not be for anybody, but it is downright funny to hear Slashdot readers making up stuff like you do about the complexity. Sounds like you should be spending your time reading "computers for dummy" books rather than hanging out here on Slashdot.
The ease and difficulty of setting up MythTV depends on what approach you take -- that is what open software is all about - the user gets to control how they install, modify, and use the software.
:)]
Here are some options ranging from hard to easy
1. [Hardest] Download bleeding edge development releases, compile from source, patch/modify as needed, download & compile dependency programs too as needed, install on multiple backends and frontends, customize every possible setting, etc [Install & customization time = never done
2. Locate and dowload pre-compiled packages, install them along with dependencies, customize settings as needed [Install & customization time = 1/2 day]
3. Use a package manager such as yum under Fedora and point it to the atRPMS repository -- all necessary packages and dependencies will be installed automatically. Run mythtvsetup and supply a few basic answers, accept defaults on everything else. Do minimal if any customization using mythfrontend [Install & customization time = couple of hours max]
4. [Easiest] Install a myth-enabled version of Linux such as Mythdora with Myth pre-loaded and configured as a default app. No extra setup or customization. Just runs [Install & customization time = install length of the custom distro, maybe 1 hour]
Option #4 is the analog of installing MCE which is the only option for Windows Media Center. The only difference is that even with option #4, under Linux, you can still later come back and customize MythTV as much as you want (even to the extent of upgrading to bleeding edge sources and toggling all the customization switches). In contrast, with Windows all you get is the base Media Center and maybe the ability to change a little eye candy. And if the conspiracy guys at Micro$oft-Media companies use DRM and other "evil" things to hobble your PVR, well you are then stuck.
So before, you gripe about MythTV, please make sure you compare apples-to-apples -- and in that context, MythTV is just as easy while having more power, customization, and DRM-freedom.
> As for succinctness and structure, I have found, over three years of being the only person in charge of a system used by most of
> management (i.e. above me), that the only way to get anything accross is to make an email a maximum of three paragraphs, each
> containing a maximum of two sentences. And anything that is mildly complicated should be in a bulleted list. I think that's
> ridiculous. Basically, most people (especially when reading stuff at work) have something like an 8th grade attention span and
> literacy level. So you have to be like a newspaper, write for 13 year olds.
I think that the way you describe above is the CORRECT way to write a business email. It has nothing to do with attention span or education level. Rather, senior managers with dozens or even hundreds of indirect reports tend to get hundreds of emails a day. They are looking for the 'snapshot' picture to identify the key details or issues that they need to know and address. Anything more and you lose the wheat with the chaff and they end up spending most of the day on emails rather than on managing the business.
I find in general that email is not a good tool for very detailed communication -- for that either face-to-face or more formal presentations (like Powerpoint or even formatted text reports) tend to work better. Now not everyone will agree and some may prefer to communicate via long emails, but conversely, I don't think you can expect colleagues to read your every email like a Bible scholar parsing and probing every word.
Friends don't let Friends use Micro$oft...
Problem was not accessing the account but OPENING the account. I only tried it for stock option accounts so I cannot comment on general accounts.
>Typical slshdot arrogance. How about IE has functionality that your sacred cow doesn't?
Even if that were true, IE is not even an option for Linux users. It would be one thing if Micro$oft really had a superior product that it offered impartially across all major platforms. Then if some users persisted on using an inferior product, perhaps you could blame them.
The reality is that Micro$oft *intentionally* develops incompatible browser implementation with proprietary extensions in order to gain advantage in the browser wars.
Micro$oft also *refuses* to develop versions for Linux so as to gain (unfair) competitive advantage in the platform wars. Even MacOS is only supported grudgingly (and often belatedly or incompletely) so that Micro$oft can claim that they are not a 100% monopolist.
The thing that amazes me even more is that some of the biggest eCommerce sites are broken without IE.
Recently, I needed to open an account on eTRADE in order to access a stock grant given to me by one of my clients.
Well, it turns out that it is impossible to open an account without IE.
I then called tech support to complain. Well, the rep said that I had no choice but to use IE. I then said that I don't use IE because of security issues and that I was surprised that the leading eCommerce financial services company requires users to exclusively use buggy and insecure Micro$oft software. He kept insisting that I couldn't open an account and access my stock grants without IE. In fact, he couldn't even access my account until I used IE to first open the account.
I then suggested that maybe he could open it for me but he said that too was impossible. Finally, I got him to go to a supervisor to ask whether there is any way for me to get access to my money (i.e. stocks) without being required to use IE. After a long time on hold, he said that if I was willing to wait for 3-4 weeks they could snail mail out a written form that I could then fill out and return by snail me -- he warned that even after I returned it, manual processing would delay opening of the account. He was not even able to fax or email the form.
Even I was not willing to take that much time and effort to stand on principle. So, after haranguing the poor rep a little longer and finally getting him to file a complaint and bug report, I had no choice but to break down and launch my dusty copy of IE on my laptop.
What makes you think that people need to read ever word of every email they receive?
Most people are busy and receive far more email than they need to. So, the most effective and efficient approach is to learn to prioritize and scan email for the key points rather than reading linearly like a kindergarten child.
Maybe they were scanning the email for key messages to see whether it needed further attention. Seeing words like "beer" might have triggered their natural inter parser to skip over that section as irrelevant.
And if in scanning their email, they missed some hidden gem of important information, perhaps the problem was that you did not structure your email properly and keep it sufficiently succinct so that the key information could be identified. I for one, always spend time thinking before composing to make sure that the gist of my email is clear from the subject and or initial sentence or two. If someone misses the key point of my email or even some hidden gem, then I may be as much as fault for not communicating clearly.
Your case is very different from the OPs where judges are presumably required to read (or at least skim) the entire brief and per the OP it seems that more than half of the brief was never even physically opened.
I go one step further... .xorig and I name the modified one .xmod. Then I create a link from the unsuffixed version to the .xmod version. If I create a new config file that wasn't included in the distro rpm, then I label it with the suffix .xnew and link from the unsuffixed version to the .xnew version. That way upgrades (of packages or distros) and deletions don't remove my changes and I can also easily use 'find' to find all the config files that I changed and use diff to find the specific changes. When I upgrade packages I just do a quick search for rpmnew/rpmsave to see if any of my config files need to be updated or re-linked.
For any config file that I change, I rename the orig with the suffix
Yes... it's "perfect" if you believe in the "command-and-control" method of computing and that everyone will obey the registry laws to the letter as they come down from Lord Gates.
/etc (for system-wide) or /home/ for user specific. Each config file is separate and independent. If someone rights a messed-up config program, it doesn't screw up everything else. And with a decent package manager, it is easy to update or delete old config files when programs are updated or deleted respectively. And it doesn't take special programs to read the config files nor does it take minutes to 'search' them -- just a simple, almost instantaneous grep. And good config files are commented so they allow for in-line documentation of various options.
But in reality -- i.e. real life -- companies don't obey it. Registry entries get orphaned, corrupted, misplaced, misused, etc.
I can't tell you the number of programs that I have had trouble removing because of registry issues. Also, every week when my Symantec System Works program runs, it finds hundreds of registry issues that seem to reappear no matter how many times they are corrected or scrubbed. And who can forget the dreaded registry corruption. Change/delete one entry and you can corrupt and lock up your computer with no easy way to recover (unless you backed up your registry recently).
I don't know about you but I am reasonable computer competent and I find the registry very cumbersome and confusing with often hundreds of entries strewn all over the place.
Linux is just sooooo easy. Config files are almost always text files in either
So, yes in theory, one grand, humongous, registry file is "elegant" in some perfect world where everybody rights perfectly Microsoft-compliant code. In reality, the registry is a huge PITA that is infinitely worth then the humbler Unix ascii config file method.