I wasn't saying that's why people haven't moved. Given a choice, I would sooner pay less per minute than per month. The cost would be one factor in moving to voip.
I think there are a lot of reasons people haven't moved. I don't own a land line and having VOIP is useless to me.
In Vancouver, BC we have a service called CityFido. You get flat rate cell phone usage -- I pay $45 a month for my phone no matter how much I use it and long distance packages are fairly competitive.
What's the point in VOIP when I can use my cellphone cheaply anywhere?
How is he less than honest? You can get the source code and binaries for free, just like any other OSS project.
He charges for support and beta/alpha software. Beta/Alpha eventually becomes stable and thus free. In fact the FSF says his model is in compliance with the GPL.
Doesn't include: mean look up your ISP's SPF records, and if there isn't one, this rule does nothing? Or worse, bounces your message (I'm not 100% sure on how this rule is processed)?
SPF isn't meant to stop spam. It is meant to stop spam that spoofs the from address.
This means all the spam that comes from AOL and Hotmail accounts that don't actually leave from there servers would be bounced at your mail servers. At this point in time, if everyone used SPF, my guess is that at least 50% of spam would be blocked.
Of course, spammers are going to register domains to use for spamming and set SPF records so that their mail appears legit to the SPF filters.
You're probably thinking, "What's the point?" Well, it's easier to understand if you have ever hosted a domain that has been either blacklisted or had an increase in bandwidth charges because of millions of bouncebacks due to spammers using a FROM address in your domain.
Using old tested software is a positive in some senses, but even servers need to be progressive. Look at how much (Net|UCD)-SNMP has changed in 2 years, or the difference between Mysql 3.23 and 4.0.
Instead of dragging everyone behind, why not have scheduled releases? Fedora is using this system and I'm pretty sure that RHEL is as well. I'm not saying that Fedora's system is great (it's about as far from a Debian style distro as you can get), but a hybrid would probably keep most people happy.
If Debian tagged a 'channel' every 6 months, people could keep their older systems running the versions of the software that they were installed with (plus patches). They'd also have an option for getting more up to date programs, kernels and so forth.
The problem with the debian system is that everyone needs to use the unstable version to get the software they want to work.
It's great that you can apt-get install just about every piece of software. Too bad the stable tag uses versions of software that were old when woody was marked stable.
I bit the bullet and bought RHEL licenses for the last round of upgrades. It works great, other than the realization that it's basically paying through the nose for what RH8.0 gave you for free. Well, that and it uses a buggy perl 5.8.0.
I went through this a year and a half ago. It worked out really well. The CEO came to us with a short list and set up interviews. Everyone had been prescreened and so forth.
The CEO wanted us to ask technical questions. That's silly -- a manager doesn't do technical things. We drilled the candidates on various hypothetical/real situations that have come up. Things like "The marketing team wants more features to make the product more sellable. What do you do?"
You really need to go by track records and references. You want someone who has gotten the job done and stood up for the technical team in the past.
We were lucky. Our top choice matched the executives. I don't know what we would have done if it didn't..
SCSI is stupidly expensive -- I can build a 750GB raid 5 (4x250GB) for the same price as a 140GB SCSI drive. Both of these solutions are roughly $1000CAD. Even throwing in a 3ware SATA controller is still cheaper than doing a software Raid 1.
I also think that everyone is going a little overboard -- I'm pretty sure that original poster does not need redundant servers running linux HA and development and production servers. They can probably afford downtime over hot swappable memory dimms and CPU's. I'm also pretty sure that they don't need E450's and one poster suggested.
My guess is that at a 15 person shop that needs to ask such questions, they probably have a file server. They probably don't have a domain controller. I bet they outsource their email to their ISP and colocated their website for cheap somewhere.
My recommendation is to poke around the community and find a small grey box company that offers IT services for small businesses. Make sure you talk to some of their clients and make sure they offer similar services to what you are looking for.
They can then build you servers with quality parts (not the crap they put in HP's or Dell's). They can also do it for a lot cheaper. They should show you how to back up your servers using something like BackupExec. Plus, they should offer some sort of warranty. If it breaks, they should be able to come on site to fix it for you.
I believe they go through brightmail or another one of the big companies. You need to pay a big chunk to get onto their whitelists every year.
One of our anti-spam guys has been engaging in dialogue with a lot of the big mail providers like aol, hotmail and yahoo. You won't find it on their front pages, but if you dig around you'll be able to contact their post masters and work from there.
I've had a 250gb SATA drive fail on a 1TB array on a 3ware card. It was about 4 hours for it to rebuild. The system was slower, but we didn't have 'cascading failures.'
In fact, the only time I've experienced 'Cascading Failures' was on an expensive Mylex SCSI raid controller. There is nothing like saying "Shit, we just lost two drives on that raid 5."
Banks can continue to use SCSI, but I'm going to use SATA everywhere. It'll save me over half the cost for the same size array.
You still have to plan for failures everywhere. That's why I have spare drives and backups for all my arrays, even my SCSI ones.
I don't know.. I was rolling on the ground laughing when he talks about the trench on the death star being a birth canal and an "edipal confrontation."
Hmm, I work in a small/medium business. By the time people were talking about it on Full-Disclosure, my mail server virus scanners were already up to date. I saw about 4 messages in total.
I haven't gotten any help desk calls about infections yet. I don't think I will either. The desktop antivirus was up to date before I left on Monday as well.
Administration isn't that difficult and there is very little difference in virus scanner software and updates now -- there is too much competition for the crap to get through.
All this from a mail system I spent $300 in licensing for -- just the Fprot virus scanner plugged into qmail-scanner. Oh, and cron+wget to update the dat files.
And you are the reason that RBL's cause so much collateral damage.
It's great that you are taking this political stand and sticking it to the virus scanner companies. I'm sure all the email admins out there make the logical jump that their virus scanner messages are causing their IP addresses to show up in RBL's. They'll all disable their virus bounce messages for you.
Actually, now that I think about it, it's more likely that people will assume RBL's are useless and don't work. They'll probably complain to their peers and convince them that RBL's are unreliable.
Sweeping generalizations like this are bad. You shouldn't say "Never update your windows system," you should say "Only update it when the problem affects you," or "Only update after you've thoroughly tested the patch." Both of these use common sense. Never patching is not.
SP2 is probably going to break a lot of software. Especially the kind that is built with bad assumptions (pinging hosts, expected levels of access etc). Am I going to deploy it day 1? No. Am I going to what a month and see what everyone else says, then install it on a small test group? Yes.
Yes, on modern systems, sh is bash. But, when you run it as sh, it runs in compatibility mode. If you try to use bash syntax when executing as/bin/sh, you will die a horrible death.
So? How is this a problem? If people don't care or have anything to contribute they don't have to vote. What's worse is people voting who have been mislead or misinformed.
While we're talking about the "real" problem, I think it's corrupt and selfish people. Why should we have to worry about people cheating? I'd be much more worried about someone buying their way into power than if some people who don't really care not voting. The fact we have to worry about that is sad.
Don't get me wrong, people should vote. But if they don't want to, that is their right just as much as it is to vote.
The productivity gained by the scroller is worth not having a real button. The scroller functioned fine and did not have a problem on every scroll mouse I've used. I use X quite a bit, which means a lot of middle mouse button clicking. Sticking with Logitech and Microsoft through the ages has served me well.
Everyone I've ever discussed it with agrees that the optical scroll mouse is the pinacle of mice technology.
Slashdot Mirror
I wasn't saying that's why people haven't moved. Given a choice, I would sooner pay less per minute than per month. The cost would be one factor in moving to voip.
I think there are a lot of reasons people haven't moved. I don't own a land line and having VOIP is useless to me.
In Vancouver, BC we have a service called CityFido. You get flat rate cell phone usage -- I pay $45 a month for my phone no matter how much I use it and long distance packages are fairly competitive.
What's the point in VOIP when I can use my cellphone cheaply anywhere?
You're probably in the minority. My guess is that most people would prefer to have lower per minute rates than monthly rates.
I supposed the ideal would be having different packages -- the more you pay flat, the less you pay per minute..
He charges for support and beta/alpha software. Beta/Alpha eventually becomes stable and thus free. In fact the FSF says his model is in compliance with the GPL.
Doesn't include: mean look up your ISP's SPF records, and if there isn't one, this rule does nothing? Or worse, bounces your message (I'm not 100% sure on how this rule is processed)?
SPF isn't meant to stop spam. It is meant to stop spam that spoofs the from address.
This means all the spam that comes from AOL and Hotmail accounts that don't actually leave from there servers would be bounced at your mail servers. At this point in time, if everyone used SPF, my guess is that at least 50% of spam would be blocked.
Of course, spammers are going to register domains to use for spamming and set SPF records so that their mail appears legit to the SPF filters.
You're probably thinking, "What's the point?" Well, it's easier to understand if you have ever hosted a domain that has been either blacklisted or had an increase in bandwidth charges because of millions of bouncebacks due to spammers using a FROM address in your domain.
Using old tested software is a positive in some senses, but even servers need to be progressive. Look at how much (Net|UCD)-SNMP has changed in 2 years, or the difference between Mysql 3.23 and 4.0.
Instead of dragging everyone behind, why not have scheduled releases? Fedora is using this system and I'm pretty sure that RHEL is as well. I'm not saying that Fedora's system is great (it's about as far from a Debian style distro as you can get), but a hybrid would probably keep most people happy.
If Debian tagged a 'channel' every 6 months, people could keep their older systems running the versions of the software that they were installed with (plus patches). They'd also have an option for getting more up to date programs, kernels and so forth.
The problem with the debian system is that everyone needs to use the unstable version to get the software they want to work.
It's great that you can apt-get install just about every piece of software. Too bad the stable tag uses versions of software that were old when woody was marked stable.
I bit the bullet and bought RHEL licenses for the last round of upgrades. It works great, other than the realization that it's basically paying through the nose for what RH8.0 gave you for free. Well, that and it uses a buggy perl 5.8.0.
Hey,
I went through this a year and a half ago. It worked out really well. The CEO came to us with a short list and set up interviews. Everyone had been prescreened and so forth.
The CEO wanted us to ask technical questions. That's silly -- a manager doesn't do technical things. We drilled the candidates on various hypothetical/real situations that have come up. Things like "The marketing team wants more features to make the product more sellable. What do you do?"
You really need to go by track records and references. You want someone who has gotten the job done and stood up for the technical team in the past.
We were lucky. Our top choice matched the executives. I don't know what we would have done if it didn't..
Hope that helps some.
SATA is the only way to go.
SCSI is stupidly expensive -- I can build a 750GB raid 5 (4x250GB) for the same price as a 140GB SCSI drive. Both of these solutions are roughly $1000CAD. Even throwing in a 3ware SATA controller is still cheaper than doing a software Raid 1.
I also think that everyone is going a little overboard -- I'm pretty sure that original poster does not need redundant servers running linux HA and development and production servers. They can probably afford downtime over hot swappable memory dimms and CPU's. I'm also pretty sure that they don't need E450's and one poster suggested.
My guess is that at a 15 person shop that needs to ask such questions, they probably have a file server. They probably don't have a domain controller. I bet they outsource their email to their ISP and colocated their website for cheap somewhere.
My recommendation is to poke around the community and find a small grey box company that offers IT services for small businesses. Make sure you talk to some of their clients and make sure they offer similar services to what you are looking for.
They can then build you servers with quality parts (not the crap they put in HP's or Dell's). They can also do it for a lot cheaper. They should show you how to back up your servers using something like BackupExec. Plus, they should offer some sort of warranty. If it breaks, they should be able to come on site to fix it for you.
That's all.
I believe they go through brightmail or another one of the big companies. You need to pay a big chunk to get onto their whitelists every year.
One of our anti-spam guys has been engaging in dialogue with a lot of the big mail providers like aol, hotmail and yahoo. You won't find it on their front pages, but if you dig around you'll be able to contact their post masters and work from there.
Yes, but unless you're making millions from your service, you won't be able to afford getting onto hotmail's whitelist.
AOL does have a feedback loop that lets you know when your mail is blocked. I can't remember what it is called off hand though.
You sound like a sales guy pushing SCSI.
I've had a 250gb SATA drive fail on a 1TB array on a 3ware card. It was about 4 hours for it to rebuild. The system was slower, but we didn't have 'cascading failures.'
In fact, the only time I've experienced 'Cascading Failures' was on an expensive Mylex SCSI raid controller. There is nothing like saying "Shit, we just lost two drives on that raid 5."
Banks can continue to use SCSI, but I'm going to use SATA everywhere. It'll save me over half the cost for the same size array.
You still have to plan for failures everywhere. That's why I have spare drives and backups for all my arrays, even my SCSI ones.
Even on a 200mhz box, you can still use X forwarding.
If that isn't an option, you can use tcpdump to log to a file. Then scp that file to your workstation and open the stream up in ethereal. Same diff.
I'm thinking Uma Therman would make a kick ass Samus. Just make sure she puts Kill Bill on her resume before she submits it to Woo.
Last I heard, OSC was working on a screenplay for Ender's Game. No URL's, but it was up on his website. I'm sure google will point to this.
I don't know.. I was rolling on the ground laughing when he talks about the trench on the death star being a birth canal and an "edipal confrontation."
Would Lucas really say something like that?
Hmm, I work in a small/medium business. By the time people were talking about it on Full-Disclosure, my mail server virus scanners were already up to date. I saw about 4 messages in total.
I haven't gotten any help desk calls about infections yet. I don't think I will either. The desktop antivirus was up to date before I left on Monday as well.
Administration isn't that difficult and there is very little difference in virus scanner software and updates now -- there is too much competition for the crap to get through.
All this from a mail system I spent $300 in licensing for -- just the Fprot virus scanner plugged into qmail-scanner. Oh, and cron+wget to update the dat files.
And you are the reason that RBL's cause so much collateral damage.
It's great that you are taking this political stand and sticking it to the virus scanner companies. I'm sure all the email admins out there make the logical jump that their virus scanner messages are causing their IP addresses to show up in RBL's. They'll all disable their virus bounce messages for you.
Actually, now that I think about it, it's more likely that people will assume RBL's are useless and don't work. They'll probably complain to their peers and convince them that RBL's are unreliable.
Way to go, jerk.
In the book version of Starship Troopers by Heinlein, there is a whole military division of soldiers paired with cybernetic dogs called the K-9 Corps.
My copy is on loan to a friend, but I did find a PDF of it on the internet through google. Don't know if it is legit though: Starship Troopers
Sweeping generalizations like this are bad. You shouldn't say "Never update your windows system," you should say "Only update it when the problem affects you," or "Only update after you've thoroughly tested the patch." Both of these use common sense. Never patching is not.
SP2 is probably going to break a lot of software. Especially the kind that is built with bad assumptions (pinging hosts, expected levels of access etc). Am I going to deploy it day 1? No. Am I going to what a month and see what everyone else says, then install it on a small test group? Yes.
Yes, on modern systems, sh is bash. But, when you run it as sh, it runs in compatibility mode. If you try to use bash syntax when executing as /bin/sh, you will die a horrible death.
So? How is this a problem? If people don't care or have anything to contribute they don't have to vote. What's worse is people voting who have been mislead or misinformed.
While we're talking about the "real" problem, I think it's corrupt and selfish people. Why should we have to worry about people cheating? I'd be much more worried about someone buying their way into power than if some people who don't really care not voting. The fact we have to worry about that is sad.
Don't get me wrong, people should vote. But if they don't want to, that is their right just as much as it is to vote.
The productivity gained by the scroller is worth not having a real button. The scroller functioned fine and did not have a problem on every scroll mouse I've used. I use X quite a bit, which means a lot of middle mouse button clicking. Sticking with Logitech and Microsoft through the ages has served me well.
Everyone I've ever discussed it with agrees that the optical scroll mouse is the pinacle of mice technology.
You've probably have never had to deal with more than 50 domains before, have you?
Finally, a negative book review shows up on slashdot. I was getting tired of all these 8's, 9's and 10's -- they were diluting the scale.