I bought a Galaxy Note, and I love it. I'm not going to put anything else in the same pocket with it anyway, so why not fully utilize that space? The screen size makes it so much more generally useful than phones I've had in the past.
and it's based on a flawed idea of how the internet works
What part was flawed, exactly? The Internet isn't just a big cloud that one pays for a connection to. Just because the customer pays for a connection to the ISP's network, and YouTube pays for a connection somebody else's network, doesn't mean that there aren't other links in between those two networks that also have finite capacity, and also cost money. Some of the big networks might have no-cost peering agreements between them, but that's certainly not how everything functions; lots of traffic flows across connections that have monthly bills associated with them. And even with "free" peering links, there is a cost to increase the (finite) capacity of peering links if that becomes necessary.
A "connection to the Internet" doesn't really exist the way you seem to think it does.
Think about what the article said, though. Your phone is connected to a small tower in the plane, and it functions like a separate country for the purposes of the cellular network, so you're paying international roaming rates. Now I don't know about your plan, but for me, that's at least a couple bucks a minute. For a sufficiently important (and short) call, I'll pay that. But I sure won't be droning on with inane chatter for a couple hours like you're implying.
To be honest, I don't really see this as being too different than the wired phones that are already present on a lot of planes. It will be expensive to use, and that alone will be enough to keep it from being a nuisance.
Actually, you have that backwards. POSIX timezone definitions (the things you find in/usr/share/zoneinfo on Linux) describe how DST works in all different years. If you convert a UTC timestamp in March 2005 to your local timezone, it won't use DST, but if you do it for a UTC timestamp in March 2007, it'll know that then it needs to use DST. Thus, you're actually better off storing everything in UTC, because then you know what time everything really took place / will take place, in any timezone you care to know it in.
Honestly, I can't think for the life of me why they haven't become a non-profit yet.
They may well be. However, they're also Canadian. That means:
Just incorporating as a non-profit isn't enough. They'd also have to register as a charity, and in Canada, that means a lot of paperwork, and a lot of restrictions.
I'm not sure how international donations work for tax purposes, but I bet it still wouldn't be easy for Americans to write off their donations, even if OpenBSD were set up as a Canadian charitable organization.
There's one key LILO feature missing from GRUB, as far as I know: lilo -R
This allows me to install a new kernel on a box I'm not in front of, and tell LILO to boot it by default for the next boot only. If the new kernel doesn't work, I only have to ask somebody near the machine to reboot it for me, and it'll come back up in my old, working kernel. With GRUB, I'd have to try to talk somebody through hooking up a monitor and picking the right kernel... when it's a headless colocated server located somewhere far away, that's not always an appealing idea.
Remember OS/2? There were a lot of "true believers" trying to spread the word about OS/2 (myself included). Heck, they even formed Team OS/2, all to promote this commercial product made by IBM.
And it worked great, too! That's why everybody uses OS/2 today... er, waitaminute.
2. There are too many addresses. There are 16.7 million addresses per square metre of the earth's surface, including the oceans. This is overkill.
It's deliberate overkill. It allows things like 64-bit subnets, which in turn allow for stateful autoconfiguration. It also allows for large chunks of address space that won't be allocated at all; if it turns out in the future that our current allocation method is inadequate for our needs, we can simply devise a new allocation method in this empty space, rather than having to migrate to a whole new version of IP.
3. The problem with a 64-bit network prefix is that routing tables become massive. Just do the math and you'll see that extreme amounts of memory are required to hold routing tables.
Yes, if an IPv6 router had to hold nearly 150,000 routes in memory like it does in the current IPv4 world, it would be massive. Fortunately, IPv6 is designed to have properly aggregated addresses, so that things are much more hierarchical, and routing tables can be stored much more efficiently.
4. The IPv6 header is too large.
Aside from the fact that more and more connections are using much larger MTUs these days, IPv6 also supports more aggressive header compression than IPv4 did, often resulting in similarly compact headers.
Sounds like something you might want to do with OpenOffice.org. It's quite capable of using data directly from a PostgreSQL (or other) database, and performing mail merges with it, which should enable you to create pretty much any sort of form you need.
Heh, I did that last year. Clobbered my boot sector, partition table, and/boot partition... but the system still worked. Every time a process tried to touch/boot, it would hang in "D" status, so the nightly updatedb process would hang every day, increasing the load average by 1. The load average was nearly 60 (i.e., two months later) before I decided to do something about it.
I used GNU parted to detect my partitions and re-create the partition table. I then re-mke2fs'd my/boot, and re-installed the kernel RPM to rebuild that. Finally, I ran LILO again, crossed my fingers, and rebooted.
Hey, I've got an old RaQ2 MIPS that I've been looking for updates or more importnatly I new kernel but finding LINUX MIPS for the RaQ2 is next to impossible.
I've found that the Cobalt port of NetBSD works quite nicely. Unlike trying to run Linux on your RaQ2, you can actually get a current version of NetBSD for it (and not an unofficial and flakey patch either!).
In regards to Access, however, I still don't see how they can claim to be an Internet Service Provider if they're going to censor the web (read: newsgroups)
They don't censor the web. They also don't censor Usenet (which is not a part of the web, inciddentally). However, they do have a feed that doesn't include the alt.binaries.* hierarchy. Is that what you're complaining about?
"hey wanna turn on the net for a half sec so I can grab a removal tool?"
This kinda got old after about the 5th time last month!
Okay, let me get this straight? You're talking about a customer that has been infected five times in one month, and you're expecting that Access is going to trust your ability to clean up an infection properly? Personally, I'd say the connection should be turned off permanently!:-P
Either way, I don't see why turning on a connection to allow you to spew more viruses is such a good thing. Access generally will do it the first time, but if you're getting infected five times a month, I don't blame them for refusing.
The above comments are my own, and I in no way represent the views/opinions of my employer, or anyone else.
* Yeah, yeah, Redhat supplies much more than just an OS, its all the apps too. I know, but even a support policy that covered the OS and the 50 most common apps would be an improvement.
I notice that Mandrake's product lifetime policy includes this idea. They support the "base" system longer than they do the desktop software, which means that servers don't need major upgrades as frequently to keep them secure.
Does anybody know off-hand what the x-height for the Bitstream Vera fonts is? Or better yet, does anybody know how I can determine that for myself?
This value is useful for the font-size-adjust CSS parameter, to make sure that if font substitution takes place, the resulting font still has the same apparent size as the designer intended. font-size-adjust basically defines the ratio between the height of the font, and its x-height (the height of a lower-case letter).
Next, build two or more tunnels to the ISP over different circuts or providers and run your own small BGP network on private IP's between the router at your multihomed isp and the routers on either end of your connection.
Yes, I suppose this would actually work. In fact, it's pretty much what I'm doing for my own network:-). You might not want to use BGP, though; it isn't really designed to run across several hops like that, and will probably flap more than you'd like.
OSPF will do the job just fine, and provide reasonably quick convergence too.
If you're doing this, keep in mind that -- to the outside world -- both your nameservers are equal. You'll get approximately half the clients querying your secondary nameserver, meaning that approximately half the traffic will hit that second link.
Of course, if they have similar amounts of bandwidth, etc., you might actually consider this to be a good thing:-).
It's still not going to work great for fail-over, though (thanks to things like DNS caches). As distasteful as it might be, the most cost-effective way for you to approach this is probably going to be with an SLA (service-level agreement) with your ISP.
Re:I'll guess I'll admit it..
on
Slashdot over IPv6
·
· Score: 1, Informative
And furthermore, i'd say the "end of NAT" is a bit too much. I find it very useful to use a NAT gateway/firewall and put insecure clients behind that. It reduces the need to think secure on the local network. I can for instance export my fileserver data rw onto 192.168 without much consern. Wouldn't wanna do that if they were all "real" IP's.
Why not? If you set up a trivial firewall using iptables and allowing only --state ESTABLISHED,RELATED (or the equivalent using some other stateful firewall), you have exactly the same level of security as you do with NAT. If you also throw in a rule to ensure that no spoofed packets come in with a source address that matches your internal subnet, you can now safely export your fileserver data.
Remember, --state ESTABLISHED,RELATED means that you're protected by exactly the same connection tracking code as you are with NAT. And, by eliminating NAT, you're no longer breaking the end-to-end nature of IP.
Actually, try it sometime. It works just fine, as long as the web server supports HTTP versions 1.0. You also don't need to hit enter a second time with this form, because there are no headers.
Taco Hell used to be "the" Slashdot Easter Egg. It began stagnating quite a while ago, and I see that it's now even lost its wonderfully hideous purple look. Ah well.
This is a showstopper for me. Turning compression off does not solve the problem for me.
Well, I installed OpenSSH 3.4p1 on a machine running kernel 2.0.38, and it's working fine. I didn't have to mess with any extra patches, or with the Compression setting in the configuration.
Then why the hell is enable by default in OpenBSD sshd daemon?
Okay, so you're using SSH. Your network traffic is all nicely secure and encrypted, and nobody can sniff your password. But then you sit down in an Internet cafe somewhere, install a copy of PuTTY, and cheerfully log into your server. Obviously, PuTTY asks you for your password. You just type it in and go, right?
Well, what if somebody had installed a keyboard sniffer on that machine? It's a public machine operated by a relatively unknown third party; there are lots of ways it could happen.
With a one-time password, this wouldn't be as big a deal.
- Does Mozilla 'do the right thing' with a read-only NFS mounted directoy yet? In the past, user prefs were stored under all various subdirs of the product, and it was unusable for a network-based install
Mozilla stores prefs in ~/.mozilla, and has for ages now. My user has no write privileges to the Mozilla installation files, and it works just fine. I'd assume that this means that read-only NFS should work fine.
- Will there *EVER* be a release of mozilla or netscape 6.X that runs on glibc-2.0 systems?
Probably not. glibc 2.0 just "wasn't there" as far as thing like threading are concerned.
Or you could try my solution when I had a client that insisted I use ASP: PerlScript. If you checkout ActiveState's Perl port, you'll find that it includes PerlScript, which basically ties Perl into the Windows scripting system alongside VBScript and JScript.
This means that you can write ASP pages with Perl code in them instead of VBScript code! They just need to install ActiveState Perl on the server, which isn't much of a stretch.
The other nice bonus is that much of the code you write this way will work equally well on a Linux server running Apache and Apache::ASP!
Slashdot Mirror
I bought a Galaxy Note, and I love it. I'm not going to put anything else in the same pocket with it anyway, so why not fully utilize that space? The screen size makes it so much more generally useful than phones I've had in the past.
What part was flawed, exactly? The Internet isn't just a big cloud that one pays for a connection to. Just because the customer pays for a connection to the ISP's network, and YouTube pays for a connection somebody else's network, doesn't mean that there aren't other links in between those two networks that also have finite capacity, and also cost money. Some of the big networks might have no-cost peering agreements between them, but that's certainly not how everything functions; lots of traffic flows across connections that have monthly bills associated with them. And even with "free" peering links, there is a cost to increase the (finite) capacity of peering links if that becomes necessary.
A "connection to the Internet" doesn't really exist the way you seem to think it does.
Think about what the article said, though. Your phone is connected to a small tower in the plane, and it functions like a separate country for the purposes of the cellular network, so you're paying international roaming rates. Now I don't know about your plan, but for me, that's at least a couple bucks a minute. For a sufficiently important (and short) call, I'll pay that. But I sure won't be droning on with inane chatter for a couple hours like you're implying.
To be honest, I don't really see this as being too different than the wired phones that are already present on a lot of planes. It will be expensive to use, and that alone will be enough to keep it from being a nuisance.
Actually, you have that backwards. POSIX timezone definitions (the things you find in /usr/share/zoneinfo on Linux) describe how DST works in all different years. If you convert a UTC timestamp in March 2005 to your local timezone, it won't use DST, but if you do it for a UTC timestamp in March 2007, it'll know that then it needs to use DST. Thus, you're actually better off storing everything in UTC, because then you know what time everything really took place / will take place, in any timezone you care to know it in.
They may well be. However, they're also Canadian. That means:
There's one key LILO feature missing from GRUB, as far as I know: lilo -R
This allows me to install a new kernel on a box I'm not in front of, and tell LILO to boot it by default for the next boot only. If the new kernel doesn't work, I only have to ask somebody near the machine to reboot it for me, and it'll come back up in my old, working kernel. With GRUB, I'd have to try to talk somebody through hooking up a monitor and picking the right kernel... when it's a headless colocated server located somewhere far away, that's not always an appealing idea.
What Linux distribution doesn't have an IPv6 stack built in these days?
And for that matter, Windows users don't have to wait for Longhorn either. Windows XP has an IPv6 stack built in too: How to install IPv6
Remember OS/2? There were a lot of "true believers" trying to spread the word about OS/2 (myself included). Heck, they even formed Team OS/2, all to promote this commercial product made by IBM.
And it worked great, too! That's why everybody uses OS/2 today... er, waitaminute.
Okay, I won't argue with you there.
It's deliberate overkill. It allows things like 64-bit subnets, which in turn allow for stateful autoconfiguration. It also allows for large chunks of address space that won't be allocated at all; if it turns out in the future that our current allocation method is inadequate for our needs, we can simply devise a new allocation method in this empty space, rather than having to migrate to a whole new version of IP.
Yes, if an IPv6 router had to hold nearly 150,000 routes in memory like it does in the current IPv4 world, it would be massive. Fortunately, IPv6 is designed to have properly aggregated addresses, so that things are much more hierarchical, and routing tables can be stored much more efficiently.
Aside from the fact that more and more connections are using much larger MTUs these days, IPv6 also supports more aggressive header compression than IPv4 did, often resulting in similarly compact headers.
Sounds like something you might want to do with OpenOffice.org. It's quite capable of using data directly from a PostgreSQL (or other) database, and performing mail merges with it, which should enable you to create pretty much any sort of form you need.
I used GNU parted to detect my partitions and re-create the partition table. I then re-mke2fs'd my /boot, and re-installed the kernel RPM to rebuild that. Finally, I ran LILO again, crossed my fingers, and rebooted.
Worked out just fine :-).
I've found that the Cobalt port of NetBSD works quite nicely. Unlike trying to run Linux on your RaQ2, you can actually get a current version of NetBSD for it (and not an unofficial and flakey patch either!).
Horrible ping times? Hmm, let's see...
An average of two and a half milliseconds? That doesn't sound bad to me. It sounds like your "mate" needs a house call ;-).
Oh, and if you're only using 20 gigs a month, Access wouldn't care either. The people running into trouble are using much more than that.
The above comments are my own, and in no way represent the views/opinions of my employer, or anyone else.
They don't censor the web. They also don't censor Usenet (which is not a part of the web, inciddentally). However, they do have a feed that doesn't include the alt.binaries.* hierarchy. Is that what you're complaining about?
Okay, let me get this straight? You're talking about a customer that has been infected five times in one month, and you're expecting that Access is going to trust your ability to clean up an infection properly? Personally, I'd say the connection should be turned off permanently! :-P
Either way, I don't see why turning on a connection to allow you to spew more viruses is such a good thing. Access generally will do it the first time, but if you're getting infected five times a month, I don't blame them for refusing.
The above comments are my own, and I in no way represent the views/opinions of my employer, or anyone else.
This value is useful for the font-size-adjust CSS parameter, to make sure that if font substitution takes place, the resulting font still has the same apparent size as the designer intended. font-size-adjust basically defines the ratio between the height of the font, and its x-height (the height of a lower-case letter).
OSPF will do the job just fine, and provide reasonably quick convergence too.
Of course, if they have similar amounts of bandwidth, etc., you might actually consider this to be a good thing :-).
It's still not going to work great for fail-over, though (thanks to things like DNS caches). As distasteful as it might be, the most cost-effective way for you to approach this is probably going to be with an SLA (service-level agreement) with your ISP.
Remember, --state ESTABLISHED,RELATED means that you're protected by exactly the same connection tracking code as you are with NAT. And, by eliminating NAT, you're no longer breaking the end-to-end nature of IP.
Actually, try it sometime. It works just fine, as long as the web server supports HTTP versions 1.0. You also don't need to hit enter a second time with this form, because there are no headers.
Taco Hell used to be "the" Slashdot Easter Egg. It began stagnating quite a while ago, and I see that it's now even lost its wonderfully hideous purple look. Ah well.
Well, I installed OpenSSH 3.4p1 on a machine running kernel 2.0.38, and it's working fine. I didn't have to mess with any extra patches, or with the Compression setting in the configuration.
Okay, so you're using SSH. Your network traffic is all nicely secure and encrypted, and nobody can sniff your password. But then you sit down in an Internet cafe somewhere, install a copy of PuTTY, and cheerfully log into your server. Obviously, PuTTY asks you for your password. You just type it in and go, right?
Well, what if somebody had installed a keyboard sniffer on that machine? It's a public machine operated by a relatively unknown third party; there are lots of ways it could happen.
With a one-time password, this wouldn't be as big a deal.
Doesn't get much simpler than that ;-).
Or you could try my solution when I had a client that insisted I use ASP: PerlScript. If you checkout ActiveState's Perl port, you'll find that it includes PerlScript, which basically ties Perl into the Windows scripting system alongside VBScript and JScript.
This means that you can write ASP pages with Perl code in them instead of VBScript code! They just need to install ActiveState Perl on the server, which isn't much of a stretch.
The other nice bonus is that much of the code you write this way will work equally well on a Linux server running Apache and Apache::ASP!