However, when you moved your VOIP, you'd have a new IP number, yes? And if that IP number was an IPv6 number, then the number would (possibly?) have a geographical correlation. Or at least an ISP correlation.
I'm talking about ways of making it work even when moving, your talking about how it needs to be movable. You did read my post, right?
I'm wondering if the heirachy structure of the IPv6 address could be used here. I don't know how IP roaming works (apparently a feature in IPv6) but could the header of the address be used to track you to your access point, and if the ISP was required to have access point addresses, then they could use those? OK, already I see a problem of shipping address versus billing address, and it would require the ISP's have open access logs, but don't you need them already to save you all from terrorists?
Depending on where you are, this may vary, but I understand that "Caller ID blocking" is just raising a flag on the Caller ID signal, that says "here is their ID, but please don't display it". I beleive that call centers, especially ones with reversed charge numbers (free calls) are permitted to have phones which ignore the Caller ID blocking flag.
I wouldn't say they were the first to use it, but certainly the first of such note.
First case I saw of some-one using autorun to screw with a system was when I borrowed a backup CD off a friend, and its autorun would open a very noticable JPG image. Since I was borrowing the CD to update something on my mothers machine... fortunately, she wasn't in the room at the time!
I believe it can also be bypassed by holding down the shift key while inserting the CD into the drive (temporarily disabling AutoRun), or by permanately disabling AutoRun.
Using a bit of tap to do it is just grandstanding.
Excuse me, but you are not correct. I don't know specifically about SIP, but I've been doing some H323 hacking lately, and I can talk about that.
How is some-one else, without an existing connection to my machine, supposed to connect specifically to one of my home computers? They can connect to my IP address, but thats my modem/router.
The only way to do it is sit a H323 gatekeeper/proxy on my DMZ. The H323 client thats out there then asks the Gatekeeper to forward the data to the machine named (for example) Larry*. This means I have to have a machine sitting there just to make H323 work right. If this doesn't seem like a work around to you...
I'm pretty sure SIP supports a gatekeeper function (they do have uses as well, mostly as a directory service). But its a work around, a way to bring seperate IP spaces.
If we were running IPv6, then the outside machine would be able to make a direct connection to Larry.
OK, sure, in cases you wouldn't want this to happen, but you can still firewall and such things. But for people who just brought this internet phone thingie because they can make free/cheap calls where ever they can get online over wireless, IPv6 is the only way to go.
So.. the original point. SIP is probably not, and H323 I can definately say isn't hobbled to not work with NAT. It's because NAT is a bugger of a hack, and the workarounds are an even bigger hack. It'd be kind of like saying why are car engines hobbled not to work on fuel thats 90% water?
As for the disadvantages of IPv6... default lame firewalls allowing too much access to LAN machines, giving internet worms more targets (although the targets are now hiding in a much larger haystack).
I think what we have here is Gates' inferiority complex towards Linux. He desperately wants to have MS in every market, and can't stand the thought of that open source demon increasingly being utilized
Screen is very useful. I think of it as providing virtual terminals (you know, Ctrl-F1 to Ctrl-F6) but over remote login (usually ssh in my case)
My normal usage is like so: ssh into the box screen -r (resume session, that was created first time I went in after bootup, with screen -S) Ctrl-A Ctrl-N (next virtual terminal) to flick through things i have running, usually centericq (text multi protocol chat client) and btdownloadcurses, or the results of a compile I left going. Ctrl-A Ctrl-C (open new virtual terminal) which gives you another command prompt. Here I'll wget a tarball, extract it, start the compile, then Ctrl-A Ctrl-N back to centericq
I'll disconnect from the box (Ctrl-A Ctrl-D to disconnect from screen, then Ctrl-D to close the login shell), go home, reconnect from one of my other machines, go through the same process.
Also, I can stay connected, go to another machine, and use screen -x to have multiple connections to the same screen. In the case of a dropped connection, and screen thinks my old login is still valid, screen -rD will disconnect the old login.
Oh, and Ctrl-A ? will show all the in-screen escape sequences. Some of them I haven't used (there is one to do horiz split display, but it crashes btdownloadcurses), but look like they could be even more useful!
But really, the best part is not having programs killed on me because the internet connection dropped (as happens all too often around these parts).
Excuse me, but LILO allows the same thing. I think I currently have 4 kernels (plus WinXP) in my LILO config because I never get around to taking the old ones out of the list.
Ah, that's the thing I was missing to make it clear, when I moved the USB modules into the kernel image the bootup took longer.
And anything worth putting a RAID controller in to should probably be on all the time, yeah. If it's not, is it really worth a RAID controller? But I do know about the long scan times, a person I know insists on having extra drives and between the on-mobo SATA and the add-in IDE expansion card, that machine takes forever to boot. (The expansion IDE card get's ID'd as a RAID card by the BIOS (based on it's PCI ID I assume).)
I think (and I'm guessing, not a guru either, just some-one who guesses a lot) that Linux kernel doesn't perform any device driver init for a device until it is loaded, and it is not loaded until it is needed. By having the modules compiled in, they are all initilized during the startup, rather then on first use (such as when hotplug does hardware detection).
Re the modules being loaded on use: On a different machine, I just recently upgraded the kernel, and forgot to make modules_install (This machine loads almost everything as modules). It booted up right fast, but nothing much worked, naturally. Including the network card. Once I'd realised what I'd done, I did the make modules_install, then restarted the hotplug daemon. Voila, all the hardware went up (coldplug), and only thing remaining to do was restart the services that failed due to missing eth0.
I think this would be similar to how later versions of Windows "boot" faster, by moving some hardware detection off until the GUI is already loaded.
I'd try this when I got home, but by then I'll have forgotten about it again. I don't leave the machine in question on, it's my home desktop machine and sometimes goes days without being used (gasp!)
I find the same thing. My kernel takes much longer (well, twice as long) as windows to boot up. This is mostly because I've compiled all my device drivers into the kernel, so everything gets detected during this stage. I suspect (but haven't bothered to find out) that if I had all the not-immediately-needed drivers as modules, and ran hotplug in the background rather then in the foreground, then everything would start up faster. Maybe, anyway.
Fair bit of angst over there in the PSP scene it seems.
That said, I hope this works out to encourage homebrew development, although will Sony release an update that fixes this buffer overflow along with, oh, something that people -need-, like maybe updated online verification code or something.
I'd say it's because they have all their SEED technology in place, and don't want to replace that. Especially since it currently works. Producing a XPCOM based plug-in for Mozilla based browsers lets them connect to SEED encrypted connections, without replacing infrastructure.
I was connected to it, just got disconnected, timed out a reconnect, tried again, and got 403: Forbidden. Have they just locked down because of the pre-announcement?
Wait till tomorrow for it to really launch, I guess.
I was wondering, when you say you have xscreensaver as your desktop background, is that under your icons? I've been able to get a similar effect by killing kdesktop and running an xscreensaver with --root, but then I loose both my desktop icons (not a biggie but still) and I think Alt-F2 to Run...
I wasn't using 3.4 though, I might try again. Especially if I use crystalgl instead of kompmgr, as my NVIDIA doesn't always correctly handle transperency properly, with the "beta" RenderAccel (and the setting to still do OpenGL with Composite enabled) turned on.
As every-one else is saying, Visual Assist X is the bomb. One of the things I like that no-one else has mentioned is much improved syntax highlighting, including in the tool-tips and other places.
Slashdot Mirror
What I was going to say. Bravo.
Your talking about eardrum carnage, at a teen concert?
Why would they bother using a device, just use the concerts sound system to carry the sound. The device itself isn't anything special.
However, when you moved your VOIP, you'd have a new IP number, yes? And if that IP number was an IPv6 number, then the number would (possibly?) have a geographical correlation. Or at least an ISP correlation.
I'm talking about ways of making it work even when moving, your talking about how it needs to be movable. You did read my post, right?
I'm wondering if the heirachy structure of the IPv6 address could be used here. I don't know how IP roaming works (apparently a feature in IPv6) but could the header of the address be used to track you to your access point, and if the ISP was required to have access point addresses, then they could use those? OK, already I see a problem of shipping address versus billing address, and it would require the ISP's have open access logs, but don't you need them already to save you all from terrorists?
Just a thought I didn't see come up.
In my experience, fonts in Linux are in points, not pixels, unless its because I properly measured my monitor for my xorg.conf file.
Depending on where you are, this may vary, but I understand that "Caller ID blocking" is just raising a flag on the Caller ID signal, that says "here is their ID, but please don't display it". I beleive that call centers, especially ones with reversed charge numbers (free calls) are permitted to have phones which ignore the Caller ID blocking flag.
I wouldn't say they were the first to use it, but certainly the first of such note.
First case I saw of some-one using autorun to screw with a system was when I borrowed a backup CD off a friend, and its autorun would open a very noticable JPG image. Since I was borrowing the CD to update something on my mothers machine... fortunately, she wasn't in the room at the time!
I believe it can also be bypassed by holding down the shift key while inserting the CD into the drive (temporarily disabling AutoRun), or by permanately disabling AutoRun.
Using a bit of tap to do it is just grandstanding.
On the webserver in my basement, of course!
Excuse me, but you are not correct. I don't know specifically about SIP, but I've been doing some H323 hacking lately, and I can talk about that.
... default lame firewalls allowing too much access to LAN machines, giving internet worms more targets (although the targets are now hiding in a much larger haystack).
How is some-one else, without an existing connection to my machine, supposed to connect specifically to one of my home computers? They can connect to my IP address, but thats my modem/router.
The only way to do it is sit a H323 gatekeeper/proxy on my DMZ. The H323 client thats out there then asks the Gatekeeper to forward the data to the machine named (for example) Larry*. This means I have to have a machine sitting there just to make H323 work right. If this doesn't seem like a work around to you...
I'm pretty sure SIP supports a gatekeeper function (they do have uses as well, mostly as a directory service). But its a work around, a way to bring seperate IP spaces.
If we were running IPv6, then the outside machine would be able to make a direct connection to Larry.
OK, sure, in cases you wouldn't want this to happen, but you can still firewall and such things. But for people who just brought this internet phone thingie because they can make free/cheap calls where ever they can get online over wireless, IPv6 is the only way to go.
So.. the original point. SIP is probably not, and H323 I can definately say isn't hobbled to not work with NAT. It's because NAT is a bugger of a hack, and the workarounds are an even bigger hack. It'd be kind of like saying why are car engines hobbled not to work on fuel thats 90% water?
As for the disadvantages of IPv6
* I do not have a machine named Larry.
Your ideas are intriguing to me and I wish to subscribe to your newsletter.
Screen is very useful. I think of it as providing virtual terminals (you know, Ctrl-F1 to Ctrl-F6) but over remote login (usually ssh in my case)
My normal usage is like so:
ssh into the box
screen -r (resume session, that was created first time I went in after bootup, with screen -S)
Ctrl-A Ctrl-N (next virtual terminal) to flick through things i have running, usually centericq (text multi protocol chat client) and btdownloadcurses, or the results of a compile I left going.
Ctrl-A Ctrl-C (open new virtual terminal) which gives you another command prompt. Here I'll wget a tarball, extract it, start the compile, then Ctrl-A Ctrl-N back to centericq
I'll disconnect from the box (Ctrl-A Ctrl-D to disconnect from screen, then Ctrl-D to close the login shell), go home, reconnect from one of my other machines, go through the same process.
Also, I can stay connected, go to another machine, and use screen -x to have multiple connections to the same screen. In the case of a dropped connection, and screen thinks my old login is still valid, screen -rD will disconnect the old login.
Oh, and Ctrl-A ? will show all the in-screen escape sequences. Some of them I haven't used (there is one to do horiz split display, but it crashes btdownloadcurses), but look like they could be even more useful!
But really, the best part is not having programs killed on me because the internet connection dropped (as happens all too often around these parts).
Excuse me, but LILO allows the same thing. I think I currently have 4 kernels (plus WinXP) in my LILO config because I never get around to taking the old ones out of the list.
Ah, that's the thing I was missing to make it clear, when I moved the USB modules into the kernel image the bootup took longer.
And anything worth putting a RAID controller in to should probably be on all the time, yeah. If it's not, is it really worth a RAID controller? But I do know about the long scan times, a person I know insists on having extra drives and between the on-mobo SATA and the add-in IDE expansion card, that machine takes forever to boot. (The expansion IDE card get's ID'd as a RAID card by the BIOS (based on it's PCI ID I assume).)
I don't think I'm using an initrd.
I think (and I'm guessing, not a guru either, just some-one who guesses a lot) that Linux kernel doesn't perform any device driver init for a device until it is loaded, and it is not loaded until it is needed. By having the modules compiled in, they are all initilized during the startup, rather then on first use (such as when hotplug does hardware detection).
Re the modules being loaded on use: On a different machine, I just recently upgraded the kernel, and forgot to make modules_install (This machine loads almost everything as modules). It booted up right fast, but nothing much worked, naturally. Including the network card. Once I'd realised what I'd done, I did the make modules_install, then restarted the hotplug daemon. Voila, all the hardware went up (coldplug), and only thing remaining to do was restart the services that failed due to missing eth0.
I think this would be similar to how later versions of Windows "boot" faster, by moving some hardware detection off until the GUI is already loaded.
I'd try this when I got home, but by then I'll have forgotten about it again. I don't leave the machine in question on, it's my home desktop machine and sometimes goes days without being used (gasp!)
I find the same thing. My kernel takes much longer (well, twice as long) as windows to boot up. This is mostly because I've compiled all my device drivers into the kernel, so everything gets detected during this stage. I suspect (but haven't bothered to find out) that if I had all the not-immediately-needed drivers as modules, and ran hotplug in the background rather then in the foreground, then everything would start up faster.
Maybe, anyway.
This is modded funny, but should be considered insightful.
Fair bit of angst over there in the PSP scene it seems.
That said, I hope this works out to encourage homebrew development, although will Sony release an update that fixes this buffer overflow along with, oh, something that people -need-, like maybe updated online verification code or something.
ROFLOL
I'd say it's because they have all their SEED technology in place, and don't want to replace that. Especially since it currently works. Producing a XPCOM based plug-in for Mozilla based browsers lets them connect to SEED encrypted connections, without replacing infrastructure.
I was connected to it, just got disconnected, timed out a reconnect, tried again, and got 403: Forbidden. Have they just locked down because of the pre-announcement?
Wait till tomorrow for it to really launch, I guess.
After reading one of the Symantec links, I noticed it appears to remove/disable a number of ad-ware programs.
So, you can tell if your computer is infected by it starts running cleaner!
And does this explain this virus writer's rationale? The end justifies the means?
I was wondering, when you say you have xscreensaver as your desktop background, is that under your icons? I've been able to get a similar effect by killing kdesktop and running an xscreensaver with --root, but then I loose both my desktop icons (not a biggie but still) and I think Alt-F2 to Run...
I wasn't using 3.4 though, I might try again. Especially if I use crystalgl instead of kompmgr, as my NVIDIA doesn't always correctly handle transperency properly, with the "beta" RenderAccel (and the setting to still do OpenGL with Composite enabled) turned on.
As every-one else is saying, Visual Assist X is the bomb. One of the things I like that no-one else has mentioned is much improved syntax highlighting, including in the tool-tips and other places.