I've finally got Bluetooth working in linux. For a while I was just looking at bluez and I couldn't figure out how I was actually supposed to transfer files. I could connect, the pin was right, but I didn't know how to send or receive files.
I installed gnome-bluetooth-manager and I still couldn't figure it out. All that did was see the phone. I couldn't do anything to the phone connection or find a way to transfer files.
Finally though the kdebluetooth RPMs finally made it to my urpmi server and after i did urpmi kdebluetooth I'm now able to transfer files. It took a while to figure out which programs to use (didn't realize that they'd be in the icewm menus too, I'm a moron). But finally found kbtobexclient and was able to transfer a demo java game (J2ME with the wireless toolkit makes development easier, not easy [yet, but i haven't worked on it much yet except to look at the demos], but easier) and actually run it.
Wednesday, September 28, 2005
Moving Tech blog entries here
My main blog is at Bounding Overwatch. I've created this new blog as my tech blog though (struggling very hard to avoid using some amusing yet faux latin for the blog name) and I've copied most of the tech related blog entries from the old blog to here.
The entries are still there too though. Didn't want links to break if people are linking to me there or search through google.
I wonder what the result of all this moving will be for jijo though. His Filipino Tech blog aggregator picks up posts via RSS. But if I posted all of these entries this morning (except I used the original datetimes), will his aggregator be suddenly inundated with all my entries? Or will only this first entry make it?
It's probably the second, but it'll be interesting if it's the first. Hahaha.
The entries are still there too though. Didn't want links to break if people are linking to me there or search through google.
I wonder what the result of all this moving will be for jijo though. His Filipino Tech blog aggregator picks up posts via RSS. But if I posted all of these entries this morning (except I used the original datetimes), will his aggregator be suddenly inundated with all my entries? Or will only this first entry make it?
It's probably the second, but it'll be interesting if it's the first. Hahaha.
Thursday, September 08, 2005
Dynamic DNS in Linux
That title is a bit over the top, of course. I only looked at one solution, and when that worked for me, I stopped looking.
I've got a sort of static IP. My internet provider, Destiny Cable Internet, as of this writing, uses the ISC DHCP Server. That server tries to allocate the same IP to the same requesting MAC if the IP is still available. So the IP is pretty stable. However, there is no promise that the IP will never change, so it's semi-stable.
I thought I'd be OK with just assuming that it was stable, but I changed my mind. It *might* change and when it does, then I'll have to go to everywhere that assumes it's stable and fix it. And then it might change again.
Instead I decided to use a dynamic DNS server. For the client software, I first looked at EZ IP-Update and, since there's a list of dynamic dns servers at that site, I chose DynDNS. Both choices were pretty random, but I'm happy with both of them.
ez-ipupdate took around 20 minutes to set up (there was a minor compile error, something wrong in the ./configure script's deciding how to use errno, so I had to figure that out and fix it, it's a quick and dirty hack though, not good enough to propagate upward, but I may email the ez-ipupdate maintainer so that he can look at his configure configuration) and DynDns took 2 minutes (plus the less than a minute to wait for the email to arrive).
So I've got dynamic dns working now. bopolissimus.homelinux.net
I've got a sort of static IP. My internet provider, Destiny Cable Internet, as of this writing, uses the ISC DHCP Server. That server tries to allocate the same IP to the same requesting MAC if the IP is still available. So the IP is pretty stable. However, there is no promise that the IP will never change, so it's semi-stable.
I thought I'd be OK with just assuming that it was stable, but I changed my mind. It *might* change and when it does, then I'll have to go to everywhere that assumes it's stable and fix it. And then it might change again.
Instead I decided to use a dynamic DNS server. For the client software, I first looked at EZ IP-Update and, since there's a list of dynamic dns servers at that site, I chose DynDNS. Both choices were pretty random, but I'm happy with both of them.
ez-ipupdate took around 20 minutes to set up (there was a minor compile error, something wrong in the ./configure script's deciding how to use errno, so I had to figure that out and fix it, it's a quick and dirty hack though, not good enough to propagate upward, but I may email the ez-ipupdate maintainer so that he can look at his configure configuration) and DynDns took 2 minutes (plus the less than a minute to wait for the email to arrive).
So I've got dynamic dns working now. bopolissimus.homelinux.net
Port forwarding with ssh
I find forwarding ports (and, actually, also reverse forwarding ports) over ssh very useful. I'll probably find it less useful now since I've found OpenVPN so easy to setup and use, but it's still a very useful technique, particularly where I don't have root access.
I always use -v when doing this because -v will tell me if the link is slow or the link is down (while i'm still connecting, less wasted time). It also shows a message everytime a request is forwarded down the ssh tunnel, useful for debugging to make sure I'm doing the right thing.
With the setups below, surfing to http://localhost:8080 will forward requests to the internal or external server specified. Note: by default, port forwarding binds only to 127.0.0.1, so from the box itself you can connect to the forwarded service, but you can't from another box. Use GatewayPorts for exposing the service (warning, think about that, it might open security holes).
1. I'm on my laptop and I want to surf the website on remote.com.ph as if I were physically there. So what I want to do is forward my port 8080 to remote.com.ph:80 (it's an internal website and is inside a firewall, so I can't surf to that site directly from the open internet).
ssh -v -L 8080:localhost:80 remote.com.ph
what that does is, connections on my local port 8080 will be forwarded to remote.com.ph. From there, it will be forwarded to remote.com.ph's localhost:80.
2. Same as #1, I can ssh to remote.com.ph but the web server isn't on remote.com.ph, it's on another internal box inside the firewall, 192.168.80.80:80.
ssh -v -L 8080:192.168.80.80:80 remote.com.ph
What that does is, connections on my local port 8080 will be forwarded to remote.com.ph over the ssh tunnel. Remote.com.ph will then forward them onward to 192.168.80.80:80
3. Same as #2, but I'm trying to surf to some external web site that doesn't let me in if I surf from the Philippines, or if I surf there from the Philippines something bad will happen. For instance, PayPal will block paypal accounts if they're used from IPs that it identifies as being in the Philippines. This has to do with very high fraud rates. But maybe I'm not a fraudster, I just want to use my paypal account, but I can't because I'm physically in the Philippines. If I have ssh access to a host in the USA which paypal won't be suspicious about, I can do port forwarding through that server, e.g.,
ssh -v -L 8080:www.paypal.com:80 my_us_server.com
There is also reverse port forwarding. I do that when the server I need to connect to (usually not for http, but for some internal server) is inside a firewall and I can't get to it directly and the firewall won't port forward to the internal server. In that case, what I do is ssh to the gateway, and from there ssh to the internal server. Then I do reverse port forwarding. What that does is, it will ssh back to *me*, and open a localport on *me* which will be forwarded down that second ssh link back to it. This, of course, only works if the internal server has a route out to the internet, if it doesn't, then some other solution will have to be found.
# first ssh to the gateway
ssh gateway.remote.com.ph
# at the gateway, ssh to the internal box
ssh my-internal
# at my-internal, open the reverse tunnel back to me, i am, client.com.ph
ssh -v -R 8022:localhost:22 client.com.ph
What that does is get my-internal to ssh to client.com.ph (my box, outside the firewall). Once it gets there, it will set up port forwarding so that port 8022 at client.com.ph so that when I (at client.com.ph) connect to port 8022, the request will be forwarded to port 22 at the my-internal computer. I use that much less often than -L, but when -L doesn't work, -R is often a lifesaver.
I always use -v when doing this because -v will tell me if the link is slow or the link is down (while i'm still connecting, less wasted time). It also shows a message everytime a request is forwarded down the ssh tunnel, useful for debugging to make sure I'm doing the right thing.
With the setups below, surfing to http://localhost:8080 will forward requests to the internal or external server specified. Note: by default, port forwarding binds only to 127.0.0.1, so from the box itself you can connect to the forwarded service, but you can't from another box. Use GatewayPorts for exposing the service (warning, think about that, it might open security holes).
1. I'm on my laptop and I want to surf the website on remote.com.ph as if I were physically there. So what I want to do is forward my port 8080 to remote.com.ph:80 (it's an internal website and is inside a firewall, so I can't surf to that site directly from the open internet).
ssh -v -L 8080:localhost:80 remote.com.ph
what that does is, connections on my local port 8080 will be forwarded to remote.com.ph. From there, it will be forwarded to remote.com.ph's localhost:80.
2. Same as #1, I can ssh to remote.com.ph but the web server isn't on remote.com.ph, it's on another internal box inside the firewall, 192.168.80.80:80.
ssh -v -L 8080:192.168.80.80:80 remote.com.ph
What that does is, connections on my local port 8080 will be forwarded to remote.com.ph over the ssh tunnel. Remote.com.ph will then forward them onward to 192.168.80.80:80
3. Same as #2, but I'm trying to surf to some external web site that doesn't let me in if I surf from the Philippines, or if I surf there from the Philippines something bad will happen. For instance, PayPal will block paypal accounts if they're used from IPs that it identifies as being in the Philippines. This has to do with very high fraud rates. But maybe I'm not a fraudster, I just want to use my paypal account, but I can't because I'm physically in the Philippines. If I have ssh access to a host in the USA which paypal won't be suspicious about, I can do port forwarding through that server, e.g.,
ssh -v -L 8080:www.paypal.com:80 my_us_server.com
There is also reverse port forwarding. I do that when the server I need to connect to (usually not for http, but for some internal server) is inside a firewall and I can't get to it directly and the firewall won't port forward to the internal server. In that case, what I do is ssh to the gateway, and from there ssh to the internal server. Then I do reverse port forwarding. What that does is, it will ssh back to *me*, and open a localport on *me* which will be forwarded down that second ssh link back to it. This, of course, only works if the internal server has a route out to the internet, if it doesn't, then some other solution will have to be found.
# first ssh to the gateway
ssh gateway.remote.com.ph
# at the gateway, ssh to the internal box
ssh my-internal
# at my-internal, open the reverse tunnel back to me, i am, client.com.ph
ssh -v -R 8022:localhost:22 client.com.ph
What that does is get my-internal to ssh to client.com.ph (my box, outside the firewall). Once it gets there, it will set up port forwarding so that port 8022 at client.com.ph so that when I (at client.com.ph) connect to port 8022, the request will be forwarded to port 22 at the my-internal computer. I use that much less often than -L, but when -L doesn't work, -R is often a lifesaver.
Wednesday, September 07, 2005
Comparisons of different SQL implementations
Online there's a pretty long (not sure how comprehensive, but informative and useful) Comparison of different SQL Implementations
Haven't read it all yet (busy), but I'll need to get to that within 1-2 days.
Haven't read it all yet (busy), but I'll need to get to that within 1-2 days.
Saturday, September 03, 2005
Destiny Cable Internet SMS number
To get Destiny Cable Internet to call me (in case of technical problems),SMS to 09175771111 MD [account name] [account phone number]
Very convenient since I have a cell phone and cell phone charges being exorbitant in the Philippines, I'd rather not have to call them. Besides, I don't know what the tech support number is and I'm too lazy to look :-).
It's weird that one has to put one's phone number in there, but I guess that's for people who have landlines. They can put their landline number there. I don't have a landline so it seems a waste to type the number in since they should just call me on my cel and they've already got that when they receive my text.
It's nice to be able to add (after the phone number) some text about the particular problem. So they can look at things on their side (e.g., signal level, problems in my area, etc) before even calling me. Saves me time explaining things, and saves them some telco costs.
Very convenient since I have a cell phone and cell phone charges being exorbitant in the Philippines, I'd rather not have to call them. Besides, I don't know what the tech support number is and I'm too lazy to look :-).
It's weird that one has to put one's phone number in there, but I guess that's for people who have landlines. They can put their landline number there. I don't have a landline so it seems a waste to type the number in since they should just call me on my cel and they've already got that when they receive my text.
It's nice to be able to add (after the phone number) some text about the particular problem. So they can look at things on their side (e.g., signal level, problems in my area, etc) before even calling me. Saves me time explaining things, and saves them some telco costs.
Friday, September 02, 2005
Windows Solutions
Long ago I saw a list of "Windows Solutions". I had it in my sigs and now I can't find them. Fortunately, I keep my old sig file around even though I don't use it anymore (I haven't seen a tool yet that will modify the gmail sig via cron).
Anyway, here they are, so I don't forget them, hopefully google will cache them and google can find them with "Windows Solution bopolissimus" :
1. Don't use Windows
2. Stop the program and start it again
3. Ctrl-Alt-Del and kill the program in Task Manager
4. Reboot
5. Uninstall the program and reinstall it.
6. Reinstall Windows and Everything else
7. Backup, format the disk, install windows, reinstall everything else, restore your data.
to be fair, you sometimes need to do all of these (or the transliteration) for other operating systems, and #2, #3, #4 aren't so common anymore in XP. But they're still much more common there than in Linux. All that spaghetti inside is turning into much and Windows is turning more and more into a big pile of mud (if I may be allowed to mix metaphors)
Anyway, here they are, so I don't forget them, hopefully google will cache them and google can find them with "Windows Solution bopolissimus" :
1. Don't use Windows
2. Stop the program and start it again
3. Ctrl-Alt-Del and kill the program in Task Manager
4. Reboot
5. Uninstall the program and reinstall it.
6. Reinstall Windows and Everything else
7. Backup, format the disk, install windows, reinstall everything else, restore your data.
to be fair, you sometimes need to do all of these (or the transliteration) for other operating systems, and #2, #3, #4 aren't so common anymore in XP. But they're still much more common there than in Linux. All that spaghetti inside is turning into much and Windows is turning more and more into a big pile of mud (if I may be allowed to mix metaphors)
Subscribe to:
Posts (Atom)