Jump to content

aloishis89

Active Members
  • Posts

    50
  • Joined

  • Last visited

Posts posted by aloishis89

  1. So I hacked my zipit a while ago and go everything working. I wanted to try out the SideTrack userland, but I'm having trouble getting it to boot. I have the bootloader screen when you turn on the zipit (the weird penguin thing) but I can't get it to go any further. I tried using ubuntu image writer under linux mint to write the img to my 4 gig micro sd card, but no luck. I also tried writing it under windows with Win32DiskImager, which didn't work, so then I tried again and expanded the ext3 partition to fill the entire disk like it's suggested to do on IronGeek's site to no avail. I'm pretty sure it's just a matter of me getting an image written correctly and not that I've somehow bricked my zipit right?

  2. It would be cool to have a web based grooveshark downloader, but I found another one that works well. It's called Free Music Zilla. Basically it does some sniffing for mp3s and then grabs then and lets you download them. I just tested it out on grooveshark and it works great. It can't recover all of the metadata from the song, and you have to play like a second of each song on grooveshark to queue the download, but other than that it's awesome. My current instance of it is maybe 10 minutes old and its taking up about 12,500 K of memory.

    freemusiczilla.com

  3. I've been using Grooveshark (grooveshark.com) for a while now. I like it better than Pandora since it doesn't have any restrictions and you can listen to any song whenever you want. It's so convenient, which makes me think that the RIAA will probably shut it down in a few years. Anyway, I know some work has been done on saving music from Pandora. I know that you can always use a recording program to save any audio that passes through your speakers, but in the techniques I've seen for pandora, they download the actual file. My question is, can you do the same thing with Grooveshark? The only embedded thing I'm seeing in the page is something called "Grooveshark.mpeg" which is a little file that doesn't play on a regular media player. It's got some data in it, I just don't know how to get at it and see if it's interesting.

    So is there a program or technique already out there? Or do we have to come up with one?

  4. yeah, i didn't know if doing it separately would help or not. Here's a thread I found that might be helpful

    :

    http://ubuntuforums.org/showthread.php?t=264614

    Specifically, this step:

    Make Sure dhclient Is Not Confused

    I've read that dhclient can get confused if you switch between using your ethernet interface and your wireless interface. The result is frequent dropped wireless connections or the inability to connect wirelessly.

    You can see what interfaces are currently configured by dhcp by looking in the folder /var/run. In that folder you may see files which indicate that dhclient is currently configured for interfaces. i.e. you may see /var/run/dhclient.eth0.pid

    (1) Bring both the ethernet and wireless interfaces down by using:

    ifconfig YOUR_INTERFACE down

    (2) Kill dhclient to make sure it's not running using:

    killall -9 dhclient.pid

    kill all -9 dhclient.YOUR_INTERFACE.pid

    (3)Also, clean things up by removing the dhclient pid's from /var/run by using:

    rm -f /var/run/dhclient.YOUR_INTERFACE.pid

    (4) Use a text editor to clear ALL text from var/lib/dhcp3/dhclient.leases and /var/lib/dhcp3/dhclient.YOUR_INTERFACE.leases

    (5) By now, dhclient should have been thoroughly killed. Use ifconfig to bring only the wireless interface back up. If you have a network applet in your task bar, you should notice your wireless interface being set up.

    (6) After a minute or two, check /var/log/daemon.log, var/lib/dhcp3/dhclient.leases, and /var/lib/dhcp3/dhclient.YOUR_INTERFACE.leases to see if dhclient is making any progress in getting an ip address from your access point.

    you can use the command line text editor nano to edit those text files if the mouse still isn't working well. If this stuff doesn't work, then I'd suggest asking on the debian forums since they will be able to give you more specific help than I will.

  5. try:

    iwconfig eth1 essid TheNetworkName
    iwconfig eth1 key key_here
    ifconfig eth1 up
    dhclient3 eth1

    or you could try to install wicd network manager with

    apt-get install wicd

    I haven't tried using wicd on the zipit yet, so i'm not sure if it'll work.

  6. it takes a bit. for me it did nothing, then blinked off, then I got the blank bootloader screen again, and then got the login prompt. see if the light on the front is blinking, that means that its reading from the card. total time from cold boot to login is maybe 30 seconds or something.

  7. I think as long as you successfully flashed the bootloader, you are still good. What's really happening (and anyone correct me if I'm wrong please) is that you are telling the Zipit to look on the miniSD card to boot from. It's kinda like a live CD (or booting off a flash drive). I've got all the files from my hack, are the ones you need gpsi8686.bin and gspi8686_hlp.bin?

  8. You sure can, I just did. I went from stock Zipit to hacked Zipit in an afternoon and got ssh, vlc media player, and a couple other things on there. If you do hack it, I suggest using a 2gig microSD card and only using the 1 gig image. The actual data is exactly the same, so with the 1 gig image I still have tons of room to install stuff. I'm working on using TVersity to stream Hulu to the Zipit. Such a cool little device!

  9. Ok, I got that much working. Now my problem is that I need to only have 1 character at any time in the input field. The only way I can think of doing this is by having javascript keep everything in the box always selected and making the maximum input length 1 so that when I type a new command, the old one gets overwritten (since it is selected). That way I don't have to use backspace or anything. So far I have only gotten as far as getting everything to select when you click in the box. I have this:

    input type="text" id="txt1" size="30" maxlength="1" onkeyup="showHint(this.value)" onFocus="this.select()" onBlur="this.select()" onClick="SelectAll('txt1');" value="Click here, then press a key"/>

    I have a couple different things going on in there to try to get it all selected (like the onFocus and the onBlur which don't work). Any idea how I could keep everything in that box permanently selected? Or is there a better way altogether to keep a single character in the box?

  10. I built a version of the hak5 rover and I've been working on the code to drive it online. It works, but right now its a page written in PHP that has several submit buttons that execute different PHP functions, which make the robot move. The webpage is hosted on an onboard laptop, and you drive it by connecting to that onboard server. So even though it works right now, the nature of the submit buttons requires that the page get reloaded each time a command is sent. I want to also implement a failsafe in case the driver gets disconnected. The best way to do this would be with a socket connection (correct me if I'm wrong there). That way if the socket closes, then I can tell the onboard server to send a STOP to everything. I am also under the impression that that means I won't have to reload the page every time. Another reason I want to do this is because clicking buttons is not a very intuitive way to drive something, and I'd like to be able to use the keyboard (WASD). I already have a nice javascript library to do this for me, but the trouble is passing these variables of the states of the keys to the PHP that is controlling the motors. I have done a little research, and heard about this thing called COMET that is supposed to be good for opening a sustained connection between the server and the client. I was also wondering if it is possible to pass the keystroke information from javascript to PHP. Sorry for the length, but here is a sum up of my questions:

    1) Is a socket connection a good idea for controlling a robot over the internet?

    2) Can you pass information from javascript to PHP without having to reload the page?

    3) Would COMET be useful in this situation?

    I've never really used Ajax or anything before, so I'm learning it as I work on this project. Any help in the form of explanations or code snippits or links would be great. Thanks!

  11. I found something that would add that will increase your range immensely. (3-5 kM) And incredibly cheap too.

    It is known as wifry, wokfi, etc. Essentially you use asian cookware to magnify the wifi signal, similar to a satellite dish.

    Info here:

    http://www.usbwifi.orconhosting.net.nz/

    http://www.instructables.com/id/%22Poorman...-cookware-refl/

    http://en.wikipedia.org/wiki/WokFi

    Wow, that's a really good idea. I had thought about making a cantenna, but I didn't think of using something like a wok instead. We will definitely look into that, there's a pretty high chance we'll make one. I am on a BattleBots team, and I also happen to know that a big steel wok makes a great housing for a robot (google for Blendo) and is way cheaper than buying a big steel hemisphere. Asian cookware can be very versatile!

  12. @VoodooTorture:

    We would like to be able to tether a cell phone for internet, but we don't really have the budget for that. Me and 2 of my friends are the ones working on it, and we all go to separate colleges during the school year, so the only time we have to work on it is during breaks or over summer. So paying for an unlimited data plan over an entire year would be kind of wasteful at this point. As for protecting the laptop, some sort of armor would be a good idea. We have speakers on the rover, so we joked that if someone got too curious, we could drive in the bushes and play dog barking audio and scare them off :D. I updated the site with some new writeups for the latest version, but I won't have pictures or video for another couple weeks.

    @stacked:

    Good to know, we'll try it.

  13. After considering building our own, we decided to buy one. Its a little pricey at $125, but its 2 channel, and has a continuous current rating of 25 amps with a 50 amp peak. This is WAY more than we are letting it draw right now, so the rover should go much faster and have much more torque. The new motor controller is a Sabertooth 2x25. We decided to buy one because we knew that it would work off the bat, we could get it soon, and it has features like the ability to recharge the battery when the motors slow down or go in reverse.

    The parts should be coming in next week, and once we get them hooked up and working, I will update the website with pictures and video. I have also been through a couple more version of the code, so I will get that up once we test it.

  14. We are using the stock gearbox. It was designed to be used for ~50 pound kids, so we figured it would suffice. Yeah, let me know about making my own controllers. You mentioned that it would involve programming and if by that you meant writing code, then thats no problem, but if you meant buying a chip programmer, then I may need to wait until I get back at college to use one.

  15. @thedude: Treads will help for driving off road, but right now I am almost positive that it is the current draw that is causing the problem. We connected the motors directly to the battery and it had plenty of torque.

    @Dead_Already: We were actually pretty curious as to how our motor controller worked. I would be interested in building my own if I could. I'm not sure what our current draw is exactly, but I think a 15 amp peak should be enough (its a 3-12v motor running at 18v). Is building my own worth attempting?

  16. You're right, I had thought about doing that but we did some more tests and it looks like the motors we have can put out enough torque, but they are being limited on the amount that they can draw. We bought motor controllers that, at the time, we thought could handle our maximum current draw but they ended up not being enough. Buying higher rated ones is too expensive, so we had to use resistors to limit that. Right now we have 2 ohms of resistance, but we are going to try to bring that down to 1/3 or 1/2 an ohm and that should help some. Really, with the motor controllers we have, there will be a limit to how much current we can squeeze through (unless we spend another $100-120 for new controllers) and we are trying to get as close to that limit as possible.

×
×
  • Create New...