A better form of Caller ID
For years I have wanted the ability to know who is calling, no mater what phone the person is calling from. I think that it should be fairly easy to implement such a feature -- how hard could it be to add some simple header information to all incoming calls?
You could use the header information to automatically prioritize your calls like you can with e-mail. This would be much better then simply screening your calls, it would allow you to set up complex conditional modifiers (e.g. if my sister calls me with a normal priority call, and I'm on the phone with my wife, send her to my voice mail, if she uses high priority, pass it through to the call waiting feature).
Just adding a simple 5 digit header would let you set up codes for different people, groups and priorities. Of course, you would need something a little more complex then the caller ID boxes they use now (maybe you could run it through your computer).
This would certainly give us more control of our communication. Now if only I could get the telephone companies to offer this feature...
You could use the header information to automatically prioritize your calls like you can with e-mail. This would be much better then simply screening your calls, it would allow you to set up complex conditional modifiers (e.g. if my sister calls me with a normal priority call, and I'm on the phone with my wife, send her to my voice mail, if she uses high priority, pass it through to the call waiting feature).
Just adding a simple 5 digit header would let you set up codes for different people, groups and priorities. Of course, you would need something a little more complex then the caller ID boxes they use now (maybe you could run it through your computer).
This would certainly give us more control of our communication. Now if only I could get the telephone companies to offer this feature...
Comments
Originally posted by Splinemodel
packet based communications can be hard to implement over generally analog networks.
They already pass on the phone number for caller ID, would it be a lot harder to pass on a, let's say a 6 digit number entered right before or after the phone number is dialed? Maybe use the # key to separate the phone number form the header (like #xxxxxx# 1-xxx-xxx-xxxx).
Maybe it is to hard, but it would certainly make my life easier.
(feel free to jump in if you have any ideas)
If I wrote something like this...
[PU] //answer
CID= &GV input string
[##] -pause
[Play directions]
[##] -pause
[Tone]
[Rec] //record a message
//compare sample to demos
[Ck = code 1]
+flaga
[Ck = code 2]
+flagb
[Ck = code 3]
+flagc
[Ck = code 4
+flagd
[OTS "CID" + "flagx"] //display
This is my suedo code... tell me your ideas. I see one problem right off... they will have to be .aiff files which means a second or two or 'thinking time'. Also if people don't press the buttons in the same spacing as the sample it won't work. So I may have to have multiple versions of the same sample to allow for variance in button pressing techniques.
Originally posted by Kickaha
Not to be dense or anything, but couldn't you just check the frequency of the tone coming across for what button was pressed instead of comparing it to an *audio* file??
Problem is with most calls those tones aren't standardized enough. You can get a digital cell phone calling a crummy land line and they won't match. (that is even assuming I could figure out how to do it) Essentially I'd need to re-invent the wheel by writing a frequency finder (which I have no idea how to do). Also the ancient 28.8 modems which I can program have really bad distortion because of abuse...
EDIT: Hey! I found a program on the internet which a Japanese radio station student wrote. Problem: It's all in Assembly and the docs are in Japanese.... >>arg>> I'm gonna email the author and see if he knows some english.... Thanks Kickaha for the advice. If I can use his program (after recoding it of course) then I might be able to do it after all.