jleblanc: redial

12.13.06: Final version (for now!)
Here is the link to my final project: SpokenLetters

Documentation coming...

11.21.06: Expanded Functionality
This new dialplan enables people to generate call files out and to target which sound file The recipeint will hear. It does this in a not very robust way, but I am waiting for Mysql to really get some good stuff going.
Here is the updated dialplan, and the php gencall file.

11.19.06: Some Issues Addressed
I mainly fixed the hangup issue on the record system, and also added a feature where outgoing calls are recorded with MixMoniter and then posted to the podcast too.
Here is the updated dialplan.

11.15.06: La Voz Humana
))) intitial thoughts
A smidgeon more information is here but I'd just listen to the audio above.

11.13.06: Final Idea: La Voz Humana
I realized after much thinking about phones and voice, that what I intuitively want to do is to strip out the interactivity from phones. Phones were arguably the first, and for a long time, the sole interactive technology widely used. Today no one writes letters anymore (few of us at least), and writing heart-felt things in email is awkward and always seems a little cheap. I am intrigued by the idea of sending longer voice letters. I recall in Apocalypse Now how Mr. Clean gets a cassette tape from his mother where she talks to him about home. I wonder if that was common then. The problem I have with writing and email is that you can edit, you can always look back and see what you've created. I think that in some ways this disrupts the flow of ideas for certain applications. The rhythm and cadence of our voices also contain a wealth of information (even more so than handwriting).
I think about how I lie in bed and talk to myself (which is how I thought of this), and I want to capture that but not with all the inconvenience of a tape recorder. If I use the phone I can pipe it straight into asterisk and from there onto the web and anywhere else.

Basically I think it is interesting how we went from email to IM, but how with voice today, person-person communications are almost always direct. But often I like to leave voicemails, and I even like to hear them. I think that there is a kind of communication which will work best when you are just recording your own voice, either for you or someone else, and then you listen to it later.
I would love to send a message to dan snow, but for some reason email is not good. I want to add the flavor of my voice. But in his case I cannot call him, as he is in Japan (and would I want to talk directly to him? sure, but I would also like to just send him something to listen to).

The kinds of stories you hear on this American Life, or old recordings of your grandmother. Somehow there is a warmth there that no amount of words can match. If a picture is worth a thousand words, maybe a sound is worth a million words and a thousand pictures? No, but sound does have a warmth that neither words nor really even images can capture. Sound leaves more to the imagination, but gives you a strong sense of reality at the same time.

I think I like this idea. It has a slow simplicity that I really enjoy. It is about bringing out warmth, and importantly, enabling anyone to do this, even people without computers.
Importantly though this is personal, it is not public like a blog or fast like a chat. It is more like the ancient art of letter writing. And so it is far less about any technology or implementation, but is more an idea of communication, a slower way to connect more deeply.

Even more important is that I think this would fill a gap that I feel in my life and my ability to communicate with people. The technology to do this has been around since tape recorders and answering machines. It is not about technology. It is about using it to reclaim a different pace of communication.

11.6.06: Stop being Lame
I like a fair number of the aspects I discuss below, and yet at the same time I feel like I should be getting a little more wild. But what does wild mean?
The thing I worry about is doing something hugely ambitious from the technical side of things, but giving it no way to have traction. In some sense the it is really about building a system that I think my friends and family would use.
So perhaps I should pose the question to them: What else do you wish your phone could do?
It should have no caveats. Because I am curious about really crazy out there ideas. I mean fuck it.

11.5.06: Thoughts on the Final Project
So I am a little bit unsure of what I want to do for the final. I have had some ideas which I will list here. I was very interested in the idea of letting people create their own dialplans from a simple language on the web. They could post messages record things, bring in rss feeds and create podcasts.

I feel like an open platform system would be very powerful. But I am not sure. I feel like I should identify a user base or some specific audience before hand to help focus my work on this.
Who would want to use this system and why? What features would be helpful? How would this be different from other services out there now?

I think I want to talk to Shawn about this, so that I can get some conviction about the idea and some sense of direction.

I looked at pheeder. It strikes me that many of these applications are pretty one to one.
* Let me think what would I want for myself? What are things that would like to be able to do? I would like to be able to look at a TNO invite and then call a number and leave myself an easily recoverable message so that I can retrieve location information easily.
* Likewise, it would be nice to be able to dump text into a file and be able to retrieve it again from the phone, having the phone read it back to me.
* Related to that is the idea of having rss feeds read to you. That seems a little dense though, and not very scalable at the moment. There might be some smaller rss feeds that would work. Recall the number that David Shulman was talking about where you could call and ask for various pieces of information.
* I should probably talk to Angela more about what her friend would like. Audio blogging is a big deal. I don't care for it so much. To me it is more about augmenting your memory if you don't have a PDA.
* Other pieces that could be brought in are remote address books. These exsist but the idea is unifying these sort of things into a central system. I should be clear here that what I mean is automatic telephone forwarding
* I should also look at the sites that Anne Hong mentioned to me...

The other side of things is what kind of technical challenges I am interested in.
* Learning MySql
* Reversing the podcast
* Doing voip overseas to Taha or Dan Snow
* setting up an actual server
* Writing my dialplan from a browser and then directly uploading that into my dial plan
* * Note the huge security issues here. They are huge potentially

11.4.06: Playing with Sphinx
I tried out sphinx following Shawn's example from week 9.
It was a little tricky to be honest. There were as you can imagine the typical permissions problems. Also, for some reason the perl script does not write out what it deciphers to terminal when called from the command line. Why this is the case, I'm not sure.
In addition I did get these types of commands to do things:

. >>perl sphinxtest.pl > outputfile.txt : Which will overwrite outputfile.txt with the result
. >>perl sphinxtest.pl >> outputfile.txt : If you want to append each time

I won't get into posting the various files here, but one note I f'd up is the Monitor outputs two files which are Your-filename + either -in or -out for which side. The actual filename you pass to Monitor contains nothing.

I had wanted to create a text input system where you could spell out the letters in a word. This is not going very well. Letters like 't' 'p' 'd', they just sound far to similar. There is no doubt that this system works much better with a small set of 2-3 word phrases. Plus this current method is absurdly slow.

11.1.06: Php, Festival + Sable
Basically I just wanted you to be able to input sable code and text from a webpage and then be able to have you hear that text spoken when you called a site

* Here is the dialplan which via extension 4 just calls a system try to convert the sable file to a wav file.
* The test.sable file lives on my web directory on social.
* On my public social site, the index page displayed a text box and submit button. When someone clicked the submit button the text was posted to input.php
* This php script first copied the sable header (sablefront.txt) into the sable file, and then it copied in the text contents.

10.23.06: Explaining the midterm
For the midterm I looked at pushing messages recorded within asterisk out to a podcast. So here is my initial documentation on what I did:

* The core part was the dialplan which copied .wav and .mp3 versions of the messages out to my public directory on stage.
* The dialplan called rss_gen.php which was located in my home directory on social to edit the rss file for the podcast.
* On my public social site songs.php displayed the .mp3 and .wav files.

Some of the tricky parts of this were how to set permissions correctly. I encountered problems because things that I could do from the command line (such as using cp) couldn't be done by asterisk because folder permissions were set to restrictively.
A problem this system suffers from is that I rely on the user to press pound after recording their message, because if they don't, then the dialplan commands and php scripts that copy the sounds files to public and update the rss feed aren't called. A fix should be made for this.

Something else I tried was making a better file navigation system in php. It hasn't worked yet, because I have been having problems with getting anything from the database_get application.

10.25.06	animal telephony
10.15.06	song recording system (dynamic I)
10.10.06	some ideas
10.3.06	messing around with some php + agi
9.26.06	branch menu
9.20.06	simple voicemail simple song menu with vm
9.18.06	untested code with voicemail
9.12.06	first asterisk experiment