[p2p-hackers] iGlance is here!

[auto43348 at hushmail.com]

Here is my $0.02 on voice p2p comms:

1.  Well implemented, OpenSource, strong encryption using standard 
algorithms which is ON by default.
2.  Multiplatform/Os including mobile devices.
3.  Firewall traversing.
4.  Infrastrctureless p2p.  Not requiring a central server.  The 
ability to bootstrap on an internal network seperated from the 
internet.
5.  Scaleable and reliable.
6.  Phone, chat, video, PTT, file transfer, SMS, whiteboard, 
ability to integrate with the PSTN.  Integrate with normal 
telephones. Desktop sharing and PTT integration with radios (like 
Dingotel) would be nice.  Security issues with desktop sharing 
would need to be firmly addressed.
7.  It Just Works.  Grandma needs to be able to download, click and 
use.
8.  ATA ability.  Ability to put on a standalone device to create 
an ATA.  Drop on something like one of the linux based routers to 
create an ATA.
9.  Good marketing and PR.  A critical mass is well, um critical. :)
10.  Ability to force relays and increase anonymity if desired 
would be nice.
11.  Effecient use of bandwidth and CPU, with an ability to choose 
low bandwidth/cpu protocols if desired to help with  slower 
computers, limited bandwidth (like mobile wireless) and sat 
connections.
12.  The pie in the sky would the ability to call (in both 
directions) an ATA, softphone, 2way radio, SMS, chat or cell and 
obtain a p2p secure SIP connection using a central 
infrastructureless design with also having the ability to interface 
with the PSTN using private gateways or paid services as needed.

Something like Skype with it's momentum, critical mass, firewall 
traversal, multiplatform, default encryption, It Just Works feel, 
PSTN integration; the 2way ability of Dingotel; OpenSource and 
multiple protocol interfacing like Gaim;  Well implemented, peer 
reviewed strong secuirty like PGP/GPG/ciphire, with an internal 
design which doesn't require a central infrastructure and therefore 
is robust and cannot be shut down (so you don't have problems like 
after Katrina that because the regional switch was down or 
overloaded, New Orleans cell phones being called from NY and 
physically present in Atlanta, don't work.), and works/integrates 
on an inexpensive black box like a Sipura.

I am repeating myself, but that should demonstrate the general 
idea. :)

rearden

>Date: Sun, 6 Nov 2005 20:23:12 -0800
>From: David Barrett <dbarrett at quinthar.com>
>Subject: Re: [p2p-hackers] iGlance is here!
>To: "Peer-to-peer development." <p2p-hackers at zgp.org>
>Message-ID: <1131337414.29029B5D at dg11.dngr.org>
>Content-Type: text/plain; charset="us-ascii"; format="flowed"
>
>On Sun, 6 Nov 2005 8:45 am, Ian Clarke wrote:
>> Its about time someone did this :-)
>
>I thought so too! :)  Push-to-talk is a nice middle-ground, and 
>the 
>video presence is really quite powerful when both sides have 
>cameras.
>
>
>> It seems like you are pretty heavily reliant on windows stuff 
>though,  
>> any plans for Mac or Linux versions?  How portable is the code?
>
>Actually, there are very few windows dependencies and platform 
>portability was a design goal from the outset.  All the file, 
>threads, 
>TLS, sockets, and base stuff currently cross compiles between 
>Windows 
>and Linux (and WinCE for a bit, but I pulled it).  Generally 
>libpig is 
>very thin, and almost all of the platform specific code is one in 
>one 
>file (GNonPortable_Win32.cpp/_Linux.cpp).
>
>The "thickest" part of libpig is the audio/video capture/playback 
>part 
>as that depends heavily on the platform, but the external API is 
>very 
>minimal.
>
>Another somewhat "thick" area is the GUI.  But I've sliced this 
>into two 
>layers:
>- A platform dependent "display RGBA bitmap and return input 
>events" 
>architecture
>- A platform independent, XML-based, HTML-like, skinnable GUI 
>engine
>
>The intent behind libpig is to create thin cross-compile layer, 
>rather 
>than a application design philosophy.  Sorta the difference 
>between 
>OpenGL and DirectX -- I want it to fit into my architecture, and 
>not 
>architect around its design.
>
>My original intent was to just use APR, ACE, Qt, Boost, etc, but I 
>found 
>myself confused by deep polymorphic trees, esoteric programming 
>practices, compilation problems, licensing problems, and so on.  
>Eventually I decided my needs were so minimal that I'd just go my 
>own 
>way, and I don't regret the decision in the least.
>
>
>> I personally use a Mac, but I will give this a try on Virtual 
>PC.
>
>Excellent.  Incidentally, I'm extremely eager to do a Mac port (I 
>spend 
>most my life in coffeeshops -- for which iGlance is idealy suited -
>- and 
>Macs outnumber PCs maybe 2:1).  But I don't know jack about the 
>platform, it doesn't work well in VM ware, and I don't want to lug 
>
>around two laptops.
>
>If anyone is interested in bringing iGlance to the Mac, I am *all 
>ears*.
>
>> All the best,
>
>-david




Concerned about your privacy? Instantly send FREE secure email, no account required
http://www.hushmail.com/send?l=480

Get the best prices on SSL certificates from Hushmail
https://www.hushssl.com?l=485