“Hey, that’s a really cool case! Where can I buy one?”

The truth is that we built the case ourselves. There is definitely a strong “modding” ethic within Voxeo (of computers, cars, equipment, pretty much anything…) so building this case was right in line with that part of our culture.

In this three part blog post series, I’m going to show how we built the case and then transported it down to New York for SpeechTEK.

First, though, if you haven’t seen the video of Voxeo CEO Jonathan Taylor talking about the server and Prophecy 10 (and giving a tour of the server), here it is:

As to the name used in the title of this post, we called this 16,000 port system internally “the Beast” and so naturally the case was called “the Beast Cage“.

Now, on to the photos…

We started with an Odyssey FZAR4 4-space rack. This box is normally used to transport high-end audio gear.
	First we modified the rear portion of the case. On the left will be a two-port ethernet connection. On the right is the hole for the power connection.
Here’s a close-up view of the 2-port Ethernet jack plate.
	Here we’re almost done with the Ethernet connection.
We next added two more holes for cooling fans.
	We added two 8mm cooling fans.
Ethernet, cooling fans, and twist-lock power connector are now installed. We paid $8 more for the fans with the blue LED lights.

Stay tuned for Part 2, where we get the system ready to go…

We’re pleased to announce that we are now able to get phone numbers in South Africa for our IVR Hosting platform. We’ve always had the ability to deploy our Voxeo Prophecy on-premise platform in South Africa (and have some very large customers using it there), but now we’re delighted to add the hosting option as well.

If you would like to find out more about setting up a phone number in South Africa, please contact us.

And if South Africa isn’t of interest to you right now, please remember that through our worldwide developer program there are over 40 other countries in which you can obtain phone numbers on our hosting platform (which is free to use in development – sign up now!). For more information about our worldwide developer program, please see our announcement (in March) and the FAQ.

Map image courtesy of Google Maps.

Developers who don’t work with python may not be aware that the new python version 3.x is deliberately NOT backwards-compatible with python 2.x. There are many reasons and the reality is that python 2.x will be in usage for many years – even as all the new development happens with 3.x.

So the quick answer is that the main branch of the python library for the Tropo WebAPI is targeted at python 2.5 or later. (Current is 2.7.)

However, for people who want to work with python 3.x, I created a branch in the Github repository called “python3″ that includes a version of the Web API that works with python 3.x. There’s a catch, though:

When you download the code or clone the repo from Github, you only get the ‘master’ branch with the 2.x code.

The “git clone” command just pulls the one main master branch. Similarly, if you download the code directly from the website (i.e. not using ‘git’), you will by default get the 2.x code only.

Here’s how to fix that… for both command-line git and for the web download.

THE WEB WAY

If you download the python module from the Github website, it’s a simple process to get the python3 code. Simply go to the module web page, click on the “Switch Branches” link on the nav bar, and choose “python3″:

Now when you click on the “Download Source” button in the upper right of the screen you will get the python 3.x source code.

THE GIT WAY

For command-line git, the process isn’t too much different – and you can nicely have both the 2.x and 3.x code in your same local directory on your system. Github nicely provides this handy “git cheat sheet” that includes these relevant lines:

TO PULL A NEW BRANCH FROM A REMOTE REPOSITORY

git fetch origin [remote-branch]:[new-local-branch]

So the magic command to get the python 3 branch into your local clone of the repo is to make sure you are in your local repo and type:

git fetch origin python3:python3

You now will have the python3 branch in your local repo – but the code you see will still be the 2.x until you switch to the new branch.

Here’s how the command-line sequence would look:

$ git fetch origin python3:python3
From git@github.com:danyork/tropo-webapi-python
* [new branch] python3 -> python3 $ git branch
* master
python3
$ git checkout python3
Switched to branch "python3"
$

The code in your repo will now be the python 3.x code and you can work with it and try it out with your applications.

To switch back to the 2.x code, just do:

git checkout master

And to go back to the 3.x code, just do:

git checkout python3

Alternatively, you could clone the repo twice into two different directories and have one be the python3 branch all the time. Different people have different strategies.

Over time, if you want to pull down any changes that have been made to the Github repo for the python 3 branch, you can type this (MAKE SURE YOU ARE IN THE ‘python3′ BRANCH IN YOUR LOCAL REPO):

git pull origin python3

That’s it.

MOVING FORWARD WITH PYTHON 3.X CODE

I created the “python3″ branch because I had a personal interest in experimenting with python 3.x for a Tropo app. I plan to do my best to keep it in sync with any changes and improvements that are made to the 2.x track… but any help is welcome! If you work with python 3 and find ways to make the Tropo module work better (or more “Python 3-ish”) please do contact me either via Github (I’m “danyork“) or via email.

P.S. If you are a regular Github user, feel free to fork http://github.com/tropo/tropo-webapi-python, make your changes in your fork and then send a pull request. (And if that last sentence made no sense whatsoever to you, don’t worry… it’s all good.)

Via a team effort spanning multiple time zones, we not only wrote all new examples for Python and Groovy, we also tested every single example in every language to ensure they are as complete, functional and logical as possible.  We’re proud of the results, and hope they make working and playing in Tropo even easier than it was before.

Now when you look through the Scripting Guide, you’ll see not one, not two, not three code examples for every demonstrated functionality, you’ll find FIVE:

But that’s not all.  In addition to the examples within the Guide itself, there are also five examples listed for every independently listed Method as well (under the Method heading, which follows the chapters of the Scripting Guide on the right side of every Scripting Docs page).   Check out hangup to see an example.

This way, if all you need is a quick example showing how to put together a ’say’ in Groovy, all you have to do is click on say under the Methods sidebar and zoom right down to the code samples; no need to browse the entire Guide (though we still recommend giving the Guide a read through if you have time; even adept Tropo users might find something new and useful).

Questions, compliments, compromising pictures of @JohnnyDiggz during his GeeksWithoutBorders Tour?  Send em over to support@tropo.com.

http://www.w3.org/TR/speech-synthesis11/

SSML is used to control how text is rendered as human-like speech. It includes elements for describing the voice, pitch, speed, and other characteristics of human speech needed to ensure proper output prosody and pronunciation.

Requirements for enhancing SSML 1.0 were collected during workshops held in China, Greece, and India. The new SSML 1.1 W3C recommendation enhances SSML 1.0 to provide better support for a broader set of natural (human) languages.

In particular, SSML 1.1 supports

• a new registry for pronunciation alphabets that describe the pronunciation of words and phrases. Developers use pronunciation alphabets to describe precisely the pronunciations of words and phrases. An example is pinyin, a common way of writing pronunciations for Mandarin Chinese.
• the Pronunciation Lexicon Specification (PLS) to allow for standardized independent collections of pronunciation information that could be also be used by speech recognition engines. (more info about PLS)
• finer author control over voice selection and behavior upon encountering unexpected language content.
• better token delimiting for languages that (1) do not use white space as a token boundary identifier, such as Chinese, Thai, and Japanese, (2) that use white space for syllable segmentation, such as Vietnamese, and (3) that use white space for other purposes, such as Urdu.

Much of the work on this specification took place in China, a country with languages that are quite different from European and American languages. This is the first W3C recommendation in which Asians played a major role.

When published, the final version can be found at http://www.w3.org/TR/speech-synthesis11/.

Voxeo’s own Dan Burnett was a co-editor of the SSML 1.1 specification and contributed this to the W3C’s Testimonials page:

SSML is an important part of the overall ecosystem of W3C standards enabling speech across a variety of applications. SSML in particular provides a key way to render richer, more natural sounding speech. We are particularly pleased that SSML 1.1 provides advancements in several key areas, including support for Asian and Eastern European languages as well as improved audio controls for authors. The headway in the Recommendation is the result of the work of the dedicated individuals and companies around the world who value the importance of standards work and support the W3C Voice Browser Working Group. Voxeo is very proud to have been involved in this significant global accomplishment.

We’re pleased to see SSML 1.1 reaching this milestone and congratulate all involved.

Working with Tropo’s WebAPI engine is fun, but there are a few ways to make it not only fun, but easier as well.  Since making developers lives easier is what our support team is here for I wanted to share these tips with you guys!! The first one is Tunnlr, which has been mentioned here before, but I figured I would bump it, since it really is an awesome service! You get 90 days free for all new signups, and there after it’s only 5 bucks a month for single user, or $15 for a 5 pack. You can sign up here and please tell them Tropo sent you :). Now a problem I had when working with Tunnlr was that among my 200 open windows I would often either forget I had a tunnlr session open, or lost track of it, and then I would have to locate the pid, kill it, and try again. This was a pain in the ____, so I figured lets make a shell function to address this issue, and now I would like to post it here to help our developers out as well!

#Create Tunnlr reverse http proxy, takes port as parameter
    function tunnlr(){
        pid=$(ps aux | grep tunnlr | grep -v "grep" | cut -d " " -f6)
        if [ "$pid" == "" ]; then
            echo "No existing pid found, starting tunnlr"
        else
            echo "Found Existing pid(s) [ $pid ], so long sucka"
            kill -9 $pid > /dev/null
        fi
    	    echo "http://web1.tunnlr.com:______" | pbcopy
    	    echo "URL Copied to clipboard"
        ssh  -nNt -g -R :_____:0.0.0.0:$1 tunnlr____@ssh1.tunnlr.com &
    }

Make sure you replace the redacted values above ( ____ ) with values from your Tunnlr account settings page.  You can find these by logging into your account, and clicking “how do I start tunnlr manually”, this will then show you the ssh command to start it, and you will see the missing values for your account.

Now since most developers are working with multiple web servers, which could be running on various ports, I figured it to be a good idea for this function to take port as a parameter; So to it, on port Sinatra’s default port of 4567 for example, we just run tunnlr 4567. The script will first check processes to make sure no other tunnlr sessions are running, if so kill them, and then starts the tunnlr session. It will them copy the tunnlr URL to your clipboard, since I kept forgetting this :), that way you can just paste it to the browser.

The next tool I wanted to mention is shotgun, which is a gem that allows yours changes to push live without having to restart Sinatra. You can find this gem here. Once installed (sudo gem install shotgun) you just run it like so:

dyer:~/dev/webapi git:master
→ shotgun -p 4567 sms_webapi.rb 
== Shotgun/Mongrel on http://127.0.0.1:4567/

-P 4567 specifies the port to run on, and I like to run this on the default Sinatra’s default port, rather then Shotguns default of 9393.

So with these two items you can now develop locally on your machine, and avoid Sinatra restarts when making changes. I really hope this helps, and if our team can be of any help please let us know. We are always available on Freenode (irc.freenode.net) in #tropo, so please drop by and chat us up, or simply you can email support@tropo.com.

John Dyer

Tropo Support

And since Ruby is one of the languages that Tropo Scripting supports, you may then use your new Ruby skills to write some Tropo apps!

Since I’d honestly forgotten the details since the site-wide search was set up almost three years ago, I logged into the server and looked.  The answer is that we currently use “GT Search all Blogs” by Giovanni Tufo:

/*
Plugin Name: GT Search all Blogs
Plugin URI:
Description: Search in all Blogs, Posts, Pages(both in the title and in the post), based on "WordPress MU Recent Posts" by Ron Rennik
Version: 0.2
Author: Giovanni Tufo
Author URI: http://joevanni99.wordpress.com
*/

However, it looks like I have a problem… that plugin seems to be NO LONGER MAINTAINED.

The web page referenced is no longer online, the plugin isn’t listed in the WordPress plugin directory, and while I did find the author’s current website it is in Italian and I couldn’t find anything referencing WordPress.

Which, to me, means that I get to add yet another action to my “to do” list:

Find a new “search all blogs” plugin.

Now, what I have works, obviously, (and works well!) but as WordPress continues to evolve and change I’m not really comfortable using plugins that aren’t also tracking the development of WordPress. At some point in the future, this plugin may stop working… and then it’s up to me to fix it.

In looking through the Plugin Directory, this MultiSite Global Search plugin looks pretty good:

http://wordpress.org/extend/plugins/multisite-global-search/

My one concern that I don’t see listed in the plugin info is whether or not it only searches public blogs on the site. I would hope that is the case, and if it doesn’t is probably a minor modification. (I have a number of “private” blogs on the site that are for experimentation and testing… but I don’t want those posts showing up in search results.)

Unfortunately, I don’t really seem to find any other “global search” or “search all blogs” plugins in the plugin directory.

What are other people using out there?

Do you want to use the python language to build applications that interact with your users / customers using voice, text messaging (SMS), instant messaging (IM) or Twitter?

If so, we now have a python module available to help you create those apps on our Tropo platform using the Tropo Web API.  The module is available for download here:

http://github.com/tropo/tropo-webapi-python

You can download it directly from the website or use the git version control system to clone a copy to your local system.

There are a couple of samples provided, one that works with Google AppEngine and one that works with the “itty-bitty python web framework” :

http://github.com/tropo/tropo-webapi-python/tree/master/samples/

We need to get some more samples in there, quite frankly, and will be doing that in the time ahead, so please do check back occasionally.  For more info on what you can do with the Tropo Web API in general, please read the Tropo Web API Guide.

The module is still in development, too, so you may see some changes to it over time as we work on continuing to improve it and make it better.  If you are git-savvy, the best way to stay up-to-date would be to clone a copy and then issue a “git pull origin master” from time-to-time to be sure you have the latest and greatest code.

This python module is freely available as open source, and I and others will be writing more about what you can do with it in the weeks ahead.  If you would like to contribute samples, find bugs or have other issues, please either contact me on Github or via email.

If python’s not your language of choice, we also do have Tropo Web API libraries for Ruby, PHP and C#. Check them all out on http://github.com/tropo/

Enjoy… and please do drop us a line if you build some cool apps with this library!

P.S. And yes, this python library has been updated to line up with our new library naming guidelines, so there are no expected changes to the module name any time soon. That’s been done.

P.P.S. I should note that Ted Gilchrist was the primary author of this python library for Tropo, with additional contributions from Saúl Ibarra Corretgé and myself.

However, if you go to our main blog portal at blogs.voxeo.com and enter text into the Search box visible there, you will search across ALL public blogs on this site.

That’s the trick to doing a site-wide search. It matters where on the site you are when you do the search. Hope this helps some of you out there…

It looks like a great place to visit! I love this shot. Very cool to see the gnomes taking in the trip, too!

http://www.readwriteweb.com/hack/

As their intro says:

Our Channel ReadWriteHack, sponsored by the Intel Atom Developer Program, is a resource and guide for developers. ReadWriteHack will outline best practices for designing and developing applications. We will also provide examples of hacks and how they work, to inspire you and show you what’s possible.

It’s great to see RWW adding a focus on developers, and it’s certainly joined my list of sites to regularly follow. Looking forward to seeing what they write about!

It will take a week or two for us to bring all the libraries in sync… stay tuned for more updates.

If you want the source code to do this your self, it is available on Github here. The original blogpost may be found here. Of course you are not limited to using Heroku and Amazon S3, but this serves as a working example.

I’m admittedly a sucker for anything related to typography, so I did have to check it out.  It’s a set of TrueType fonts for Windows, MacOS X and Linux developed by Mark Simonson and available for free under the Open Font License.  Code looks quite nice in the font when editing.

Another similar font I’ve enjoyed using is “Inconsolata” developed by Raph Levien and also available for free under the Open Font License:

You might ask – why does this matter? I mean, programming code doesn’t have any formatting when it’s compiled, so who cares about the font! Well… if you are working for much of the day using a code editor, it can be great to use a typeface that makes your code very readable on the computer screen or printed output.

What do you think? Are there other fonts that you’ve found you like to use for working with code? (There are many out there…)

http://www.voxeo.com/outboundivr/

The great news is that you can sign up in our Evolution developer portal and try out creating such an app for free!

More resources and links can be found at the bottom of that web page.

Today’s Chart of the Day from Silicon Alley Insider (SAI) is Comscore’s metrics on the use of various IM networks. For the venerable AIM, it isn’t pretty. After years in the #1 spot, a sharp decline starting at the end of 2008 led SAI to title their graphic “The Death Of AOL Instant Messenger.”

Yahoo had a huge spike around the same time, but within 6 months declined back to around where they started. I’m not sure what caused the Yahoo spike, but a few things may have contributed. Yahoo 9.0 was released around that time, as was the Yahoo iPhone IM client.

Facebook and Google Talk aren’t represented on the chart. Comscore is unable to break the IM traffic out from Gmail and regular Facebook usage.

http://www.w3.org/TR/2010/WD-voicexml30-20100831/

This newest draft listed only these changes:

• Revised Legacy profile description to match current thinking.
• Removed SIV Resource (section 5.4) since it is now covered along with the recognition resource in section 5.3.
• Update section 4.4 (Event Model) to match our current thinking about DOM events as the underlying model for all flow control
• Cleaned up text in sections 1, 2, and 5

Per Voxeo’s Dan Burnett, co-editor-in-chief of the specification, the biggest change is the third bullet about the updates to the event model (section 4.4).

Personally, I’ve found the best way to understand the changes is to view the “diff” of the version from the previous version at:

http://www.w3.org/TR/2010/WD-voicexml30-20100831/diff.html

Given our interest in VoiceXML and VoiceXML 3, we’ll be continuing to write about it here… you can see related stories at:

http://blogs.voxeo.com/speakingofstandards/category/voicexml/

http://www.etsi.org/plugtests/SIPit27/SIPit27.htm

As I’ve written about in the past and recorded a video interview about, these SIPit events are critical, in my opinion, to helping drive the overall adoption of SIP and open standards in communication systems.

If you are a creator of software or hardware devices that use the SIP protocol, SIPit events are a great way to test how well your equipment works with other SIP implementations. There is a fee, but it’s small for the week-long testing you get to do. More info can be found on the registration page.

P.S. You’ll note that this SIPit event is the week following IETF 79 in Beijing, China… so if you can make the travel work, it’s a great way to combine two weeks of open standards / SIP – related events.

http://github.com/tropo

However, as those libraries have evolved, we’ve found that they’ve developed in slightly different directions with regard to the naming of the repository on Github and to the module names, etc.  We gave some thought to the best way to provide guidance to make it easier on developers, and came up with these goals:

• Provide a consistent repo name so developers using multiple languages can easily locate their downloaded/cloned directories on their local systems.
• Provide consistent naming of module/library names and function/class names so developers can move with ease from one language to another. (Use case: a developer builds a Ruby app for one client for Tropo. Developer then needs to build a PHP app for another client. Developer should be able to move from one Tropo library to another without having to relearn or guess what the names are.)
• Allow for the future creation of additional Tropo-related modules/libraries

To get there, the existing repositories are going to migrate to these new names:

• http://github.com/tropo/tropo-webapi-ruby
• http://github.com/tropo/tropo-webapi-php
• http://github.com/tropo/tropo-webapi-python
• http://github.com/tropo/tropo-webapi-ruby

Going forward, we’ll ask that anyone creating a new Tropo Web API module please follow this same naming convention.

We’re also going to move the names of the actual modules to be more consistent. The result will be something like:

You’ll note that we’re treating “webapi” as one word in the names. (Note: These may be the final names, but they may change as the changes get made to each library.)

We’re in the process of working with the library authors to make these changes now.  The migration is a pain, quite honestly, because it’s going to break existing applications, samples, tests, etc., but we want to do this once so as these modules start to get more widely used, there is a good degree of consistency between them.

We’ll let you know as the libraries are moved over to the new convention – and please let us know now if you have any feedback on this issue.

Thanks – and we’re definitely looking forward to continuing to see the awesome apps you all are building with Tropo!