Posted by Kevin Micalizzi on Sat, Apr 11, 2009 @ 05:11 AM
Sramana Mitra continues her work with entrepreneurs in her Strategy Roundtables. Working in collaboration with INSEAD students and graduates, Sramana gives these entrepreneurs guidance in vehicle fuel monitoring, makeup brush cleaning, computer-based simulation solutions and tools, and personalized online career consulting ventures.
[There's a slight echo at the beginning due to one of the speaker's Skype connections, it gets resolved quickly.]
Posted by Steve Chazin on Wed, Dec 10, 2008 @ 12:23 PM
Talk about eating our own dog food - we just held our board meeting on Dimdim. There was 1 attendee from California, 1 from New York, 2 from Massachusetts, 2 more from London, 1 from Mumbai, and 1 from Hyderabad. If we had flown to California, then we would have spent at least $4,000 per person in airline, hotel and car rental costs. By avoiding the travel of 7 people, we saved about $28,000 by conducting the meeting over Dimdim.
This year we have had about 8 such board meetings over Dimdim. So we have saved about $224,000 this year.
On top of that I have another Dimdim meeting with a potential partner in 30 minutes. If I had a face to face board meeting then I would never be able to make it to this next meeting. And so my productivity increases too. Bye, gotta run to the next meeting.
Warm regards,
ddg
CEO
Posted by Steve Chazin on Tue, Oct 21, 2008 @ 08:53 AM
Folks, today at 2PM ET, Leading entrepreneur and strategy consultant Sramana Mitra partners with ReadWriteWeb to bring you a Web 3.0 product strategy roundtable online, using our web conferencing platform. To attend, simply click this link at 2PM or dial-in to hear Sramana give live feedback to 8 entrepreneurs. If there is time, she will take your questions submitted live using Dimdim. This is going to be fun. We'll see you there!
Dial in: 1-419-400-0203
Conf ID: 918878
Here is the replay!
Posted by D D Ganguly on Fri, May 02, 2008 @ 08:18 AM
One of the things that prides us here at Dimdim is being able to quickly respond to your needs. We read and react to all your emails and we encourage you to voice your opinions. Since our public launch a little more than 3 weeks ago, we heard that you really like Dimdim but that there were a few areas you wanted to see improved. One of those areas is our Open Source Community Edition. You didn't like the fact that the current version is not on par with our Dimdim Free hosted version, you hated the 5 person limit and only one meeting at a time, many wanted more than one video stream, some of you didn't like the server software installers and a few even complained about our documentation. Well, I'm proud to say that your voices have been heard! (and that we've been really busy...)
In the next week we will release a new version of the Dimdim Open Source Community Edition to SourceForge.net. (UPDATE: our new Open Source release is now available) That version is on par with the features of our hosted offerings, removes the 5 attendee limit, enables multiple simultaneous meetings and even adds a 2-way video chat feature. We have also packaged the Dimdim Servers into a single VMWare Virtual Appliance to ease the installation process. You'll also notice much improved documentation and a new admin console.
We hope you will spread the word to everyone that we're committed to the open source model and that our actions here speak louder than words. We're seeing some really innovative uses of our product and we really want to see more. We simply ask you to tell everyone that Dimdim is serious about democratizing the web meeting space and to come and try it. Remember, you don't need to download a thing* to test out Dimdim (just sign up by clicking that orange button on the upper right of this page) and your invitees don't need to join Dimdim to attend your meetings. (and we won't make any of them install a thing either.) And if you want, you can download and test our Open Source Community Edition on your own servers. Enjoy.
*Some of you have also said we've been a bit misleading about this "no download" thing. Please read this article to clear the air. We're sorry if we've caused any problems with our overly aggressive marketing. Blame the CMO.
Posted by Steve Chazin on Wed, Mar 26, 2008 @ 01:40 PM
Blog post by Saurav Mohapatra
The first version of Dimdim was (and the conference server part of the current version still is) a java web application running in Tomcat. We’ve slowly been moving towards more native components in our server stack on a case by case basis. Not to be drawn into Java vs. Other server tech debates, this move is a part of a growing consensus about leveraging the OS provided capabilities to the fullest to achieve high throughput, optimal performance while balancing that out with development cost / learning curve.
For our latest feature, the Screen share capability, we switched to a C++ based FastCGI application behind the NGINX webserver.
The move was done after a lot of thought and exploration. For the traffic pattern of something like desktop sharing, the current java technologies simply did not add up. A typical screenshare traffic pattern transmits around 1-2 mb of data for a 1280 X 1024 for per minute of share for something “heavy” like Powerpoint slide show or Full screen image viewing. Thus we switched to a robust native technology server. The app was written using FastCGI technology and first version ran behind the lighttpd server. As we continued testing, we found better performance especially for longer duration shares with the NGINX web server and thus that was chosen finally as the front server for the desktop screencasting application.
We built on top of the tried and trusted RFB protocol of VNC adding HTTP/S communication layer using libcurl for publishing the desktop stream.
The viewer was adapted from FVNC code base from OSFlash and we added tight encoding to it along with HTTP/S communication layer.
The point I’m trying to make here is that we built on top of existing robust components and application and were able to achieve significant progress in a shorter duration than if we’d have developed the whole thing from scratch.
Now this whole exercise kind of paraphrases and proves what I have always believed about open source. It is an organic philosophy, perhaps one of the best displays of the most positive sides of human nature. You build something based on stuff that someone has been built before and others after you will use the stuff you’ve developed to build their own.
It is a really wonderful philosophy and in this geek’s opinion perhaps one of the best ways to live life.
Posted by Sundar Subramanian on Sat, Dec 09, 2006 @ 07:48 AM
We feel, in software, user experience and user talk fulfillment is more important than feature set expansion. A software product which fulfills the needs and expectations of the user is better than a product which is more feature rich but difficult to use. This has been the core of Dimdim's product management philosophy and it is this principle which guides us.
Features which are not central to the core functionality set of Desktop/Application/Presentation sharing and which will suck out our scarce resources are not given priority unless our user communities ask for them. This means that our engineers can focus on making the core features easier to use and the product in general will be more stable. This underlines the philosophy we are following for our Beta release slated for release around January 2007.
New features in this release will include two way Audio and Video along with a much more scalable backend server. The UI (User Interface) is also being completely redesigned to ensure optimum usage of screen real estate. More features like collaborative whiteboard and recording and archiving will follow. To achieve this, we may have to sacrifice the "Remote Control" feature, where the presenter can control the Attendee's computer. This feature will follow in later versions of the product.
Of course, this kind of development where the stress is on user experience means that the focus should be on a task oriented fulfillment of user expectations. What I mean by this is, adding features only for the sake of adding features is not the way. Features will be added if they are required to complete the user experience as expected. For example, look at this ridiculous situation of Windows Vista presenting the user with Nine (yeah, that's right) different options when a user wants to shut down the machine. Now, as reported by a developer who has worked on this particular feature, such a situation arises because decisions are not made with user intention (and expectation) in mind but by a group (in this case 43) people who take decisions depending on requirements/limitations defined by other groups of developers working on related code.
In Dimdim too we had a situation where many options could be set by the Presenter of a web meeting and the straight forward way of doing this would be to present all these options upfront before the start of the meeting. But this would have resulted in a situation where by even before the meeting begins, the presenter is confronted with a long list of options which have to be defined. This is not desirable user experience. So we took a call that expected values of parameters (like all participants will be able to chat with each other by default while they will not be able to start their video feeds by default) would be already defined when the meeting starts while at the same time giving the flexibility to change these values at any time.
Of course, Dimdim does not even have 43 employees! That's where our user community helps us ensure that all of our product designs are community decisions and not "committee decisions". So when you receive an email invite to participate in our hosted Beta, please let us know what you think about Dimdim and how we can make it better.
Ps: Did I mention that you can register for the free hosted invite from our website at www.dimdim.com
Posted by Saurav Mohapatra on Fri, Sep 08, 2006 @ 02:20 AM
Want to lose weight quickly? Seeing things a tad too clearly? Want to be "Comfortably Numb" like the Pink Floyd song says (without getting assistance from illegal chemicals)?
Don't worry! Dr. Mohaps got a new lifestyle for you! It's called "BrUnNerTM".
BrUnNerTM = Breakfast +lUnch + dinNer
[note: The meal is to be taken at one usually at 18:30 hrs and has to be washed down with generous doses of caffeine]
[Caveat Emptor : Dr. Mohaps is not a Real Doctor and BrUnNerTM is not a real meal/diet/lifestyle. Side effects might include seeing things, late nights and heavy duty coding binges. Prolonged BrUnNerTM lifestyle might result in severe burnout and addiction to watching crappy hindi films over and over again.]
We finally pushed out the alpha build of dimdim Web Conference out and boy! Did BrUnNerTM help!
We had a marathon bug fixing session just prior to the launch, which helped us set new uptime records. Jay was the hands down winner managing to stay awake for nearly 48 hours and Prakash got a honorary mention for doing 24+ with Jet Lag (he basically continued the same day across the globe traveling from India to US. I think there is a twilight zone episode here somewhere)
Even new hire Rohit was thrown straight into the fray working on the build system.
But anyways with a small dev team stretched to its limits and working round the clock (The only "employee" of dimdim who does not work 7 day weeks is our cleanup lady/maid/cook who has Saturdays and Sundays off ) ] , we have managed to quell a lot of doubts that people initially had as to whether or not we could deliver on the promise of the dream that is dimdim Web Conference.
Now there is celebration (mostly involving catching up with long deserved sleep and sampling other meals of the day individually) but soon (within 16-24 hrs) we are going to drive full steam on the beta path.
Hope to talk to you soon and long live the "BrUnNerTM" lifestyle!
Cheers
mohaps
Posted by D D Ganguly on Mon, Aug 21, 2006 @ 09:30 PM
In my previous post I wrote in general about how we take most of our decisions based on input from the community and in this post I am going to detail that with examples and practices that make user led product development possible.
Initially as we brainstormed (or is it bullshitted) within the company regarding the target space for an open source web conference product. E-learning space was one which emerged due to its cost-sensitive nature at the same time being pretty demanding in terms of features and scalability. So I contacted the developers and lead users of a few top open source e-learning products and immediately we started the process of learning from their experience. The founder of one such product helped us so much that we turned to him to get feedback every time we felt we had taken a significant step. This resulted in building trust between us and he even recommended other lead users and developers in that particular community who might be interested in Dimdim. This helped us immensely as each person on that list was very knowledgeable about the needs of users and ways to fulfill the needs. We even found a developer who wanted to integrate Dimdim with the e-learning product so that he could use it in an upcoming conference. This was god-sent manna for us as such a move would expose our product to the wider community leading to much needed feedback from early adopters. Through word of mouth we also found other value added resellers and solutions providers of the e-learning products who were interested in knowing about Dimdim.
Well enough of this self-congratulatory stuff, let me now talk a bit about what I feel are necessary attributes to ensure positive interaction between a company and the wider open source community:
- By default most people have more good in them than bad. No I am not launching a new religion or cult but what I mean to say is: most people want to help a positive effort even if it has no direct bearing on them.
- Respect the users by being genuine; don’t tell lies, ideally under promise and over deliver. They are as intelligent if not more, than you and can easily see though a gimmick or sleight of hand. When we gave the first demo to a potential user I started out by saying that it was pre-alpha build with potential bugs and it didn’t even have the main features he wanted. He thanked me for being polite enough to say that upfront and when some bugs were found out, he understood the situation enough to have confidence in us to go forward with the process.
- Talk and more importantly, listen. That way you will not only know the problems and issues encountered by users but also most probably the users themselves will give the solution. For example, when we talked to a user of an established web conferencing product the main points which came out were: Slow to respond, Needs software to be installed on the attendee side even if the attendee does not participate actively but is more of a viewer and the need for integrated audio instead of calling in through a separate phone line. So solving these issues became part of our core features of our product.
I feel these principles are true for any interaction: be they between a company and a community or between two individuals. Now that I have established that, did I mention I am an uncommitted bachelor looking out for a suitable female companion? :)