I've been trying to teach myself how to create videos from [java, brms] user group presentations for over a year now, and it's mostly been a failure despite some tremendously successful efforts. Only recently have I learned enough to make it a success, and even now I can't actually succeed, I just know how I might succeed if I wanted to invest even more effort.


  • The "Extra Half Hour" benchmark to put a presentation online is not yet reasonable.
  • The Flip video camera can be your friend.
  • The coding on screen is the hard part, everything else is much easier to "get".
  • Hosting your app on YouTube or other is still wrought with challenges.
  • Software can be maddening
  • InfoQ is still the measuring stick
  • The Zoom - a final trick.

The "Extra Half Hour" Benchmark:

Here's my benchmark for success and why I haven't come close yet. My thinking is I should be able to spend the time shooting a presentation at my favorite user's group, invest an extra half hour, and that presentation would then be "live" for anyone who didn't get a chance to see it. 

We have hundreds of people in our local java user's group, but only 50-70 can take the time to show up for the usual presentation. Quite often it's a pretty serious speaker, so it would be nice for anyone, anywhere, to see this great presentation even a year later.

So far, this benchmark for success is so wildly ridiculous that it has made me a laughing stock. They haven't banned me from the meetings yet, but they threatened to because I show up with a camera, record it in front of everyone, then it never shows up online. I've tried and tried, but I never can get all the way through the problems in the time that my other life allows.

The Flip

One of the first things you will need is a way to record. To make a long story short, I invested in more expensive stuff, but it's kind of hard to beat the Flip. 

Not even the creators of the Flip intended it to be such a great tool, but it is, Amazing ease of use, perfect picture quality, great sound. You can do much better for ten times as much, but you might not to lug all that good stuff around for just a user presentation.

Coding On The Screen

Take this example: A recording of one presentation for coding examples can be 6 gigs. Try putting that on YouTube. "Fugetaboutit". 

If you want to segregate your efforts into "easy" and "hard" I'll give you a hint. The easy stuff is the talking head stuff and the slides. No biggie, there. The hard part is all that pesky screenshot type stuff.

Why ? Because the talking head stuff can be small format because you don't need to see much besides the guy's headshot at low fidelity. Everyone listens, they don't want to see much detail. Because the slides can be easy because they are just slides, you can just provide a link to the slide deck pdf and you're done.

But if you want to show the guy coding on the screen....now that can be a real challenge because that's where it's a mamoth file. It has to be high enough resolution for you to actually show what the guy is doing.

Hosting Your Video

YouTube would be perfect except for nothing more than 10 minutes, so a typical 90 minute presentation would be 10 different videos.

Then there are dozens and dozens of others, each with limitations, requirements etc. From vimeo to blip to you name it. I never did make it through the experimentation stages, still watching others though, I might copy a successful user's group if I can find one.

We also tried hosting our own but after mucking with the quicktime streaming settings for a couple weeks on our hosted server we finally just gave up. No way, no how. But we might have come close....

There's also EC2 or other similar Amazon options, but too many to experiment with in that "half hour" window, and they cost money anyway.


I have experimented now with many different pieces of software for editing the videos and publishing them online. Don't even ask how bad it's been. Everything has great promise starting out, but then it won't do ..... or ..... or ..... and meanwhile I've invested hours on being teased. No, I'm not going to buy a Mac with a thousand dollar editing program, which would solve all my problems. I'm doing this for free, for other user guys in the user groups.

Examples of problems

  • creates great videos but garbled when rendered (2 different vendors)
  • firmware locks videos from access except from it's own program
  • underpowered, won't handle the files
  • lack of features such as zooming in etc


Want to see all this done right ? It's also a pretty good demonstration of why some things just have to be commercial to have enough critical mass to make it worthwhile. 

InfoQ does a great job of putting presentations online. Nice separation between headshots, slides, audio, and the related code samples. And it all works. 

The Zoom: A Final Trick

If you want to learn from my follies and make it all work, the trick is in the zoom up on code samples. I learned this from InfoQ's video of the Rod Johnson SpringOne presentation.

What you do is this. Make your video of the headshot serve also as your video of the code samples. In other words, shoot the camera at his head unless he's doing code samples, then shoot the screen.

Then - this is where it all becomes possible - edit the code samples video to zoom in super close only on the part of the code that is changing. That makes it possible for the screen to follow the code and still see the type big in a small format.

Now, because your entire presentation is either headshots or zoomed up tight code, you can render your massive file down to a micro size that is easily publishable, such as InfoQ uses. Publish it along side the pdfs of the guy's slides, and presto! Maybe still not within the half hour, but at least it's a possible success.

I don't have the editing stuff to do this yet. But I'm going to see if I can, in future months. Maybe if I can use a combination of software products that didn't work previously, just change the combination around.

Wish me luck. I've still got some great presentations that I haven't put online yet.