Technology

Converting the videos is easy enough. ffmpeg is the tool to use:

$ ffmpeg -i yourfile.avi -target pal-dvd -aspect 4:3 output.mpg

You can use any format for input that your ffmpeg is compiled to support, so this works just as well for Youtube and Google Videos that you cannot find the mpgs for:

$ ffmpeg -i youryoutubevid.flv -target pal-dvd -aspect 4:3 output.mpg

The problem with making your DVD is the actual authoring process. I've been using ManDVD with some success, although it seems to be slightly incompatible with modern ffmpegs (make sure to add a menu sound file, as ManDVD fails to instruct ffmpeg correctly to make a silent one to multiplex in), and can be frustrating to use: no working project save, for example. It will author and burn disks, but sometimes, if you're using ManDVD or another tool, the process will fail after the creation of the authored content in AUDIOTS AND VIDEOTS directories. If that happens, you can do this:

$ growisofs -dvd-compat -Z /dev/dvd -dvd-video .

That will burn you a new, properly authored DVD if you run it from the directory where your VIDEO_TS directory resides.

As far as the authoring goes, what we need is a simple frontend to dvdauthor which will create an XML document and design the titles, menus and masks nicely. I know that I could make all the menu files in the GIMP, but I just don't have the time. I want a simple, templates-based approach. I'm happy to do all the file conversion, up front, and I'm happy to use cdrecord or growisofs to handle the burn. I need help with the DVD authoring, that's all.

Let's stop this from being a compatibility crap-shoot, and get a process that produces consistent, standards-compliant DVD content into order, without loads of scripting.

Edit

Ok, maybe I was ill-informed. I'm trying VideoTrans at the moment. It seems to be making a consistent and quite sophisticated menu system for the four videos I want to turn into a DVD. I'll see what the results are like.

Many people are aware of the power of PHP heredocs, in that they allow variable expansion without the escape-hell of " and ' quotes having significance. The only downsides are the strictness of the EOF marker locations and that constants cannot be used in heredocs.

I always used to be bothered by the fact that my code in heredocs was all-black, and not highlighted at all. I now realise that this is due to the fact that to get the highlighting you want, you need to choose your markers correctly. See this comment for details. No more drab EOF. I can assure you that the syntax highlighting allows for longer markers, like HTMLform and so on.

So this method works for HTML, JavaScript and SQL. Great. But what if you, like me, create "dynamic stylesheets"? Here's an example:

Read More

This is my first hidden service, for testing purposes. smbzeu3ydgy63zlk.onion. Give it a click and tell me if you see anything!

Like Alisha on Soft Tech Reviews, I feel like a late adopter of Twitter. Truthfully, I ignored the hype when the service came out, and the idea of using it to tell the world about the minutiae of my daily activities seemed rather stomach-turningly boring for writer and reader alike. Seems like I was just being uncreative.

Twitter is a tool like many others: it needs some thought in order to get the most out of it. Firstly, though, we've got to know what we're working with:

• A means of posting 140-character messages to an audience, who can opt to be instantly notified by mobile text message, IM or e-mail wherever they are, and can respond if they wish.


• An API which can be exploited to allow one's systems to integrate with Twitter, either by indexing and parsing twits, or sending and receiving them. This includes allowing computer systems to have their own accounts for the purposes of sending and receiving instructions and content to a human user.

That last point links directly into how I am using Twitter with Sandy. Sandy is a reminder service. It has some vaguely human characteristics, for example the ability to parse commands for times and dates, and to send a reminder by e-mail, mobile text or IM, even on a recurring basis. The Twitter interfaces are very useful. When Twitter's Jabber IM service is working (it isn't at the moment...), I can send a remind command through by that means, and also request information from the diary that Sandy keeps on my behalf. I can do the same through Twitter by text message, and this makes it very useful: I can set a reminder on the go, as though I were texting a savvy assistant.

People have written Twitter extensions for Firefox, and I can type a message in my Firefox address bar and twitter it directly to Sandy (so not a public message). Setting reminders as I think about them like this is very effective.

So, I've found a use for Twitter with a bit of ingenuity. I'll be expanding my use to include communicating with clients and colleagues over the next few months, but for the moment, I'm very content to have it in my arsenal of contemporary, web 2.0 communications tools.