Exactly one month ago, I released MPUI 1.1.4. The first bug reports came in only one day after the release, so I planned that 1.1.5 shall be a bugfix release some days later. For reasons I can’t remember, I missed this date. This may be in part due to the fact that my TODO list grew, no, it exploded because of countless suggestions I received. In addition, there were the preparations for the CLT that left almost no time for other stuff.

In the time that was left, I started to implement the lesser features (as in »requiring the least time to implement«) on my list. If I hadn’t decided to draw the line today, development of 1.1.5 would have continued forever, I guess.
Read more …

For quite some time now, I always wanted to record songs from my favourite radio station. However, there was a problem: They only offer a RealMedia stream (at only 44 kbps, but the codec is remarkably good, on par with WMA or Vorbis), which is not as easily recordable as, say, HTTP streaming. However, with the help of the original binary codecs, MPlayer is able to play the stream – so it had to be possible to record it. Unfortunately, MPlayer’s –dump options only record the compressed stream, and MEncoder refuses to transcode it into something usable (.wav, .mp3, …) because it always expects a video stream to be present. Furthermore, none of the dump or transcode options are able to play the stream while recording.

This had to change. And since I had a MPlayer-for-Win32 build environment on my box anyway, this was the right time. After some reasoning, I came up with a functional solution that doesn’t require twiddling around in MPlayer’s innards too much: an audio filter. If –af lame is put into MPlayer’s audio filter chain, it simply encodes all audio samples that pass by into an MP3 file on disk. MPlayer’s audio output is still active, thus an MP3-encoded audio dump is created on-the-fly during playback.
Read more …

I just submitted my second patch to a public open-source software project. The first one was about one or two years ago, a new feature for a not-too-famous image viewer named QIV. The project I submitted today’s patch to is, erm, somewhat bigger: it’s MPlayer. It’s a patch for the DirectX video output driver that used to take 100% CPU time on some (but not all) systems.
Read more …

I love OpenGL. Whenever I want to do some real-time 3D stuff, OpenGL is the API of my choice because of its simplyness and platform-independence. But honestly, OpenGL can also be a bitch sometimes. For example, there is no portable way of using textures with non-power-of-two dimensions. Another major annoyance is the way »modern« extensions are handled. To use recent technology like shaders, you need C header files that contain all definitions of the new functions, constants and so on. Annoyingly, these aren’t available in some development environments: Out-of-the-box OpenGL support ends at version 1.1 in both MSVC and my favourite dev system, lcc-win32. Read more …

Remember what I said about the tedious job of maintaining software projects? This time, it even struck me twice: Just two days after the release of MPUI 1.1.1, I received a batch of bug reports and other suggestions, so I did another rework of MPUI yesterday. Next to half a dozen minor bugfixes, I finally implemented a »clip info« feature that relieves the user from the task of parsing the (sometimes cryptic) MPlayer output.

This morning, even before I left my bed, a good friend …erm… persuaded me to add a new feature to CENSORED: He is preparing a presentation to be held at the university. The problem is, the computer where he’s going to show the slides is not his, and it’s equipped with an ancient GhostScript installation. The 7.x versions of GS have some serious issues with rendering PDF files generated by pdf(La)TeX. For example, hypens (‘-’) are shown at completely wrong places. To overcome this, I added an option to »render« a PDF into a directory with one PNG file per page, thus allowing to generate images on one computer (that is equipped with a decent version of GhostScript) and presenting them on another one.

I love coding, but to the same extent I hate project maintenance. This is a problem I share with most other coders, I guess – but well, sometimes it just has to be done. And today it was MPUI‘s turn to get some fixes.
Read more …

During some random discussion with my colleague today, I rediscovered an experiment I did quite some years ago: MP3 compression of image data. This is, feed raw 8-bit-per-pixel grayscale image data into an MP3 encoder, let it compress it, and decode the MP3 file again. After some conversion (the decoder will always produce 16 bit signed output, but 8 bit unsigned image data is needed), a raw image file will result of this whole process, ready to be displayed with an arbitrary image viewer. Sounds simple, eh? :)
Read more …