Hi there! It's time for a new status report about the progress for our new major release, transcode 1.2.0.
The key and no-more-delayable feature of this release will be the introduction of a full-NMS export layer, featuring all the long-needed goodies like support for modern codecs/formats. So, back in time in january, that was the roadmap: (reference: http://tcforge.berlios.de/articles/tasks/index.html ) * import_alsa power up. * import_v4l2 tuning support. * Deep review of tcframe*. * Split libtc in smaller libraries. * NMS-compliant export layer. * Import the Synchronizer patch into HEAD. * switch to graphicksmagic (http://www.graphicsmagick.org/). And this is what we have as in today's HG: * import_alsa power up. To be done. It's a small task, anyway. * import_v4l2 tuning support. Done, fairly stable. Needs more testing. * Deep review of tcframe*. Delayed for 1.3.0 or beyond. * Split libtc in smaller libraries. Done. No regressions so far; needs more documentation, even though the avalible one is pretty decent. Some minor bugs can still spawn whatsoever. * NMS-compliant export layer. I actively searched for design flaws into current architecture, (of course) found some of them, addressed then solved. All that is nice, but that required (and requires) some time. The whole thing it's coming along pretty nice, however. All the preview modules so far published will still work of course, and new ones are coming (libavformat muxer support, mkv direct support very likely). The merger with public repo is not too distant into the future, but some weeks more for development are still needed. * Import the Synchronizer patch into HEAD. Done. Needs testing and tuning. No regressions spotted so far, but wider testing can eventually let them pop out. * Switch to graphicksmagic (http://www.graphicsmagick.org/). Development done. It even compiles! But testing it's just at the beginning, expect problems e.g. from the affected filters (especially). Floating (not "if", but "when") tasks: -------------------------------------- - tcframe* review: will be triggered by the new framebuffer model discussed some time ago. - new framebuffer/executable model: needs NMS modules up, running and stable enough. We have to wait. - scripting frontend: on hiatus. No brilliant ideas about how to implement it in an *useful* manner. Needs new executable? Maybe suitable for a socket 2.0/better IPC interface? Both of them? - GUI. I changed my mind: it will be a frontend on steroids, likely coded in a scripting language (python). Needs a stronger IPC, and a review of how transcode behaves (no sudden exit()s). Maybe it's just better to pop out a new executable more frontend-friendly? (tcbackend? tcengine?) - Versioned installed files (see recent email message): scheduled for 1.3.0. Short term plans: ----------------- Focus on NMS and merge the code as soon as possible. After that, most of the development is done, the remaining bits are self contained (mostly modules). A first 1.2.0 alpha version will be (hopefully) released this summer. Bests, -- Francesco Romani // Ikitt http://fromani.exit1.org ::: transcode homepage http://tcforge.berlios.de ::: transcode experimental forge
