Tir is much newer than Kepler or LuCI, so the code isn't nearly as mature. I would rank Tir as experimental, right now. The same factor also means that it has significantly fewer features.
It does have a very pleasant continuation passing style of development available though, through its coroutine based flow stuff.
I would rate it, personally, as fun for experimentation, but probably not ready for heavy lifting until Zed stabilizes it more :-)
This video from PyCon 2011 says basically you scale on multicore or multiprocessor by running more workers, under high load condition the memory advantage gives better performance. In the video it's said that at Meebo's they have used this approach for last months with huge load.
The video is python specific, so it's just for the scaling of coroutine approach part of the question. Video length is about thirty minutes.
发布评论
评论（2）
Tir 比 Kepler 或 LuCI 新得多，因此代码还不够成熟。我现在将 Tir 列为实验性的。同样的因素也意味着它的功能要少得多。
不过，通过其基于协程的流程，它确实具有非常令人愉快的连续传递开发风格。
就我个人而言，我认为它对于实验来说很有趣，但在 Zed 使其更加稳定之前，可能还没有准备好进行繁重的工作:-)
Tir is much newer than Kepler or LuCI, so the code isn't nearly as mature. I would rank Tir as experimental, right now. The same factor also means that it has significantly fewer features.
It does have a very pleasant continuation passing style of development available though, through its coroutine based flow stuff.
I would rate it, personally, as fun for experimentation, but probably not ready for heavy lifting until Zed stabilizes it more :-)
PyCon 2011 中的 这段视频 基本上说明了您可以通过运行更多工作线程来在多核或多处理器上进行扩展，在高负载条件下内存优势提供了更好的性能。
视频中称，Meebo 过去几个月一直在负载巨大的情况下使用这种方法。
该视频是特定于 python 的，因此它只是针对问题的协程方法部分的缩放。视频长度约三十分钟。
This video from PyCon 2011 says basically you scale on multicore or multiprocessor by running more workers, under high load condition the memory advantage gives better performance.
In the video it's said that at Meebo's they have used this approach for last months with huge load.
The video is python specific, so it's just for the scaling of coroutine approach part of the question. Video length is about thirty minutes.