从性能/速度的角度来判断 swt 与 swing（是的，这是经典的 :D）
最近，我一直在寻找上述参赛者中哪一个在性能方面更好，我发现了一篇相当有趣的文章，与 swing 相比，它摧毁了 SWT 的性能评级：http://cosylib.cosylab.com/pub/CSS/DOC-SWT_Vs._Swing_Performance_Comparison.pdf 。值得一提的是，基准测试是在 2005 年完成的，从那时起事情可能已经发生了变化（不知道对谁有利）所以：
1）有人知道最近进行的任何性能测试吗？ （最多 2-3 年：D）
2) 你在这件事上的个人经验对我来说非常宝贵，所以请告诉我你是否注意到 swing 和 swt 在性能方面的差异
Lately i've been looking to see which of the above contestants is better performance-wise and i've found a rather interesting article that demolishes SWT's performance rating compared to swing's : http://cosylib.cosylab.com/pub/CSS/DOC-SWT_Vs._Swing_Performance_Comparison.pdf . It's worth mentioning that the benchmark was done in 2005 and since then things might have changed (dunno in who's favor tho ) SO:
1)does anybody know any performance tests done recently? (2-3 years at most :D)
2) your personal experience on this matter is invaluable to me so do tell if you noticed differences between swing and swt when it comes to performance
如果你对这篇内容有疑问，欢迎到本站社区发帖提问 参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱，如果其他用户或者作者回复了您的评论，将不能在第一时间通知您！
发布评论
评论（2）
这不是一个完美的比较，但 MigLayout 有一些可能性来对 GUI 性能进行基准测试。请参阅官方网站上的使用 MigLayout 演示进行 GUI 基准测试部分。
这也是在探查器下检查 miglayout GUI 基准测试以查看 SWT/Swing 内存使用情况的好方法。这是一个巨大的差异。
无论如何，我同意安德鲁·汤普森的评论：“运行速度更快的将是拥有更好的程序员和更好的客户端代码的人”
It's not an perfect comparison, but MigLayout have some possibilities to benchmark GUI performance. See Using the MigLayout Demos for GUI Benchmarking part on official sites.
It's also good way to check the miglayout GUI benchmarks under profiler to see memory usage of SWT/Swing. That's a huge difference.
Anyway, I agree with comment from Andrew Thompson: "The one that runs faster will be the one that has better programmers & better client code"
运行得更快的将是拥有更好的程序员和技术的人。更好的客户端代码。
当然，这是一个好点。人为因素对于性能（以及可用性、可维护性以及所有功能和非功能品质）非常重要。但 OP 专门询问了技术/工具差异...
我们可以假设设计师/程序员是一个固定的群体，并且会对这两个 API 应用类似的技能，从而排除人为因素。或者，我们可以应用标准基准测试来测量 API 之间的行为差异，按照 http://cosylib.cosylab.com/pub/CSS/DOC-SWT_Vs._Swing_Performance_Comparison.pdf 。
这留下了一个非常有效的问题。 Swing 和 Swing 之间的客观差异是什么？ SWT？如果能够看到过去两三年的彻底“客观”性能基准测试，那就太好了。对我来说这个Q是活的。
The one that runs faster will be the one that has better programmers & better client code.
Sure, that's a good point. The human factor is very important for performance (& usabilty, maintainability, and ALL functional & non-functional qualities). But the OP was specifically asking about technological/tool differences...
We can take the human factor out of the mix by assuming the designers/programmers are a fixed group and will apply similar skills against both APIs. Or we can apply a standard benchmark test to measure difference in behaviour between the to APIs, as per http://cosylib.cosylab.com/pub/CSS/DOC-SWT_Vs._Swing_Performance_Comparison.pdf .
This leaves a very valid question. What are the objective differences between Swing & SWT? Would be good to see a thorough 'objective' performance benchmark test from last 2 or 3 years. To me this Q's alive.