|
From: | Dmitry Gutov |
Subject: | Re: emacs rendering comparisson between emacs23 and emacs26.3 |
Date: | Mon, 6 Apr 2020 17:55:50 +0300 |
User-agent: | Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101 Thunderbird/68.4.1 |
On 06.04.2020 05:36, Richard Stallman wrote:
If and when such a mechanism is in place for us free software activists, we might agree that open-paren-in-column-0-is-defun-start is not worth while any more. But it appears that has not happened yet, so I think that the open-paren-in-column-0-is-defun-start guided heuristic IS still worth while for us.
This question (whether syntax-ppss is "too slow") was raised a couple of times already, but in all previous discussions, as well as in this one, nobody has done the bare minimum research (e.g. profiling some real-world scenario) to prove this claim. Nor, if it indeed showed up in profiler reports, examined which part of the current mechanism is the bottleneck. Nor tried to fix that part somehow.
This is not how we should approach performance problems.The only arguments I see are nostalgic (e.g., somehow, Emacs was faster years ago, maybe? the files were smaller, and the C standard was simpler?)
[Prev in Thread] | Current Thread | [Next in Thread] |