emacs-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Performance degradation from long lines


From: Ihor Radchenko
Subject: Re: Performance degradation from long lines
Date: Fri, 26 Oct 2018 16:58:09 +0800

       > Would it help to "say the same"?

       > I'm trying to come up with ideas to help people get bearable
       > performance in practical use cases.  I can shut up if people don't
       > want to hear about partial solutions that don't work in all cases.

       > Assigning blame and/or venting steam will not help us make any
       > progress in this area, you know.

Sorry, I meant no offence there.
I just wanted to say that, if possible, we should try to find
workarounds for non-file buffers.
Ideally, these workarounds should work in all the non-file buffers
without a need to address the issue for each buffer type separately.

Eli Zaretskii <address@hidden> writes:

>> From: Ihor Radchenko <address@hidden>
>> Cc: address@hidden, address@hidden
>> Date: Fri, 26 Oct 2018 16:05:25 +0800
>> 
>> > Not in all cases, but certainly in some.
>> 
>> I can say the same about the whole long lines problem.
>
> Would it help to "say the same"?
>
> I'm trying to come up with ideas to help people get bearable
> performance in practical use cases.  I can shut up if people don't
> want to hear about partial solutions that don't work in all cases.
>
> Assigning blame and/or venting steam will not help us make any
> progress in this area, you know.
>
>> > How about asking Org developers to do something about these cases,
>> > like not using the entire buffer as a literal string argument, to
>> > avoid such problems?
>> 
>> Well, in org-mode, the buffer is parsed into s-exp containing all the
>> buffer elements and the associated text, which can be even longer that
>> the buffer string itself.
>> So, the long line is what the parser returns.
>> This approach is a part of the core implementation of the org-mode and
>> cannot be changed easily. 
>
> Then perhaps we could find a solution specific to backtrace buffers,
> like displaying ellipsis or a special button instead of too-long
> lines.

-- 
Ihor Radchenko,
PhD Student
Singapore University of Technology and Design,
8 Somapah Road Singapore 487372
Email: address@hidden, address@hidden
Tel: +6584017977

Attachment: signature.asc
Description: PGP signature


reply via email to

[Prev in Thread] Current Thread [Next in Thread]