Re: article selected when doing B m

["François Pinard" <pinard@iro.umontreal.ca>]

Robert Bihlmeyer <e9426626@stud4.tuwien.ac.at> écrit:

>  FP> Yet, I would guess that it might be achievable to lazy-render
>  FP> HTML from beginning up to a particular point in a buffer, so it
>  FP> could be rather speedy (once all modules loaded) [...]

> Rendering is quite fast, compared to the loading.  [...]  But there may
> be unfortunate people getting a lot more HTML messages ...

HTML rendering is usually quite fast for me as well.  Yet, sometimes,
I receive those huge HTML messages, and for them only, lazy rendering
would be a blessing, as it currently takes a quite long time on a rather
fast machine.  Of course, I just then `C-g' out of rendering, reload with
`C-u g', and try to read my way through the SGML tags...

> So for my usage pattern, lazy HTML drawing would not bring much gain,
> and the cost is probably very high.

I agree with you that, on average, we may cope with the current situation
for HTML.  My point is more that Gnus, all over, would definitely be more
fun with lazy rendering, as it occurs frequently to me that rendering is
slow (because of frequent big messages).  I was just stressing that _even_
HTML could be lazy-rendered, if we were really aiming for all that laziness,
so to speak :-).  I hope it will be addressed in Gnus, one day.

-- 
François Pinard   http://www.iro.umontreal.ca/~pinard