On Wed, Aug 01, 2001 at 11:26:11PM +0200, Asger K. Alstrup Nielsen wrote:
> On Wed, 1 Aug 2001, Andre Poenitz wrote:
>
> > A bit. But speed does not really matter. I've never seen a complaint that
> > mathed is slow, and I've never got that impression myself.
>
> I promise you, size does not mat
> Ok, I have 6528 insets, and about 54000 total elements in the insets
> (element = single letter or command). So average length of an inset
> is ~8 elements.
[ With 'inset' you probably mean 'LyX inset' not 'Math inset'? $x^2$ is
one LyX inset and two MathInsets (one for the formula and one fo
On Thu, Aug 02, 2001 at 06:54:50PM +0300, Dekel Tsur wrote:
> >
> > Each math formula is on average less than 1k with all your proposals. Now,
> > in an extreme math thesis with maybe 10 formula on each page, for 500
>^^
> This is a too low estimate. On my
> > I promise you, size does not matter either.
>
> Andre would something similar as we have it for paragraph work for you?
> A placeholder META_INSET it the char array and a map to
> find the right inset? Or is that to complicated?
Probably... there are on average only about two or three items
On 01-Aug-2001 Asger K. Alstrup Nielsen wrote:
> On Wed, 1 Aug 2001, Andre Poenitz wrote:
>
>> A bit. But speed does not really matter. I've never seen a complaint that
>> mathed is slow, and I've never got that impression myself.
>
> I promise you, size does not matter either.
Andre would som
On 02-Aug-2001 Allan Rae wrote:
> On Wed, 1 Aug 2001, Asger K. Alstrup Nielsen wrote:
>
>> I promise you, size does not matter either.
>
> That's not what you were saying about cucumbers at last years developers
> meeting.
I can remember that to!
> Perhaps, cucumbers are the exception.
I see
On Wed, 1 Aug 2001, Asger K. Alstrup Nielsen wrote:
> I promise you, size does not matter either.
That's not what you were saying about cucumbers at last years developers
meeting.
Perhaps, cucumbers are the exception.
Allan. (ARRae)
On Wed, 1 Aug 2001, Andre Poenitz wrote:
> A bit. But speed does not really matter. I've never seen a complaint that
> mathed is slow, and I've never got that impression myself.
I promise you, size does not matter either.
Each math formula is on average less than 1k with all your proposals. Now
> Alternative 1: "Everything is an Inset" - the expensive and clean way:
>
> 0 4 8 12
>
>
> makes 12 bytes + 12 bytes for the vector + 3 new insets * ~50 bytes per
> inset. Makes ~175 bytes.
Read this as:
16 bytes in the vector + 12 bytes for the vector
