Message-ID: <4C91254D.50209@morningstar2.co.uk>
Date: Wed, 15 Sep 2010 20:58:05 +0100
From: Joseph Wright <joseph.wright@morningstar2.co.uk>
User-Agent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.6; en-GB;
	rv:1.9.2.8) Gecko/20100802 Thunderbird/3.1.2
MIME-Version: 1.0
To: LaTeX Team <latex-team@latex-project.org>
Subject: Re: Allocation of registers
References: <19109.1284579373@cl.cam.ac.uk>
	<4C9122A1.4060008@morningstar2.co.uk>
	<4C91239B.2040301@morningstar2.co.uk>
In-Reply-To: <4C91239B.2040301@morningstar2.co.uk>
Precedence: list
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain; charset="windows-1252"; Format="flowed"
Sender: latex-team-bounces@latex-project.org
Errors-To: latex-team-bounces@latex-project.org
Status: R

On 15/09/2010 20:50, Joseph Wright wrote:
> On 15/09/2010 20:46, Joseph Wright wrote:
>> On 15/09/2010 20:36, Robin Fairbairns wrote:
>>>> I guess for the time being that is a good approach
>>>
>>> is there an actual reason that inserts can't start working down from
>>> 64k?
>>>
>>> etex.sty can't do this, because the basic set of inserts has been
>>> allocated by the time it gets to run, and that plainly isn't a
>>> constraint on latex 3. but does it matter for expl3?
>>>
>>> (i have once spent time on thinking this issue through, but i'm not well
>>> -- still -- and can't remember what conclusion i came to.)
>>>
>>> robin
>>
>> At a format level we don't currently have any allocation routine for
>> inserts at all, as we've not used them :-) On the other hand, expl3 is
>> 'LaTeX3 on 2e' so has exactly the same requirements as etex. (Indeed,
>> the whole reason for simply using etex for allocation in expl3 as a
>> package is that it avoided making errors in the interaction with the
>> LaTeX2e allocation routine.)
>
> I also notice that the e-TeX manual says
>
> 'The additional registers, numbered 256=9632767, can be used exactly as
> the first 256, except that they can not be used for insertion classes.'
>
> and also
>
> 'The additional registers are realized as sparse arrays built from TEX=92s
> main memory and are therefore less efficient.'
>
> (Not sure if the later means in terms of speed or in terms of memory use.)

This is all a bit disturbing in a LaTeX3 context, as the current =

allocation approach does not take account of this at all! (Of course, =

this is only an issue when building a format, but one day ...) I guess =

it suggests that we should reserve a block for inserts, and on the basis =

that you can never reserve enough space (history warns us) pushing most =

allocations to the extended pool.

Joseph
_______________________________________________
Latex-team mailing list
Latex-team@latex-project.org
https://lists.dante.de/mailman/listinfo/latex-team