Unwanted Optimizations

Silviu Julean juleans at arad.ro
Tue Apr 2 08:26:21 PST 2002


> You could use a <stage> element do simply replace a <prebuild> element
> if you wanted to. But it could do more as well.

Are you going to REPLACE the current DTD with <stage>-like tagging or going
to use it in addition, to specify more stage information *if wanted*?

> <stage>
>    <stageinfo>
>       <name>Prebuild</name>
>       <user>lfs</user>
>       <chroot>&LFS;</chroot>
>       <base>/usr/local/src</base>
>    </stageinfo>
>    ... Do some stuff
>
>    <stage>
>       <stageinfo>
>          <name>Prebuild - cleanup</name>
>       </stageinfo>
>       ...
>    </stage>
>    ...
> </stage>

If it's that versus

<chroot dir="&LFS;">
	<prebuild>
	... Do some stuff
	</prebuild>
</chroot>

i'd choose the last one... Why get complicated when the current synthax is
doing the job? And to specify more information,

<prebuild info="Fooing" comment="This performs a foo-ing (using foo-3.2.5 in
/usr/bin/foo) before the actual compilation">

just like the alt="" parameter used in HTML <img> would do.

IMHO your DTD specifies way too many stand-alone tags, when more parameters
could be used. I find the HTML DTD an excellent example, and, as i am using
HTML for nearly 4 years now, i guess i never used a visual editor because
it's so easy to write. And though the new, fancy additions like <embed>s and
<object>s have a somehow nasty synthax, the simplicity making a visually
attractive document can't get any better. The same could apply to your
synthax, if some condesation would be done.

Neven: i'm not here to complain, i know it's been a damn hard (team)work;
all i'm saying is that, thinking of all end-[l]users, this great automation
system could be used by non-LFS users too.

PS: if you're not replacing any tags in the current DTD, just ignore this
(tho my remark goes a bit deeper than that)

-- 
Unsubscribe: send email to listar at linuxfromscratch.org
and put 'unsubscribe alfs-discuss' in the subject header of the message



More information about the alfs-discuss mailing list