[RFC] SRS Section 2

Jeremy Huntwork jhuntwork at linuxfromscratch.org
Wed Feb 2 10:59:44 PST 2005


Hui Zhou wrote:
> On Wed, Feb 02, 2005 at 12:44:46PM -0500, Jeremy Huntwork wrote:
> 
>> Ok, so are we any further with this section? Here's the issues as I 
>> see them:
>>
>> 1. Are we using nALFS's DTD are are we creating a new one?
>>
>> 2. I definitely think there should be *some* type of validation before 
>> the server even tries to parse the XML. As far as I'm concerned, there 
>> really is no question about that (or about our utilizing XML for 
>> profiles). So, should that be done with RELAX-NG? It seems the LFS 
>> Book is headed in that direction, so it may make sense to follow that 
>> trend. Personally, though, I don't know enough about RELAX-NG to make 
>> that call.
> 
> 
> Yeah, only valid reason stated here is the "trend". No question that XML 
> has been, is, and will be the "trend". :)


Hui, you're starting to annoy me.

Re-read what I said. I applied the word 'trend' to the use of Relax-NG 
in the LFS book. And I said I don't know anything about Relax-NG, I 
asked others to fill me in on why they feel it's good.

As far as the other decisions, including the use of XML, I and others, 
through this and other threads, have clarified why we're going in those 
directions. And we have given you valid reasons - it seems you've just 
got it stuck in your head that we are trying to build a tool that is 
perfect for you individually. In reality, the tool may not be perfect 
for any of us individually, or meet all the ideals we as individuals set 
on an automated parser.

The way I see it, we are trying to build a tool that will be as 
generally useful as possible, and in a way that will allow us to be 
flexible and extensible.  It feels like you're missing that rather big 
point.

--
Jeremy H.



More information about the alfs-discuss mailing list