Subject: Re: [xsl] GByte Transforms From: Kevin Jones <kjones@xxxxxxxxxxx> Date: Wed, 2 Jun 2004 22:33:35 +0100 |
On Wednesday 02 June 2004 20:26, Niclas Hedhman wrote: > Small comment only; Doesn't the these kind of transforms have to be so > basic, that it makes more sense to do them directly in a SAX application?? Up to a few months ago I would have agreed with you. I am now more viewing this as just another case of the tools eventually taking over what we used to coded manually. The processor can provide you with facilities that you just would not have time to write yourself, like full XPath, making coding easier. The cost of this used to be a memory & CPU hit over a manually coded SAX solution. We are attempting to remove memory from the equation which leaves performance as the the only advantage of the SAX based approach, if it follows previous models, the advantage will reduce with smarter XSLT compiler code. Getting to the point of replacing a custom SAX handler by an XSLT processor is probably a way off. I think extending the range of what problems people can attack with XSLT is what we are really interested in, but doing it in a predictable way is an issue. > > Cheers > Niclas Thanks, Kev
Current Thread |
---|
|
<- Previous | Index | Next -> |
---|---|---|
Re: [xsl] GByte Transforms, Niclas Hedhman | Thread | Re: [xsl] GByte Transforms, Jeff Kenton |
RE: [xsl] encoding shift_jis into a, Josh Canfield | Date | Re: [xsl] encoding shift_jis into a, M. David Peterson |
Month |