Cart
[Home] [By Thread] [By Date] [Recent Entries]
On Sunday 19 January 2003 18:32, Mike Plusch wrote: > XML does not have to be verbose in the future. > > It _could_ be suitable for ALL applications. > > It is desirable to have a common, standard syntax > that can be used for ALL purposes. > > _Mike Plusch > > > How about changing the name to: > Conci-sex-ml. :) Now you're talking my kind of language; check out ASN.1, which solved this problem a zillion years ago. You have a single type system at the core of it, then multiple encodings of values of those types, from portable byte streams for interchange to direct representation (access!) as C structs, Pascal records, Java objects, etc. There's a textual notation for debug dumps and hand-editing, and there's a shiny new XML notation too. Since all of these encodings are formally defined in terms of the ASN.1 type system, automated software can map between them trivially, so you can just use whatever encoding is appropriate for the situation at hand. The counterargument is that with XML the set of software needed to edit it is likely to come with your OS rather than something you have to download seperately, helping a sysadmin who has to edit an XML configuration file in single user mode on a broken system to make it boot again, who therefore couldn't download or install a BER editor, and hadn't had the foresight to install one beforehand, even knowing that one might be needed to recover the system later... *shrug* I like to optimise for the common case myself. ABS -- A city is like a large, complex, rabbit - ARP
|
