On 12/14/09 18:57, Volker A. Brandt wrote:
[snip]
As for trying to count the number of people who "know" one
language or
another, well that's a specious argument in this context. XML
syntax
might be easy to know, and a given XML DTD might be simple to
understand, but SGML style syntax overall is a _horrible_ syntax for
both humans and computers to use.
I respectfully disagree. I find it not so bad, and computers don't
care.
While computers do not care, XML processing is heavy, given all the
tools required to parse just one simple XML file together with its
DTD/Schema.
I have to manually tweak XML config files at $DAYJOB on a weekly
basis, used in an environment that generates a configuration,
Makefiles and ROM image for embedded devices. It is unbelievably
heavy, and inconvenient to edit. While I understand that some XML
files can be structured in a human readable and editable way, the
vast majority of the files I have seen are just a pain.
Especially when you throw in markup languages like YAML or JSON
which are, IMO, much easier to read than plain XML.
Anyway, I hope you won't end like HAL and its fdi files. While
xorg.conf is not something expected to be understandable easily (way
behind sendmail anyway ;) ), it was easily readable. Now that X.org
uses HAL, we didn't win on the "understandable" side, but IMHO,
regress on the "readable" one.
I think the point also is that sysinst(8) _really_ NEEDS an embedded
scripting language anyway
+1
Why does everyone seem want to shout "XML" in answer to such a wide
variety of questions when that isn't even a proper or complete
answer to
any possible question!!?!?!?
Maybe because "everyone" is not interested in a "pure" way for its
own
sake, but cares more for established standards, easy
interoperability,
and accessibility for the non-guru level user.
"When all you have is a hammer, everything looks like a nail" syndrom.
What about JSON or YAML then? Do you consider them as "established
standards, easy interoperability, and accessibility for the non-guru
level user" ?
Anyway, I would have liked to see a "parameterized sysinst + native
packaging tool + system packages + automated netboot" installation
system
for NetBSD. It would have gone a long way towards acceptance of
NetBSD
outside the "hardcore" user community.
I very much agree.