[ntp:questions] Re: Latest changes toTWikiGettingStarted

Peter Boettcher boettcher at ll.mit.edu
Tue Oct 7 18:24:52 UTC 2003

"David L. Mills" <mills at udel.edu> writes:

> Just so I understand the itsy bitsies here, can you tell me specifically
> what got broken or repaired in the "old" and "new" HTML formats? The
> tools here evolved over three eras. First everything was done by hand
> and there were lots and lots of mistakes, inconsistencies and just plain
> junk. The next era was using tidy and a lot of these things got fixed.
> The current era uses GoLive and a few formatting glitches got fixed, but
> not a major change in style or content. Frankly, the value to me of
> GoLive is to insure consistency and connectivity in my total web
> collection, of which the NTP collection is only a small part.

Sorry about that previous comment regarding "crappy HTML" from GoLive.
It actually produces amazingly vanilla and nicely structured HTML.
The trick is that any conversion program needs to guess at the intent
behind much of the HTML, so my hack, at least, has a bunch of
heuristics that seem to produce good (?) man pages, most of the time.

The older HTML had a bunch of inconsistencies that needed more hacks
(missing /dl tags, present or absent /dd and /dt, etc).  The new
system is prefereable, but involves some growing pains as the script
must be changed (I think it's better now, or will be).

> I was persuaded to use style sheets on the expectation, I was told, that
> folks could jimmy up their browsers to produce whatever font, color and
> layout scheme they preferred. That's why I used the absolute minimum
> font, color and style adventures. However, it may well be that some
> little thing on my part will significantly simplify conversion to other
> formats. I am certainly warm to suggestions.

Definitely, it's nice.  If style-sheets support stuff like this, it'd
be nice to add a few more logical elements, like use (pseudocode)
<command option>stuff</command option> instead of <tt>stuff</tt>.  But
I have no idea how that stuff actually works.

Peter Boettcher
MIT Lincoln Laboratory

More information about the questions mailing list