This is the mail archive of the docbook-apps@lists.oasis-open.org mailing list .


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: Questions Concerning Style Sheets


Nothing like a good religious debate ;)

>>>>> "M" == Matthew Harrison <harry@inpharmatica.co.uk> writes:

    M> I only partly agree. Presentation helps people understand the
    M> content and structure of a document. Why can't the writer have
    M> the benefits of presentation as they write? What you see is
    M> what you mean.

Spoken like a true child of print media.  

I know at least two large web content shops who approve their content
designs based on a blow-up poster of a colour laser print set before a
conference table.  As a result, they produce beautiful content ... but
ugly and useless when seen on a monitor, and they then stand around
and wonder why they are not a top-rated site.  They look at Yahoo and
wonder why such an ugly site is so popular. What they see may be what
_they_ mean, but what is seen is not what they meant.

    M> Commercial tools like XmetalPro give you views
    M> of the XML, and the processed version at the same
    M> time. XML/SGML is a good for a file format, but the markup can
    M> be very intrusive when all you have is a text-editor. Even in
    M> emacs with syntax colouring, it can be hard to see the wood for
    M> the markup, er I mean trees.

Agreed ... until you flip the switch to hide the markup.  Then you
have an ascii representation showing all the proper structure, which
is admittedly in need of a measure of imagination to map colours to
meanings and some fixing to the indents, but no more confusing than if
you look at a GUI-rendered page set by the recommended style-guide of
the world's largest trade publisher and see that source code,
functions, variables and screendumps are _all_ courier bold 10pt.

    As a sidebar, I have recently been reworking some of my writings
    done in that old GUI template method, and I have been shocked and
    appauled to see the _large_ number of errors introduced by the
    copy editors and subsequently missed during my proofreading.  In
    having to conceptually re-tag everything, I find, for example,
    commands that should have been filenames, filenames and URLs with
    extraneous periods (sentence ends in a filename), and punctuation,
    capitalization and hyphenation by language rules being imposed on
    critical filenames.  Had this been in DocBook, even copy editors
    would have trouble making these mistakes.

When I last worked in WordPerfect (c.1985), I remember having to
toggle "Show tags" more than once per session, and I find, for
example, the WYSIAYG Netscape Composer is useless for serious webpage
work because I waste substantial precious time trying to guess the
hidden markup.  It is far faster to write if (a) I have hotkeys and
templates for frequent markup patterns and (b) I can _see_ what I am
writing _as_ _it_ _is_, not as some robot _thinks_ it is.

I will admit, if I had the time, I would customize Emacs PSGML mode to
use different fontfaces instead of just different colours, and I would
like to see the text contained within certain tags to be rendered
differently when I hide the tags.  If I had the time, I'd fix TkSGML
(except the authors don't trust me and keep their sources closed) but
I haven't the time for either, so I'm stuck with PSGML mode and my
imagination.

-- 
Gary Lawrence Murphy <garym@teledyn.com> TeleDynamics Communications Inc
Business Innovations Through Open Source Systems: http://www.teledyn.com
"Computers are useless.  They can only give you answers."(Pablo Picasso)

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]