OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

docbook-apps message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]


Subject: DOCBOOK-APPS: (hopefully easy) Problem with Jade, OpenJade and/or NW stylesheets


Hi all, 

Great list, this is my first post.

We recently decided to dive into DocBook for our documentation needs,
and went about collecting up the needed free software.  A separate
document detailing our experiences with getting a DocBook system going
from scratch (on Unix machines *other* than Linux and FreeBSD) will be
posted once it's complete.

Before I can get there, I'm having a strange problem that's causing me
some consternation.  This has *got* to be something stupid that I did,
but for the life of me I can't figure it out.  I'm definately going
to look dumb when one of you whips out the answer quickly, but that'll
be better than me floundering around and wasting more time on it.

First, the background:

=-=-=-
Sparc Ultra 5 running Solaris 5.7
Compiled all the various stuff with gcc 2.95.2

Docbook 4.0 SGML and XML (docbook.zip and docbookx.zip) unpacked into
a work spot.  (/usr/local/tmp/dbtmp/)

Mr. Walsh's style sheets (DSSSL and XSL (db154.zip and dbx113.zip))
unpacked under that directory too (they unpack into their own
"docbook/" subdirectory).

copied '/usr/local/tmp/dbtmp/docbook.cat' to '/usr/local/tmp/dbtmp/catalog'

Add 'SGMLDECL "docbook/dtds/decls/xml.dcl"' line to 
/usr/local/tmp/dbtmp/catalog  (per "catalog trickery" on pg. 84 of the book)
=-=-=-

Created a quicky little XML docbook document to test stuff.  Everything
worked like a champ, running Jade 1.2.1, and this command:

(cwd /usr/local/tmp/dbtmp/docbook/print)
jade -t tex -d docbook.dsl ../../firsttry.xml

Would generate the .tex file, and I could get pdf out of it with the
other tools.  HTML worked similarly (but in the other directory).

OK, all's well.  Now, I hear about "OpenJade" and think "Ah!
Somebody's taken up active development, I'll try the current version."
After getting the latest CVS checkout to build on Solaris (no small
feat, more details later), I try the same command, and get this:

=-=-=-
openjade:docbook.dsl:1:73:W: cannot generate system identifier for public text "-//James Clark//DTD DSSSL Style Sheet//EN"
openjade:docbook.dsl:247:0:E: reference to entity "STYLE-SHEET" for which no system identifier could be generated
openjade:docbook.dsl:1:0: entity was defined here
openjade:docbook.dsl:247:0:E: DTD did not contain element declaration for document type name
openjade:docbook.dsl:150:0:E: notation "DSSSL" for entity "dbl1pt" undefined
etc...
=-=-=-

"That's odd, it worked before."  I exclaim.  No amount of poking,
prodding, and kernel tracing will tell me why it used to know about
that sysid, and now doesn't.  I tried various combos of the -c, -d and
-D options, with little to no affect.

Ok, I figure I've just screwed up the OpenJade compile/install somehow,
and since we had so much trouble with it, I figure I'll just go back
to using Jade until I figure that out.

Of course, I switch back to Jade, and *get exactly the same problem*.

"Hunh, that's weird," I think, "maybe the OJ install mucked something
up with Jade too.  I'll try a different machine."

I install Jade and all the stuff on a new machine (this time BSD/OS)
and...  have exactly the same problem.

This is just weird.  I swear it worked once, just as I've explained
it.  I even have the HTML and PDF output of my XML source to prove it,
but damned if I can get it to work again.  The only thing I can figure
is that the previous Jade installation that I was using was compiled
by somebody else, possibly (probably?) with different options.  Perhaps
I missed something during my builds that had been included before...?

Help.  And please don't make fun of me when you tell me just how
stupid of a "problem" this turns out to be ;-)  (on second thought, you
can make fun of me, I can take it, I just want this to work.)

Thanks in advance,

Sean Ross
Corvallis, Oregon, USA
rosss@research.oneworld.com


[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [Elist Home]


Powered by eList eXpress LLC