@gnome
#GNOME powers libxml2 powers xsltproc powers #Docbook automation for Ada & Zangemann #FSFE book translation powers the next generation of #FOSS hackers https://git.fsfe.org/FSFE/ada-zangemann So yeah, donations matter! #AdaZangemann
@gnome
#GNOME powers libxml2 powers xsltproc powers #Docbook automation for Ada & Zangemann #FSFE book translation powers the next generation of #FOSS hackers https://git.fsfe.org/FSFE/ada-zangemann So yeah, donations matter! #AdaZangemann
I'm excited for the future of this project! Developed by #paligo
I tried validating #Docbook v5 using #xmllint from #libxml2 via #RelaxNG and #Schematron but wasn't successful. The RNG validation threw unexpected errors and the Schematron validation threw an internal error. It could be my source file, but it seems fine. Does somebody here have a working setup or tips to share?
The more I delve into #Docbook and #XSLT the more I start to love #XML As a source format it can be very expressive and customizable, plus you can reuse standard elements and attributes from others. Follow my journey here: https://git.fsfe.org/FSFE/ada-zangemann/issues/13#issuecomment-36782
This opposition of org-mode vs. (LaTeX, SGML) XML in publishing is something that resonates very strongly with me - though I'm afraid it's hard for many org users to understand.
I do almost all of my daily work in org-mode, but whenever I start a writing project that needs to be published, I use XML. There are all kinds of reasons for this, some very specific to my case of academic writing, where the handling of citations, footnotes and bibliography can get very complex and specific. But I feel that in the end it comes down to something much simpler.
For decades I have been using docbook-xml for all my writing projects, starting with drafting in asciidoc (and this co-existence of asciidoc and org-mode as two "markdown" dialects still leaves me unsatisfied). The deeper reasons however seem to be that in org-mode you start designing your text from an outline. For me, this is the wrong approach, as I need to write a text as a stream-of-consciousness, adding paragraph after paragraph, and only later get to an outline. The way I'm used to using org-mode seems to make this impossible for me.
I want to spin up replacement mailing lists for #DocBook users now that they won't be hosted by OASIS anymore. (Nevermind that IT chaos at OASIS means none of the lists are working very well.) Is there an ethical open source collective sort of place that will let me pay them to do this? The lists are free for users, of course, and I don't generate any revenue from them, so it would have to be a not too expensive place.
fwiw, I used #DocBook #SGML in #Emacs to generate LaTeX/LilyPond, even when Macmillan demanded I submit MSWord¹ docs, I'd just emit them as an output! Mind you, that was last century, but I'd still do it that way today, if I had to.
___
¹ I call those apps "font painters', or WYSIAYG (what you see is ALL you get
OK, so my fantasy #worldbuilding / #writing project notes live in a massive #Scrivener project. 164,686 words.
Now, *editing* the thing in Scrivenever is fine and good and I have no problems coming up with new stuff and just putting it there. However...!
The export process is kinda janky. At current state I can't export this to HTML, for starters. If I want a usable PDF document out of this (compiles into a 344 page document, half of which isn't even formatted right because I'm lazy and it looks OK in Scrivener side), I need to compile it into a LaTeX file via MultiMarkDown. Involves a little bit of hand-fiddling too.
It's *hhhella slow*. It's *janky*.
I really would like this to be a much more easily editable and compilable thing. With fully text-based markdown(esque) input, able to export to static HTML and PDF (via LaTeX, preferably).
Lots of options to choose from. I'm currently looking investigating #DocBook solutions. Am I mad? ...I am mad, aren't I?
Tonight is the first DocBook TC meeting since 5.2 was promoted to an OASIS Standard. Someone joked there should be bubbly. One does not joke about bubbly. Not even when it's served in a novelty glass! #DocBook #potentpotables #ducks #Photography
Although it's easy to sympathise with this criticism, since there's no doubt that #DocBook is a beast and can drive you mad, this rant is also caused by the author's perspective.
If you use #DocBook as a source for producing man pages, it reminds me to learning #org-mode for writing your shopping list. Man pages have a very rigid structure, but the main strength of docbook is its flexibility. And with this flexibility comes ambiguity and redundancy.
If you look at the discussions between the people involved, it's not that they didn't see the problems, but it's about how #DocBook evolved. It's true that it was originally designed for writing "technical documentation". But what is technical documentation? My main use of it is for writing projects that are not "technical" in nature - but where the structure is often unpredictable. And as much as I've wanted to get rid of DocBook, I keep coming back to it because it allows me to even misuse it for purposes it wasn't designed for.
I can see a parallel there for why I like #Perl (which most people seem to hate as much as XML in general). I'm not a programmer, and I don't care if you have trouble understanding my code. But it gives me the ability to do things my way without having to rebuild everything from scratch. For me, these tools have certain values that I miss in most of the others out there.
And to make things even more complicated: I love #OpenBSD and hate man pages ;-)
@ndw Why the #DocBook
Stylesheets (in oXygen) not include <result> and <remark> under <procedure>?
Wrote my own customization templates for d:result and d:remark, but me and my students runs in each course in the error trap.
See: https://social.tchncs.de/@xmlarbyter/110326832727060473
@lobrien I have seen this technique used for tutorials of #Java frameworks. I have also seen build tools distributed as containers. Back in the day I would have loved a containerized #DocBook build environment for PDF and HTML because the default toolchain on Linux would invariably fail due to the many incompatible/brittle stages involved.
@donwatkins
I also love #DocBook, even if I'm (still) struggling in setting up a proper toolchain in Windows WSL (not much a Linux expert anymore).