3 ms·
Is this as staggeringly naive as it looks? Some of the people responding are steering the discussion to a layout language with a preview window. I don't know i
by Zigurd 13y ago
Is this as staggeringly naive as it looks?
Some of the people responding are steering the discussion to a layout language with a preview window. I don't know if they are doing it because they prefer to work in such a user-hostile mode (I did this for a book, in Eclipse. Ugh.), or if they think this is a more sane goal.
WYSIWYG has its own issues. most users of word processors have no idea that paragraphs are objects in an object model, but the command structure only becomes clear when you realize that. Most users just hack at a document it until it looks right enough. At the really diabolical end of the spectrum I could show you an Ericsson documentation template that manages to manifest dozens of bugs in Word, laying in wait to eat your previous hour's work. I'm sure you have inherited documents like that.
It's all more or less a kludge, and WYSIWYG never is quite, nor is it real direct manipulation. At best it is something like "moderately friendly visual document CAD, if you get the trick behind the slick appearance."
- specialist 13y agoa layout language with a preview window aka digital typesetting http://en.wikipedia.org/wiki/Typesetting#Digital_era http://en.wikipedia.org/wiki/Typesetting#Digital_era If not quite the Correct Answer, it might be the Best Available Answer. My mother used to be a Linotype ninja, blazing thru the sunday paper's ad inserts. The user input language was a page description language, little different from PostScript or HP's PCL (or HPGL/2). While modern systems are more accessible, I've not seen a system since that matched that concision and productivity. I loved the precise control of WordPerfect (reveal codes) and FrameMaker (parametric styles). I can't comment on modern InDesign (stopped around version 2), but early on it was not considered a feature complete replacement for FrameMaker. People speak highly of LaTex. I've not used it for real work. I've done my share of UI, scenegraph, layout engines, plotter/print drivers, prepress / imposition, etc. I gave up trying to integrate general purpose constraints systems into layouts. Too many edge cases, hard to debug. I now believe layouts heuristics should be implemented imperatively. So I created DesignGridLayout, which captures grid based rules for UIs in a "fluent" API (aka method chaining). https://designgridlayout.java.net/examples.html https://designgridlayout.java.net/examples.html (The awesome Jean-Francois Poilpret has maintained, extended the project for years.) Though the box model sucks, I haven't yet given up on some mythical, simple scenegraphs + typography + content hybrid. I still don't know what that'll look like.
- informatimago 13y agoI would not say it is naive. I wonder if it is an itch Stallman has had for 25 years, but it is certainly an itch I have. Not for me, I would do perfectly well without word processed documents, but for customers and coworkers, they like PDF if not powerpoint or MSWord documents. For now, I have edited and generated them with reStructured Text, but the workflow is not entirely satisfactory (and I don't feel like dipping into python to improve things there, I would rather write rst document preprocessors in lisp when possible). Another aspect is that if we improve the graphic abilities of emacs, it may be easier and faster to implement chrome like lighttable to attract newbies. Sublime Text probably has capted a percentage of users who would have chosen emacs, if it could have competed in the esthetic plane. Also, it is a little regretable to be limited to ascii art for the few diagrams that are sometimes needed in the documentations we write as programmers. If we could edit vectorial graphics in emacs, it would be great (we can already render svg, but perhaps thru an external program). So I would not say naive, but on the contrary, it may condensate a critical mass of itches to motivate some scratches.