[dev] Suckless word processing solution?

Hi, all!

I am looking for a lightweight solution to create rich formatted content in any MS Word-editable format - I think RTF is more likely to happen, since it is an open format.

Basically, what I am looking for is a lightweight, distractionless (preferably no menus or toolbars) word processor with support for all common formatting option. Basically, I don't need the word processor to be capable of doing things I wouldn't be able to do with a real piece of paper - which means that I just need common beautifiers, font styling, paragraph styling and footnotes/TOC support, tables and images, since I often need to write academic papers following a certain style. However, I don't need any math support.

I was looking for something that would suit my needs for quite some time. AbiWord is bloated, slow and buggy and has numerous GNOME and other dependencies. Ted got updated recently, which means that it finally got UTF-8, gtk interface and proper font rendering, but at the moment is still buggy and it is uncertain, in which way will it improve. Other than that, it is a really nice word processor. WordGrinder has a nice interface concept, but CLI interface can show many style features of the text, which is why WG supports only a limited number of formatting capabilities. Besides, it can only export to troff and html. Finally, it wasn't updated since late 2008.

Then I looked in the area of humane markup languages. txt2tags is nice (the syntax is really clean and easy), but doesn't support RTF and footnotes (not to mention the proper paragraph formatting). I tried MarkDown extensions, such as pandoc (which involves having a Haskell infrastructure installed) and multimarkdown. Both support RTF export, but still look more like an easier way to get HTML output than a word processing solution. The same is true with the other lightweight markup languages: they are either tools to get HTML source, or an easier way to produce man pages.

Finally, i started looking at the full-blown typesetting systems. I admire LaTeX, but it's just too big for my needs. Besides, latex2rtf utility wasn't updated for quite some time and still doesn't work properly. Then I tried lout. Lout is nice, because it's small and has a pretty straightforward manual, but it only supports PS and PDF(?) output. Besides, I had some issues with producing texts in Russian (since it is my native language). Then I tried Groff. Groff look uber-geeky and traditional to me, it is smaller than LaTeX (bigger than lout, though), but there are still a lot of problems here. First, there is a huge lack of documentation - basically, there is only a Unix Text Processing textbook back from the late 80s (and it's not clear as to whether one could use it as a guide to contemporary troff). Second, groff devteam seems to be more focused on the needs of man writers (which is understandable). Which is why many issues specific for common word processing and desktop publishing are ignored or are being solved really slowly. Specifically, I couldn't solve the localization problem. Furthermore, troffcvt utility (a troff converter, supports RTF) is also deprecated and is of inferior quality - basically, it just ignores many formatting options. I also checked other implementations: Heirloom project might be nice (at least, it is said to support UTF-8 and modern fonts), but again it is unclear as to which documentation should I use. Besides, the project wasn't updated since April 2008. There is also a new C implementation called mdocml (designed by BSD people to replace groff), but it only supports man macros (although it is pretty active and should run on Linux too). Furthermore, there should be another flavour in MirOS BSD source tree (which is said to be an original AT&T version), but it is actually broken. Finally, I couldn't find any mention of Plan 9 version of troff being used outside of Plan 9 itself (but I suppose it should definitely support UTF-8).

Currently, I am really desperate. IMHO, there were always two main problems for those, who wanted to build a lightweight Linux/BSD environment: there were no lightweight graphical web browsers and no lightweight word processors. The situation with web browsers gets improved by surf and uzbl developers. But what about word processing? Do you have any suggestions on the original problem?

P.S.: Sorry for this post being so enormous, but I wanted to sum up my efforts for somebody who would like to solve the same problem.

--
wbr, Илембитов
Received on Tue Sep 22 2009 - 15:05:30 UTC