Re: [dev] (x)HTML-based office suite? (aka suckless word processing solution-2)

From: Ilya Ilembitov <ilembitov_AT_yandex.ru>
Date: Sun, 18 Oct 2009 19:49:48 +0400

Yes, you got it totally right. But why do you believe that html would be on par with proprietary formats? I mean, is there any possibility that html produced by markdown (or similar languages) interpreter won't be displayed correctly in most of browsers (including IE)? I thought it's basic and standartized enough. I do agree, though, that S5 is likely to have some issues.

> 2009/10/18 Илья Илембитов <ilembitov_AT_yandex.ru>:
> > I believe I made some progress with my little research concerning suckless word processing.
> > Basically, I realised that things get much easier as soon as we realize that there is no need for paper-centric formats.
> >
> > First, there is no standard here, that is correctly supported by all office suites on all platforms: ODF (as well as OASIS) is not getting any decent popularity any time soon, common MSO formats often lack needed compatibility (even now when I have to bring .doc files created by OpenOffice to be printed on a Windows machine with MSO I get my formatting partially lost).
> >
> > Second, there are lots of restrictions being imposed by those formats. They all employ the paradigm of paper-centric documents (documents that are created to be printed), whereas lots of documents people create today (I would say, almost the majority) won't ever be printed and are actually intended to be viewed and edited only between computers: be that sharing over the Net/email, passing them on portable media, etc. Thus, one could actually switch from paper-centric paradigm to screen-centric paradigm, which changes the way we think of formatting - e.g. we don't need page numbering, the references, links and TOC appear in a whole different form. Many other changes come here.
> >
> > For documents that will actually get printed one could probably employ something like LaTeX/lout/troff/etc - and that's the whole different story.
> >
> > Then I learnt about the fact that Opera employees actually use HTML for all the documents they share within company. Opera is actually interesting due to the Opera Show format, which is an approach to creating presentations using HTML/JavaScript/CSS only + some easy markup. The only drawback (AFAIK) is that Opera Show is supported only by Opera Browser (there should be addons for Firefox, though).
> >
> > Then there is S5 with a similar approach (and I believe, better compatibility - not sure about IE6, though).
> >
> > But what about word processing and spreadsheets?
> >
> > Word processing is easier, since there are lots of lightweight markup languages (MarkDown being the most famous ones) that support all the basic formatting and produce an XHTML output. Here one could see what I called a screen-based approach: when I define the formatting for my text, I actually define the structure of my text, instead of defining the beautifiers (bold/italic/underline) or font sizes manually - the language interpreter does that for me. Besides, lightweight markup languages are nice here, because usually their tags don't get in the way of a spell checker.
> >
> > Spreadsheets are a bit trickier. I was looking for a lightweight tool to produce plain text files that can be easily converted to HTML. There is SC, which is really nice (only depends on Curses), has VI-like keybindings and produces .sc files. Which are actually just plain text files with a specific formatting, that can later be converted to CSV (and be used in xhtml documents). And I think that it should be pretty easy to convert CSV to any lightweight markup language syntax for tables. The only drawback with SC is that it doesn't have Unicode support (the last update of SC was in 2002, when Unicode wasn't popular enough), so I couldn't get it to work with non-Latin characters. So I am looking for something similar, but more recent. Any ideas here?
> >
> > Second, tools like SC naturally can't produce graphs. I know there are lots of plot utilities, but I couldn't really find what I needed. What I wanted was a small (C, preferably) console tool (so that I could invoke it right from VIM) that could produce images with graphs (or diagrams) using the provided CSV file. That image can later be used within the resulting xhtml file.
> >
> > So, basically, to sum up, this way one can get a pretty flexible single document type that can easily mix text elements with tables (produced by a spreadsheet tool like sc) and graphics (be that images or graphs for the provided data in the tables). This document can be later viewed and edited (provided we send the source files as well) on literally any machine with a web browser (including IE6). And of course it can be easily published on the Web.
> >
> > This way, one can use almost any text editor that can:
> > -highlight syntax (for MarkDown or any other lightweight language)
> > -invoke external tools (for conversion of table files, producing graphs and the final conversion to html)
> > -properly wrap lines and words (which is why nano, being a really nice, easy and tiny editor is not a good solution here)
> >
> > to replace existing office suites.
> >
> > What do you think? Can this approach work in real life - for creating academic papers with non-strict formatting, lecture notes, articles - for sharing with your friends, teachers or colleagues that may or may not have any computer skills?
> >
> > I believe this approach can be easily implemented within Emacs environment (using org-mode and muse-mode, but I would like to implement this idea in VIM (or any other text editor), too - but I need a spreadsheet calculator and a plotter for that. Don't want to cause a flamewar here, of course.
> So reading your mail can be summarised as a 2x2 matrix:
> WPS WPP
> SPS SPP
> WPS means word processor screen presentation, WPP means word processor
> print presentation, SPS means spreadsheet screen representation, SPP
> means spreadsheet print presentation.
> Ok, as far as I understand what you are proposing is WPP and SPP are
> mostly not interesting and can be avoided in 90% of all cases.
> For WPS and SPS you come up with HTML as the presentation format +
> some source tools right?
> I mean that can work, but is not any better as using propietary formats.
> To sum up my view: I think there is a lack of decent WYSIWYG word
> processors (for both, screen and paper presentation), but MS
> Word/Powerpoint is still the leader in this area, and there is a lack
> of usable spreadsheet processors (again MS Excel is the best one can
> get).
> I agree there should be good replacements, but I don't expect them to
> appear in the OSS world very soon, and I pretty much doubt we at
> suckless can produce something good in the short term.
> Though I'm open minded to see what people use and why they think
> that's good. Your proposals however are just HTML imho, and I'm not
> very convinced about that, not because I think HTML sucks, but because
> HTML is a moving target and never settles and will never settle, and
> your documents will look totally different every once in a while (esp
> if you enrich it with CSS and JS).
> Kind regards,
> Anselm
>

--
wbr, Ilembitov
Received on Sun Oct 18 2009 - 15:49:48 UTC

This archive was generated by hypermail 2.2.0 : Sun Oct 18 2009 - 16:00:02 UTC