I may be misapplying the concept of “Clean State Changes” but most of Luke’s points seem to be beneficial consequences of clean state changes. Size, Compression, and Resilience come because there is no extraneous or over-encoded data. Parsing and Search are side-effect free state transformations, the output of a search is a new subset of the file for example.

Privacy and Clarity are kinda linked concepts, private data can be included in a plain text file, but it can’t be hidden from most plain text editing programs. Future Compatibility is pretty much just another flavor of Clarity. These are more about the clean state changes inside the users thought process as they reason about a given file.

Of course, until you wrote about it I hadn’t really been able to express why I switched to Notepad++ for all my writing as soon as inline spell check showed up. I’ve pretty much always understood that there was a lot of stuff lurking in those Word files, but never connected that with my dislike of the beast, since I never try to do anything complicated with it.

Hi Luke&errbody,

you might want to add another advantage of plain text (it may be considered as a manifestation of resilience you already mentioned, but i consider it important enough on its own): recoverability.
Several months ago i occasionally wiped out my home folder. I had a backup, but it was 3 weeks old, and there was some important stuff created during that time, so i turned to data recovery; surprisingly, my data appeared to be more fragmented than i expected (i suspect that’s because my laptop has an SSD). I had luck and recovered some pictures and some spreadsheets, but there was a significant amount of corrupted binary/XML files – on the other hand, when it came to text files, i was able to rescue information even from the ones that were partially overwritten.
Actually, since majority of recovered files lost their names, i had to grep their contents to find some data i needed. I was quite surprised – it worked fairly well. The worst problem when searching was… lots of xmlish garbage showing in results of the search ;)
Oh, and don’t forget about duplicates! Since SSDs copy all files around, i got up to 10 (maybe even more?) duplicates of some files. In case of plain text ones, it was easy to clean this up, but the other ones…. Some still wait unexamined.

[…] Why Plain Text (Terminally Incoherent) […]

[…] I needed was a format to keep my spreadsheet data in plain text. I used CSV (comma separated values). I exported my current milage data as a csv file, examined the […]

[…] Why Plain Text (terminally-incoherent.com) Share this:EmailTwitterRedditMoreDiggLike this:LikeBe the first to like this post. Filed under: software   |  Leave a Comment Tags: Autokey, Keyboard shortcut, text expansion, typing […]

Word would be so much better if it just offered a ‘source’ view where you can edit all the formatting tags just like you can for Latex or HTML.

@ IceBrain:

Well, look at that! Lyx 2.0 was finally released. It took them quite a while. I remember seeing that wiki page back in 2010 but 2.0 was nowhere near release then. Hell, I think the first feature requests for inline spellcheck in their bug tracker date back to like 2004 or something.

Ah, the joys of Open Source. :)

@ Morghan:

See, I like the kind of cloud that “syncs up” with local file system – like dropbox. It gives me a redundant copy somewhere else. By keeping data in both places I minimize possibility of loss. If I fuck up, the cloud can bail me out. If cloud fucks up, I have local copies. Likelihood of both me and the cloud fucking up at the same time is small – not impossible, but smaller than the probability of individual fuck up on either side.

I always find it amusing to see all the HTML code in my emails. I have my client set to text only as GNUPG integration for HTML messages was buggy. There is often more formatting and image links than actual content in messages.

It’s not just programmers. I know a little BASIC and some outdated HTML from playing around back in the day, a bit of bash scripting, and some LaTeX but I’m definitely not a programmer. It comes down to be same reason I prefer my own server to using the cloud. If my stuff gets lost or screwed up I want to have only myself to blame.

Trusting someone else with my data and its security is really not something I’m comfortable with. I do have some services, like my Kindle, but I crack the DRM and back it up to my server and external media as soon as I buy something from them.

Plus, Lyx has no inline spell checker.

Apparently, they fixed that recently: http://wiki.lyx.org/LyX/NewInLyX20#inline-spelling

@ Victoria:

Well, editing DLL’s is a nightmare no matter how geeky you are. But you bring up a very interesting tangentially related topic: how to store application configuration. Hard coding settings is obviously bad, but what is a good format to store your config files in?

XML? YAML? JSON? Generic config format?

@ Greg, @ MrJones2015, @ IceBrain, @ Parminder:

Good point about Lyx. The wonderful thing about it is that it gives you access to full access to the LaTex source it auto-generates while at the same time giving you a WYSIWYM style abstraction layer:

http://i.imgur.com/6U08m.png

It’s a good compromise I guess but most of the time I prefer to write my own LaTex by hand. Plus, Lyx has no inline spell checker. Like none whatsoever. In the day and age where my browser has that feature I consider it a huge flaw.

@ Chris Wellons:

20KB? The average size of an email here is close to a meg – including the three color rich text email footer and the obligatory embedded images and backgrounds. :P