We Need More .txt

Everytime someone requests me to send my resume or a small document in .doc, I remember the long forgotten, simple yet versatile format – the text file. While we have been fighting over standardization of office documents, we have neglected it enough to make our lives difficult by hovering around resource-hogging proprietary tools and formats even for most basic tasks. The text file is like a meta format (though format is not exactly correct, I assume it is enough for our purpose), that lets you easily create your own schema and use any extension to build your own type using it. Look at the list, if you do not believe me. Here are some of the reasons, it should rule our GUI world, as it did the command line interfaces.

  • It is quite light on the hard disk space, and consumes only a fraction of what a corresponding office application document would use.
  • It is light on the network bandwidth too. it is one of the ideal file formats to use while sending information over a network. It is so easy to use a text file attached to an email.
  • A text file carries only characters, and reduces the danger of spreading viruses.
  • A text file is a basic format in all the operating systems, and hence their editors are inbuilt. No need of third party software to use the text files. Isn’t it cool that you are independent of a tool to compose, edit or read a file!
  • A text file is compatible across platforms, barring the end of line character, but they can be easily resolved if you use one of the network tools, or through converters. However, they are nowhere near the problems that we face with other binary formats, where we have to stick to a tool and to the platform(s) it is available on.
  • A text file can be programmatically parsed quite easily, without requiring third party libraries. The facilities are usually provided by the operating system API. Check out the power of plain text to see how it can help programs.
  • On many systems, finding in text files too is very easily supported. Utilities like grep and find make it a lot easier to search through a database of text files.

Having said that, the office documents do have their own place. Desktop publishing and collaborating on huge documents can be cumbersome through the text files. We have still seen innovative usage, e.g., the vim help system. It is extremely fast and usable through its marking system. A similar system is also seen in many code editors, which let you navigate using your own marking system.

I have built my own repository of tips and tricks using plain text files. And I cannot imagine using any other format, that will create scores of dependencies and make them it less portable.

Unfortunately we use text files only for log files and READMEs today. I think they are apt for many more cases, where we assume usage of the heavy office documents or some proprietary formats. They not only free us from a lot of dependencies, but they also make efficient use of resources. Let us make use of the text files by default default and use the other formats only when necessary.

Discussion [Participate or Link]

  1. Dave said:

    txt compresses well.

  2. Dorai Thodla said:

    I agree with the benefits of text. It will be nice if we can come up with a format that has a provision for adding some meta data to it (like XML does). Probably JSON is a good candidate.

    On a slightly different topic, I would love to see some standardization on WikiText and wiki markup too.

    When some one asks my resume, I just send them a link to my site/blog. You can so easily create and keep it updated with tools like WordPress, Jottit.

  3. Andy Gayton said:

    Hey Abhijit, man I could not agree more.

    That’s why a couple of colleagues and I built rst2a – a site to help convert reStructured Text into
    different formats. Currently these are HTML and PDF – but we’re hoping to add
    more, particular S5 so you can
    use plain text to replace power point as well.

    We’d love to hear what you think.

  4. Carl said:

    I would love to see some standardization on WikiText and wiki markup too.

    Google for “Markdown”. It’s the wave of the WikiText future.

  5. ryan said:

    I agree completely. I started using .txt files for a lot more lately (the last 6 months) and they work perfect for what I need.

  6. Aperculum said:

    Remember that one of the best publishing formats is based on plain text: LaTeX so I don’t really see where those huge proprietary formats are better.

  7. Spacebat said:

    I use outline-mode in emacs a lot, which breaks down to lines starting with N asterisks to indicate hierarchy and I name those files *.otl

    At my last job there was a shared drive with a directory for each client, containing among other things a file called network.txt. There had been a few pushes to replace this with a database but the engineers stopped this and insisted they needed to be able to edit the info in vi. However the structureless nature of this file was limited the usefulness of the information. I suggested they use YAML, possibly with a wrapper script to make sure the file saved parsed as valid YAML. That way scripts could iterate over the customers and do the right thing, much of the common information would be in predicable places in the file, and engineers could edit it from remote sites in vi. I found a better job before this was implemented though.

  8. http://lukeplant.me.uk/ said:

    Of course, you forgot about encodings. It makes plain text entirely *non* cross platform and non portable. Plain text in emails works fine, as the headers contain encoding information. There is no standard for doing this in plain text, and guessing in unreliable. Attempts at specifying encoding can make things worse e.g. putting unicode BOMs at the beginning of files can cause breakage with some tools. So, when you receive a `plain text’ file, it is basically impossible to parse it at all. http://www.creativepro.com/story/feature/9763.html

    It’s possible that this might become less of a problem if UTF8 becomes the defacto standard, but that isn’t going to happen perfectly anytime soon, and until it does plain text will continue to be painful and limited in usefulness.

  9. Nelz said:

    I was *totally* flustered when I was applying for jobs, and I sent someone a resume in .txt format… They insisted I send them a .doc format one. Why!?!

    It turns out that Microsoft Word actually *can* read and display documents in .txt format.

  10. Ben Hoyt said:

    In my experience companies themselves are fine with a website URL or whatever format for your CV/resume, but recruiting firms require Word docs because they fit in with their big icky HR software. I must say I dislike the red tape and overheads introduced by recruiting firms, and believe companies would do well to skip them and hire direct.

  11. raveman said:

    text files are dead because we live now in Ajax world. Yes, plain HTML is better and faster, but people want it to look cooler. Word is just plain text file on steroids.

    Its funny that someone here said that utf8 should be the standard for text files, but some of you cant agree that .doc is standard for documents.

    The only good argument is that text file is smaller, but we all have now uber-connections, so … size doesnt matter.

  12. Pratik said:

    I am forced to accept ugly .doc files, because I am a college student. Sadly the truth is a lot people don’t care of technical merits of using .txt files, they only use what is default, and we get stuck dealing with their mistakes.

  13. Montoya said:

    I disagree, I would much prefer if there were an XML format and viewer for resumes, and that became the default for sharing them. It would then be *so* easy for software to work with resumes.

  14. Abhijit Nadgouda said:

    Christian, I would agree with you in a fully automated world. XML files are still not readable for a layman. However, I think the next step from a .txt is .xml, especially towards automation.

  15. Brian Rock said:

    Let’s not forget that .doc isn’t the only (or best) office file format out there. Open Office Writer’s .odt format is automatically compressed on saving, and the resulting file is about half of the equivalent plain text.

    That said, I think a move toward XML would be the best direction to go in. It’s easier to work with on the back-end. A lay-person might not be able to read XML source, but that’s where the Word Processor (or browser?) comes in and renders it. As long as it’s a standard data format, people can use whatever app they want to access it.

  16. Document Formats » Bin-Blog said:

    […] The simplest format there is. If there is something I should remember, I just put it in a text file and save it to the desktop. I used to use it a lot earlier – but I don’t use it much nowadays due to searchability issues. There quite a few advantages in using the text format […]

Say your thought!

If you want to use HTML you can use these tags: <a>, <em>, <strong>, <abbr>, <code>, <blockquote>. Closing the tags will be appreciated as this site uses valid XHTML.



Abhijit Nadgouda
iface Consulting
+91 9819820312
My bookmarks


This is the weblog of Abhijit Nadgouda where he writes down his thoughts on software development and related topics. You are invited to subscribe to the feed to stay updated or check out more subscription options. Or you can choose to browse by one of the topics.