dodo

From twext

(Redirected from dodo data)
Jump to: navigation, search

Dodo data stores data in the URL. Why? Simple data in the URL name may be easy to read and share, for both machines and humans.

Contents

[edit] "dodo"?

"dodo" is short for "dot dot". Two dots ".." separate dodo data into parts:

..title..author...txt

[edit] flat file data

Dodo data is stored in .txt flat files. Many filesystems permit long titles (200+ characters), so there's room for data in the filename:

..title..author..time..0.1...txt

[edit] machine read

programs read dodo data from right to left:

.txt
..
..0.1..
..time..
..author..
..title..

text file
dodo data
which version
when saved
who wrote it
what the title is

..bad.is.good..richard gabriel..080124.195559.893486..0.1..dodo.txt

Dodo does allow single dots within fields. For example, version data can be named "0.1". Dodo data does *not* allow two dots within a field. A key to dodo data may be time:

[edit] timestamp

a timestamp may be a useful "single point of truth" to:

  • unite dodo data with a common thread
  • divide or make unique separate files
  • record wiki-like title history

note: hash maybe better to make files unique

[edit] variable program

dodo wants a simple data system, so various softwares in various programming languages can work with the same data. For example, a similar tool written in Arc or Java or PHP may be able to easily work with the same dodo data. DodoXml?

[edit] variable database

a single simple data system may also work with variable DataBase Management Systems. Soon, a simple twexter software may actually test dodo data with MySQL, for example. Maybe later it can work on Rails?

[edit] find *.dodo?

.dodo may be a file extension to identify dodo data. If .dodo files are online, search "*.dodo" to find them in diverse places. Since dodo data is easy to read, multiple databases may be able to gather and share disperse dodo data. So we can more easily find, read, modify and share data.

[edit] human read

In general, dodo data for machines goes to the right. To the left, dodo shows data humans want to read:

..bad is good..richard gabriel..080124.195559.893486..0.1...txt

Simple software might hide .dodo files behind some folder, then show humans just simple links to file contents identified by titles:

Bad Is Good

Or simple software may read dodo data in filename and report:

title: BAD IS GOOD
autho: Richard Gabriel
clock: 24 January, 2008 07:55 PM

Lifting the hood to see the actual data may be as easy as opening a folder and looking at filenames. Dodo data want to be easy to read for machines and humans. It wants to be easy to share.

[edit] number namespace

Number namespace means associate names with numbers. First, define dodo data with numbers. Then associate the numbers with names:

..0.. = ..number..

Variable versions of dodo data can assign variable names to constant numbers. Here's a possible variation:

..3..2..1..0...txt
..what..who..when..how...txt
..title..author..time..version...txt
..bad is good..richard gabriel..080124.195559.893486..0.1...txt

Numbering dodo namespace wants two things:

  1. variable versions of dodo data
  2. internationaling dodo data

[edit] versions

Variable versions of dodo data give us flexibility to play around. For example, you might want to think dodo different:

..3..2..1..0...txt
..solve..problem..when..how...txt
..dodo..easy to read and share data..080125.102001.034091..0.101...txt

Games to play with dodo data may be:

  • tags
  • license
  • talk
  •  ?

Dodo data is free to play with. With enough fooling around, we will eventually get to a more serious dodo data version 1.0.

[edit] international

Numbering namespace also might help us translate dodo data to other (natural) languages:

ENGLISH
..3..2..1..0...txt
..what..who..when..how...txt
ESPAÑOL
..3..2..1..0...txt
..que..quien..cuando..como...txt

[edit] twext

Twext is twin text translation within a single document. An "0.2" version of dodo data might play around with with twext:

..3..2..1..0...txt
..TITLE..LANG2.lang1..when..how...txt
..GOOD IS BAD..ENGLISH.español..080125.110259.999678..0.2...txt
..MALO ES BUENO..ESPAÑOL.english..080125.105631.834831..0.2...txt

To format twext, software needs a simple way to find chunked text and translation in variable languages. Is dodo data worth trying?

[edit] kick me

NOTE: dodo data comes from a programming and DBMS illiterate. Some programmer friends thought this idea may be worth trying. So the idea is now updating here. Please fix errors, ask questions, comment, etc..

Dodo data sees two immediate options:

  • Death with no mercy. Please say why this dodo data is bird brained.
    • list
    • reasons
    • to kill
  • Less sucking. Please share any feedback to make dodo data better.
    • questions
    • comments
    • feedback

talk

We finally have a little support to make twexter software work, so we now need a simple way to organize twext data. Is dodo data worth trying, now, soon or ever? Thanks

rolando says: multilingual filenames problem makes dodo no go until idn (if then)


dodo0 is a simple spec to save and edit data needed to get text twext.. simple enough for any translator or natural language learner to grok..



  • long urls: 414 error happens if over 2048 or 4096 charaters.. how many actual characters to represent a punycode character.. yo no se.. if 5, then we still get a 400 character url... if UTF8 stores ascii simple and most twext is from TEXT in english then much of title will be in english so no problem: the punycode part of url is only the twxt language, maybe translator name etc


Retrieved from "http://twext.com/dodo"
Personal tools