Projects and Whatnot

Project Discussion

Both data file metaformats (DFM) and language specific datatypes (LSD) deal with how data is stored
- DFMs define data using ASCII or UTF-8 characters making them suitable for storing as files on disk. They are portable and language independent and human readable. To use the data stored in them in a program some kind of parser is required.
- LSDs define data in some binary form and are language and architecture dependent. For instance in C the number of bits in an int and how they are ordered in memory is architecture dependent. And int in Python may have a different binary representation than an int in C. The data stored as LSDs are not portable, it usually doesn’t make sense to use them outside of the specific program that define them. Because of this datatype values are usually only stored in memory while the program is running (depending on the language and compiler, certain LSDs may be stored as part of the program data on disk).

Different metaformats may parsed faster than others. How much concern should we give to the parsing speed of a particular format when choosing one to use for a text-based adventure game?
XML would also work well for representing the data in a text-based adventure game. It also provides another layer of meaning by allowing us to define a scheme: a set of rules that dictate how data can be organized, e.g. we could specify that a room could contain one or more items, but an item can’t contain one or more rooms. YAML or JSON doesn’t provide this level of detail:
```
SURGE 104C: 
    desc: a generic classroom
    items: [ desk, chalk, a mysterious note ] 
    north: SURGE Study Area
```
and
```
SURGE 104C:
  desc: a generic classroom
  items:
  - desk
  - chalk
  - a mysterious note
  - SURGE Study Area: {desc: an open space, items: vending machine, south: SURGE 104C}
  north: SURGE Study Area
```
are both valid YAML but the latter example is not a well-formed representation of a multi-room game environment. Is this motivation for XML?