Protocols

[Adron]

To whoever said that bandwidth is a factor: text compresses very well, so use a text protocol and compress it.  Compare the file size of MS Word documents (binary files) to equivilant OpenOffice files (gzipped xml), and you'll see a big difference. 

To whoever said that things like dates/escape sequences need to be standardized: xml.

Haha, and try writing an XML parser... Luckily there are libraries available for a few languages, but XML really isn't as simple a solution as a binary protocol would be.

The reason MS Word documents get so large is that they contain a lot of information apart from the document text itself. And, they do zip pretty well. Generally, binary storage containing the same information as an XML version will always compress better.

[MyndFyre]

To whoever said that bandwidth is a factor: text compresses very well, so use a text protocol and compress it.  Compare the file size of MS Word documents (binary files) to equivilant OpenOffice files (gzipped xml), and you'll see a big difference. 

I'm not entirely sure what is included inside of a Word file, but given that a completely empty Word document is roughly 20kb, I would have to say that there's a problem with that argument somehow.  Plus as Adron pointed out, you can zip a Word document; you're comparing an uncompressed file to a compressed file otherwise.  I believe all the stuff inside a Word document is there to enhance backward-compatibility; note that you can probably open a Word 2003 file in Word 97 with maybe a few bugs.  But if you try to validate an XML file against a strict DOM, a good parser (*ahem*, an accurate parser) will toss you errors.