[tcaudilllg]

[LeoDraco]

[tcaudilllg]

I've been considering using entities. But what if the icon text itself is the entity?

[LeoDraco]

I'm not sure what you are getting at. You mean you want some sort of meta-entity? While that should not be needed, you could always do:

[RuneLancer]

[tcaudilllg]

But how do I code that? It would be possible to have meta meta meta meta meta meta meta entities!

[LeoDraco]

Given your proclivity towards JavaScript, I am going to assume that you are implementing this fascinating project with it. Taking that as given, you could probably just parse the XML tree with the DOM functions.

Now, about the character entity things: again, for what use do you forsee even needing the whole meta-entity thing? The only current reason why one would need that is for situations where entities show up that you do not want processed, and, even then, it usually only has a single level of escaping. (E.g. escaping ampersands in url's sent to a server, or escaping entities that show up in a form.) One solution would be to write some sort of parser to handle them. The better solution is to make use of JavaScript's regular expression engine to parse the entities. I.e.:

Escaping:

[Nephilim]

[tcaudilllg]

The only solution is to code the length of the string before the start of the string.

I want icons to be capable of binary text too, that is, text that is essentially random, binary code. Like image data, or compressed data. So it is impossible to differentiate between escape characters and valid data, because the escape characters can emerge.

Basically, this is a case of Godel's incompleteness theorem: if the one system's data system is another system's energy system, then it is impossible for the two systems to communicate, because the confusion is endless.

Unless... an auxiliary system converted energy to data and data to energy between the two. But I can't do that now.

[RuneLancer]

1- There are many ways to encode a variable-length string without resorting to storing the length. While not a bad solution, it means you have to constantly keep it updated. Null-terminated strings don't require such maintenance. On the flipside, you can't know the exact length until you've parsed the string. Keeping the length is definately NOT the only solution, but wether it's the best or not depends heavily on the situation.

2- There's absolutely no situation where escaping characters will cause any problems. The only reason why such a system would fail is because of a horrible parser that ignores escaped characters at random! If, for instance, a backslash escapes a character (including a backslash itself, so two backslashes would translate to a single, normal backslash), short of using a different parser that treats backslashes differently, you will never find yourself in a situation where there'd be confusion between escaped and unescaped characters. Try it.

3- There is no number 3. If there were, however, it would probably involve pie and cookies. Instead, though, I think I'd rather make it on how computer science concepts and physics concepts are not the same and don't carry over to one another.
_________________
Endless Saga
An OpenGL RPG in the making. Now with new hosting!

[tcaudilllg]

LeoDraco is "a" number 3, ...

Um, hmm... I never thought about the escape thing being a problem before....

[LeoDraco]

[tcaudilllg]

I think I'd rather go with the length tracker, given the circumstances.

Base64... c^2, where 8 equals C?

Question: would c^3 be enough to create a generally relativistic environment whose maximum translation rate was c?