These are not problems with the Gemini protocol itself.
The Gemini clients that I have seen do usually allow the user to change the colours, as far as I can tell. (This is usually better than the document specifying the colours, in my opinion.)
Gemini does have a write protocol (I think it is to change "gemini:" at the beginning of the request to "titan:", add some URI parameters for the size and MIME type of the data (I think it might have been better to put those things on the next line (I am not really sure), but well, now it is what it is), and then the data to be written starting on the next line). However, for some kind of things, other protocols would be better.
For interactive fiction/text adventures (or other interactive applications), Telnet/SSH will be better than Gemini, I think. For message forums, NNTP will be better.
Gemini does have the advantage that the file format is easily readable/writable even if only treated as plain text, you do not need a software to interpret it, but that you can also use a program to interpret it too if wanted and if you do then it is still simple.
I would also propose a unencrypted variant. The differences are: The URI scheme is "insecure-gemini" for the unencrypted variant, and 6x responses are not allowed (if a client certificate is needed, it should issue a 3x to redirect to the encrypted version and then the encrypted response will be 6x).
I have no issues with the protocol - I really like it. I really want this to succeed. I'm talking about the way the protocol is presented to the world, the project as a whole. You need to show people it's cool, it's flexible, it can do useful things while also embodying privacy and low cruft.
Otherwise it ends up like Brainfuck - impressively clever, but you wouldn't want to use it for anything. It could be great for education or in many other contexts, but it's important to show something.
The Gemini clients that I have seen do usually allow the user to change the colours, as far as I can tell. (This is usually better than the document specifying the colours, in my opinion.)
Gemini does have a write protocol (I think it is to change "gemini:" at the beginning of the request to "titan:", add some URI parameters for the size and MIME type of the data (I think it might have been better to put those things on the next line (I am not really sure), but well, now it is what it is), and then the data to be written starting on the next line). However, for some kind of things, other protocols would be better.
For interactive fiction/text adventures (or other interactive applications), Telnet/SSH will be better than Gemini, I think. For message forums, NNTP will be better.
Gemini does have the advantage that the file format is easily readable/writable even if only treated as plain text, you do not need a software to interpret it, but that you can also use a program to interpret it too if wanted and if you do then it is still simple.
I would also propose a unencrypted variant. The differences are: The URI scheme is "insecure-gemini" for the unencrypted variant, and 6x responses are not allowed (if a client certificate is needed, it should issue a 3x to redirect to the encrypted version and then the encrypted response will be 6x).