Topic: No mods or even third-party mod tools. (Read 934 times)

Hm?

It's moddable. The mod tools are called g++ and {vi,emacs,joe,ed,...}

And there are plenty of 'mods' already existing.

this isn't a game. there's no need to make this moddable. if you want to make some bitcoin service, use the json-api.

Well if you just want to change the images, languages and other resources that'd be pretty easy.

Currently all the resources are compiled into the executable.

However, we could add the bitcoin directory as resource directory, in that case you can put your own icons and images there and they will be used rather than the built-in ones (it may even already be possible with some Qt built-in command line option).
 

That's pretty unreasonably inconvenient, cumbersome, and centralized for common users looking for minor changes, on top of not being geared toward allowing optional targeted improvements only some users may want/need, but to improve the standard code for everyone. - Not that to say it's bad in any way, just not geared toward what I'm thinking of.

Trust is an issue, but this is probably one of the most reputation-emphasizing communities on Earth. I mean - the devs didn't originally say "No, you can't release new Bitcoin clients, because they may contain malicious code" but AFAIK, there isn't a rampant problem of malicious clients out in the wild. People pop up fairly frequently with malicious software to steal Bitcoins, but it isn't the result of Bitcoin mods. It would create new avenues to distribute it, but for scammers, there are already viable ways to log keys from people willing to download and run software (sometimes even without consent to run it!).

It just seems weird to me, that there's this revolutionary community-driven open-source product, and nobody can change it without recompiling it, leaving it to the democracy/dictatorship/whatever of Github. So, we get Client A, Client B, Client C, Client D, and Client E. Take your pick, but what you get is almost certainly all you'll get unless the developers either themselves implement or approve implementation of a change. Plugins are superior for end-users looking for alternatives to every decision made for QT, functionally, because they aren't necessarily broken each time the core software is updated. With code changes, you have to recompile every single time.

This isn't really something I should be throwing at Bitcoin-QT devs, either, because AFAIK, there are no easily-modified Bitcoin clients. Mostly just a curiosity devils'-advocate question I don't expect people to waste their time answering should they think it ridiculously far-fetched -- thanks for the responses.

(I don't know what I'm talking about! [bump])

I don't think mods are a good idea either, but for people wanting them, wouldn't it be possible to allow soft mods that only modify graphics, language and other non-critical data?
I'm just asking, I don't have any knowledge about how mods work

Diapolo and Jonasschnelli have been doing great work lately in making the client more friendly and useful and fun. Thanks!

Code contributions are always welcome.

As for a plugin system, that's not going to happen due to security and trust issues, sorry. Sure, a plugin replacing the splash screen with an animation of a Bitcoin on Bernanke's head or singing llama's would be fun, but not if it steals your private keys behind your back and that's a huge risk (see the proliferation of all kinds of malicious toolbars in IE).

No need for mods, you can just open pull-requests via https://github.com/bitcoin/bitcoin/pulls or create a fork of it and make your very own client .

Dia

You guys ever think about that? I don't think I've ever heard of it being discussed. I can think of some of the obvious and very severe drawbacks, but I'd mostly be talking out of my ass. It'd be nice to hear some intelligent thoughts on it. Was the decision to make modifications to the client fairly cumbersome (requires recompile -- no plugins possible - AFAIK) a conscious decision? Is this something which is slated for the future - either to be implemented or re-examined? Would implementing a more modular, plugin-friendly (don't forget, I don't know what I'm talking about!) Bitcoin require a large number of man-hours?

I can think of a few relatively minor changes which might make a significant improvement in some folks' perception of Bitcoin's functionality/look:
*New animated splash screen of a giant Bitcoin falling on Bernanke's head, knocking it into his body, and replacing the head with the Bitcoin. (this one's very important!)
*Integrate a price-checking mod to quickly grab the Last price off a couple exchanges without some cumbersome independent program.
*Sound effects when we receive coins.
*Additional confirmation messages on certain operations.
*SDice plugin which lets me click a button on my dashboard to gamble.
*Ability to color-code my transaction log, or have additional details in my "Receive Coins" tab.
*Ability to edit Address Book filter and/or color-coding rules.