46
Aeva Media / Re: [Announcement] Fork you, SMF!
« on: 12 March 2011 à 1h55 »Quote
They've started work on it.Yes, I can confirm there is a branch in SVN for 2.1. Given where it is, the fact it's not in the trunk but is a branch and so far contains a variety of untested micro-optimisations... (I know they're untested because at least one has broken functionality) it strikes me as not being worth the effort so far. I haven't backported anything from there that I hadn't seen off the tracker (which to me seems fair game if it's in public) but most of it is full on micro optimisation rather than anything actually interesting.
Quote
Now, all we're waiting for is for them to release SMF2 with the BSD license and then we can officially take over. Problem is, they know how much work we've put into our project... The same way they were scared of taking me in, they're probably now scared of allowing our fork to go public. Because it will only show the incredible gap between the SMF team's work and ours.The official line on this is that it's still being tossed back and forth between the lawyers - yes, they had to bring in lawyers to negotiate the contract of sale of assets between the LLC and NPO. I'm not against that, given how much distrust and dissent there was, but it certainly doesn't make it any easier.
One of the comments was that it seemed like every time they got somewhere, it would be the weekend again, which is the real issue at stake here - it's not fear of us, because they'll do what they'll do regardless of us, but I suspect that once 2.0 final happens and we can start being more open, the real fur will fly once it becomes clear what might have been.
They're not worried about us right now, but I suspect it's going to be like Luke and Yoda - "I'm not afraid." / "You will be. You WILL BE."
Quote
But overall, it's still based on the SMF engine.It is for the moment. As time goes on it's morphing further and further away. When I get round to getting something committable for the ACP overhaul, then probably the single biggest element of SMF's codebase will have a fire lit under its ass - since the ACP accounts for a scary percentage of the code, and once it's gone (bearing in mind how much other stuff changes to accommodate it), I see the figures skewing much more towards 75/25 the other way.
Quote
Hooks are not the be all and end all of plugin programming, though. Just look at WordPress... Everytime they release a new version, some older versions of plugins become incompatible.Sure, hooks aren't everything - but it's a massive step up from where we are right now. I've seen a few mods appear using hooks, which are interesting, but the implementation of some of the hooks basically knee-capped anything seriously awesome with them.
Quote
Actually, I'd really like for our fork to completely prevent the ability of modifying core code...I'm not planning to remove that functionality, but I am going to make it harder to use; perhaps a prompt in the ACP if someone wants to do that ("This add-on modifies your forum's core files, are you sure you want to proceed?"), but I'll be looking at submitted mods to see what people are doing that warrants a raw code hack. Some ultimately will need it, it'll be unavoidable for some, but the vast majority case shouldn't.
Quote
(Or even us...)I fully expect to get into writing mods, I'm actually looking forward to the prospect to a point since I'll be able to get mods out there to fulfill the typical functions that don't need hacks (I have been trying to figure out how to engineer points in so that hooks or similar can be used, without having to use the joys of raw edits)
Quote
You can't port anything without the original author's approval, of course.This is a matter of no little debate on sm.org, so here's the full legal position:
* If you are the owner of the mod, go nuts.
* If you have permission of the author to do so, go nuts.
* If the mod is under a licence that expressly permits code remixing and forks (e.g. BSD, CC-BY-SA, LGPL) without any nasty side effects (e.g. GPL), go nuts.
Anything other than a fresh rebuild is off the table, though. It should be noted that I've been thinking about how to make mod data compatible; I've already said I wanted to make getting data from SMF+others to be an import process, rather than a conversion (you go into the ACP, press a button and say "I want data from <here>" and it gets it and pulls it into the forum that's already installed). What I'm aiming for with that is the ability to make it so modular that it can support pulling in boards and posts and stuff, but also that it can be used to inherit from mods - for example, if the user is coming from SMF+AeMe, that it can support importing the SMF core stuff and then AeMe. Or SMF Gallery Lite. Or any relevant content from SimplePortal or TinyPortal. I think you get the idea
In closing, I'll leave you with a few thoughts.
Quote
People with a vision scare them. Why do you think they've been stuck for the last 3 years on betas and RCs?I don't think it's that people with a vision scare them. I think it's people who 1) have a vision that doesn't see exactly the same way as theirs and 2) have the ability to take that vision and make it real. Which certainly includes both of us.
As for SMF's vision and future, two things come to mind. Firstly, who remembers the "I'm a PC" / "I'm a Mac" ad campaigns? SMF is in that, holding a placard, "I'm a forum." Which is great, it knows what it is. But that's all it is. It's a capable, extensible forum. But there's no place in the vision for half of the features we're doing because they're not strictly forum features and so don't fall into their vision of what SMF's future should look like.
Daft as it sounds, I worry about SMF's future, because they don't HAVE a vision. The most solidly agreed points for 2.1 are: IPv6 support, maybe draft posting and decluttering the core to avoid IE6 support. Everything else is maybe and don't know. I don't know about you, but that doesn't exactly bode well for much happening in 2.1 or beyond.