Future paths for SMF ?

[roxpace]

Does anyone have information about how SMF is supposed to progress in the future ? New planned features, change of parts in the current system ?

Also, are there any bugdatabase ?

All this would be wonderful to have in the progress when we in my company are developing new features and plugins for SMF, which we will of course make public, but it's hard to develop anything without knowing about the future of the product and the current status.

Sincerely
/R

[[Unknown]]

Yes.  Many.  Maybe.

Porting things to the new version shouldn't be that drastic, although there are already a lot of changes.

-[Unknown]

[roxpace]

But why keeping it secret ?  We wanna help you to develop the product with our code and if you like it we of couse lets you use it all, but we can not work professionally if we do not know about anything.

Or atleast a small CVS would be good so we knows what will come soon so we are prepared when the days come for a new release of your wonderful package SMF.

Sincerely
/R

[[Unknown]]

Alright, alright, fine.  For 1.1, at least, we've changed/added/implemented:

  - a lot of cool (but optional for older browsers) xml and javascript stuff.
  - sha1 password hashing instead of md5, with challenge login (javascript needed for challenge.)
  - basic caching framework using memcached/eaccelerator/turck/etc (which can be extended easily.)
  - improvements to the pm system, making it easier to use and sort.
  - improvements in the database structure and index usage, as always.
  - better support for safe mode/phpsuexec.

And a lot of other things.  The changelog.txt for 1.1 is over 60 kilobytes larger than the 1.0.3 one.  Trust me, people will like it - indeed, it's still not beta and we've got a small handful of beta testers using it in production because of the improvements it contains.

So, yes, we've got ideas on how it can and should progress.  We do have a private bug system, which I've considered many times opening up to the public.  The problem is simply that it adds to maintenance, and I don't want that right now.

As for CVS access, we can't give that out (even read) because of the way Charter Memberships work.  If you were to get a Charter Membership, you would be able to see the code a lot sooner... and possibly even acuire CVS read access (although I have people who disagree with me on this in the team, so I'm not promising anything.)

-[Unknown]

[roxpace]

I appreciate a lot that you listed some of the new things, it certainly interests us here a lot and I bet many more also.

If you make that bug system available to the public or atleast a serious range of developers it would be great.

I just say IF we would get CVS access in the future it would be a benefit and very much appreciated.

Thanks a lot so far for all kind of help

[Louis (CSpotkill)]

Edit: I wrote this before you upgraded these forums to 1.1 Beta 1. I don't have time right now to check, but if what I'm asking about below is obvious from the html or js that's live on the site, just say so and I'll take a peek tomorrow.

- a lot of cool (but optional for older browsers) xml and javascript stuff.

Any major changes to the search feature? Like my Friendly Search mod idea? By xml and javascript, do you mean XmlHttpRequest? And how advanced will this be? Basically, what I'm getting at, is - how much of what I posted should I work on before 1.1? Some things, I could probably finish fast enough to release this week, but other ideas might take a lot longer to test.

And if it turns out that parts of what I'm doing has already been done (like transforming search results to XML and presenting them via XmlHttpRequest), then should I buy a Charter Membership so I could possibly view the new XmlHttpRequest structure, or maybe I could build a semi-compatible version on my own, and make minor changes when 1.1 is out to use the stock 1.1 version?

I really feel that forums are locked into this "Category -> Board -> Topic" navigation structure, which like the old folder system (C:\Folder\File) we all know and love, but ... navigation can always be improved, and search can be a great asset to navigation. Suddenly we can group, sort and view posts without worrying what category, board, or even topic they're in.

That's why I feel having a strong search feature will help any forum succeed, by reducing duplicity so people can find what they're looking for, faster, or know when to ask questions. And after being a part of so many forums, I've learned - and I think everyone here will agree - people just don't search enough. So I figure, why not work on making search faster, easier and more intuitive? It's one of the primary ways people navigate to forums, and much of the traffic on forums actually comes from search engines, so people should know how a search box works

[[Unknown]]

Any major changes to the search feature? Like my Friendly Search mod idea?

Find as you type would make mollasses look like a road runner.  A lot of people might like the mod, but it'd never be installed on this forum.

-[Unknown]

[Louis (CSpotkill)]

Well, there's a lot more to the Friendly Search idea besides Find-as-you-type. And I'm hesistant about it myself, because the only time it really helps is if the search interface is slow, and if we use XmlHttpRequest, plus a more dynamic "overlay" interface, and real-time filtering & presentation through Javascript, there's not much point for "Find-as-you-type", I guess.

What I really wanted to know, though, is how much of 1.1 is XmlHttpRequest, if any - and would there be some way I could tie into that without rewriting large parts of the 1.0.x code?

[[Unknown]]

Again, using XMLHttpRequest for every last thing, and making everything work that way would only, in my opinion, complicate matters.  I haven't changed my opinion since I last posted that.  While 1.1 uses a XMLHttpRequest in a quite a few places, it does not use it for everything.

-[Unknown]

[Louis (CSpotkill)]

Okay, thanks for clearing that up. It makes sense and I agree, it would double the maintenance tasks when you need to change something. (Not to mention the extra complication for theme designers and mod writers.) But I'm still going to work out the XmlHttpRequest JS for my mod, because it could vastly improve the speed of the search filtering and presentation, almost eliminating whatever extra load the mod would place on the server, in modern browsers.