What should be core?

hcgtv · 2008-07-22 01:54:35

This discussion started over here.

And it continues here, what some feel should be added to core are:

Split and merging capabilities for topics and posts.
A more robust thread tracking system.

Connor · 2008-07-22 01:59:19

Its not a discussion really, neither will be added to the core. Not for 1.3 anyway.

Smartys · 2008-07-22 02:07:51

Or rather, it's not a feature request (at least not one that we're going to implement). What should be core is certainly a worthy topic of discussion.

hcgtv · 2008-07-22 02:24:46

Connor, no way is this thread intended to affect 1.3's feature freeze.

But I feel we could calmly discuss what we feel could one day be core material. I mean it could start out as an extension, at least to work out the how's and why's, then integrate it into the core at a later stage.

Let's take split/merge as an example, if an extension existed, it would be #1 on the hit parade. It's mere existence shoots it up to officially sanctioned, must be ready to coincide with the next release kind of thing. So if the core devs are having to deal with this extension's code as if it's almost umbilically tied to the core, then should it be core?

Smartys · 2008-07-22 02:27:41

Now then, to address something brought up before the split:

hcgtv wrote:

Once people start adding extensions left and right, lean and mean will be used less often in these here circles. I'm not implying that things will slow down on your forum as you add widgets, but a different mentality will take hold, one desiring more.

It's a very, very good point.

1.2 was "lean and mean" by virtue of the way it was coded (essentially, by its lack of extensibility). If you wanted to add something, you had to code it yourself (or find someone who already had and who had released the code), modify the files, etc. It's a lot of work to heavily modify code that way. Therefore, most 1.2 installs are/were fairly close to what the core is, which is lean and mean.

In 1.3, we have this wonderful new extension system. It makes it much easier to add small features and tweaks to your forum. However, this system also allows people to (easily) turn their forums into a fully-featured, everything-including-the-kitchen-sink website. Therefore, we're likely to see at least some forums moving away from the "lean and mean" philosophy and toward a more feature driven avenue. FluxBB's core can be as trim as it wants to be, but its defining feature has become extensibility, not a trim core.

Smartys · 2008-07-22 02:31:01

hcgtv wrote:

But I feel we could calmly discuss what we feel could one day be core material. I mean it could start out as an extension, at least to work out the how's and why's, then integrate it into the core at a later stage.

Agreed. There's no reason we can't merge (hah, pun) it (I'm talking specifically about splitting/merging) into the core once it has been fully developed and tested as an extension.

hcgtv wrote:

Let's take split/merge as an example, if an extension existed, it would be #1 on the hit parade. It's mere existence shoots it up to officially sanctioned, must be ready to coincide with the next release kind of thing. So if the core devs are having to deal with this extension's code as if it's almost umbilically tied to the core, then should it be core?

Except such an extension is not likely to require much maintenance at all once it's working in the first place. And if you're saying extension popularity should translate to what we're putting in the core, then I'll have to disagree on the basis that it runs counter to everything we've been developing (other than the extension system, I suppose ).

MattF · 2008-07-22 02:41:40

What's the point of core? That feature may be number one on your 'Hit Parade' list Bert, but it sure as hell ain't on mine. If two people disagree on whether it's a must have feature or not, it ain't no must have feature. Extension material. Pure and simple. That's what the extension system is there for.

Last edited by MattF (2008-07-22 02:43:13)

hcgtv · 2008-07-22 02:56:03

Smartys wrote:

FluxBB's core can be as trim as it wants to be, but its defining feature has become extensibility, not a trim core.

Yes, the 'x' in Flux overtakes the other letters

Smartys wrote:

And if you're saying extension popularity should translate to what we're putting in the core..

Not so much popularity, as much as it becomes another thing the core devs have to deal with on a release.

I look at extensions as being of two distinct classes:

They extend the system.
They compete in the market to be the best.

Split/Merge would not be something that just anybody could write, you'd have to have a really good understanding of the core, hence no competition. So the flux_sm extension, or affectionaly known as Flux S&M (because only a masochist would take it on), would be core.

MattF · 2008-07-22 03:00:16

hcgtv wrote:

So the flux_sm extension, or affectionaly known as Flux S&M (because only a masochist would take it on), would be core.

Give just one valid reason for it being core.

Edit: Just on a slight sidenote, what is this bloody obsession that people have with their desired features being in the core rather than just installing an extension?

Last edited by MattF (2008-07-22 03:02:52)

Smartys · 2008-07-22 03:04:53

hcgtv wrote:

Split/Merge would not be something that just anybody could write, you'd have to have a really good understanding of the core, hence no competition.

Don't be too sure about that. I'm evaluating the difficulty of adding the feature into the core (don't get excited, I may decide it's too much trouble for now) and it doesn't seem like it takes much more than a basic knowledge of the code and some copy/pasting of existing stuff.

And just because a theoretical extension is tied tightly to something that we would somehow manage to break every release (I'm not quite sure what it would be ) doesn't mean it should be in the core.

Matt: Split/merge is one of the most useful moderation features I've ever run across. This is different from, say, adding an avatar gallery to the core. It adds functionality that many, many forums could benefit from.

hcgtv · 2008-07-22 03:08:54

MattF wrote:

hcgtv wrote:
So the flux_sm extension, or affectionaly known as Flux S&M (because only a masochist would take it on), would be core.
Give just one valid reason for it being core.

Because it doesn't matter if it's an extension, by it's inherent popularity you dare not ship a release until flux_sm is oiled up and ready to go.

MattF wrote:

Edit: Just on a slight sidenote, what is this bloody obsession that people have with their desired features being in the core rather than just installing an extension?

I don't need this feature, I'm really good with phpMyAdmin

MattF · 2008-07-22 03:16:44

hcgtv wrote:

MattF wrote:
Give just one valid reason for it being core.
Because it doesn't matter if it's an extension, by it's inherent popularity you dare not ship a release until flux_sm is oiled up and ready to go.

So, in other words, writing an extension in time for the Flux release would be fine? Then why ask for it to be part of the core?

Reines · 2008-07-22 03:17:38

Merge/Split is one of the few missing features that I think would fit into the core without being out of place. Saying that I would think a better approach is to release an extension for it for 1.3 then judge if it is even worth discussing adding to the core later on, based on how many people actually install the extension.

Last edited by Reines (2008-07-22 03:18:35)

MattF · 2008-07-22 03:22:22

Smartys wrote:

Matt: Split/merge is one of the most useful moderation features I've ever run across. This is different from, say, adding an avatar gallery to the core. It adds functionality that many, many forums could benefit from.

I'm not saying it's not useful, but having it should be a choice, not enforced. There are quite a few addons that I personally class as necessary, but I would always want those to be extras and not inbuilt. Lack of features can be a pain if something you need or want doesn't exist, granted, but having to strip unnecessary crap out that you don't want or need is far worse. I thought one of the main features of the 1.3* branch was not having to edit the core anymore? That should apply for removing as well as adding features.

It should be a choice. Pure and simple.

Last edited by MattF (2008-07-22 03:24:34)

hcgtv · 2008-07-22 03:41:50

Smartys wrote:

And just because a theoretical extension is tied tightly to something that we would somehow manage to break every release (I'm not quite sure what it would be ) doesn't mean it should be in the core.

You're looking at it from the inside, I'm standing on the outside and wiping off the morning dew from the glass.

My view of extensions is based on using other systems that are themselves pluggable. Let's take Textpattern for instance, the current stable code base is full of deprecated code that only exists so as not to break a plugin. The TxP plugin system filled many a core void, now the devs have no choice but to haul it all around.

So the way I see it, I would make core, what would allow the code base to evolve more easily.

fmimoso · 2008-07-22 07:08:00

Speed and perfomance.

Last edited by fmimoso (2008-07-22 07:08:31)

SuperMAG · 2008-07-22 08:13:36

hcgtv wrote:

This discussion started over here.
And it continues here, what some feel should be added to core are:
Split and merging capabilities for topics and posts.
A more robust thread tracking system.

neither of those should be in core now and ever, if you want to maintain fluxbb.

Strofanto · 2008-07-22 08:42:44

A Captcha or any antibot system should be core IMHO.

pjalar · 2008-07-22 09:52:53

Strofanto wrote:

A Captcha or any antibot system should be core IMHO.

I don´t think so. Have no use for neither.

Last edited by pjalar (2008-07-22 09:53:28)

yemgi · 2008-07-22 10:13:23

Neither do I, captcha and antibot are not features of a forum software but protection mechanism you can add if needs be IMHO.

The split/merge feature is a very useful moderation tool that was solved in 1.2 by fpouget's move post mod. I really think this should be part of the default moderation option of a forum software. I agree that some people don't see a need for it but on large or popular boards it is necessary if you want to keep threads organized and prevent off-topic discussion

MattF · 2008-07-22 11:08:45

Is a sitemap generator in core, or going to be if not?

Jérémie · 2008-07-22 12:24:46

MattF wrote:

Edit: Just on a slight sidenote, what is this bloody obsession that people have with their desired features being in the core rather than just installing an extension?

On top of what Bert said (which is very true): maintainability.

If a core forum feature is handled by an extension, each time you install and each time you update boards (and one can have many, many, boards, unless the dev wants the web professionals outta here) you would have to:

Search with (to this day) unknown tools the repository and the forum for extensions that would fulfill your need, read them, read the user comments about them, check the author's rep, and select one
Check compatibility with your Flux version, and every extension you already have installed, by hand
Install it
Translate it, and hope the future updates of that extension doesn't erase or break your translation
Test it, to see if anything breaks, if the extension doesn't produce bad HTML, doesn't rely on heavy, slow, unnecessary javascript, and so on
Keep an eye on it for future updates

For each basic, small, feature you need.

It's not a personal view, it's what happened to Textpattern to take one example of a software where the devs answered every feature request by “wrote it yourself as a plugin”.

If a Flux's install, at some point, has a more time consuming maintenance than vBulletin, something has gone very very wrong somewhere.

I'm not pushing for everything in core... PM is a good example of something many people want, and that needs to be an extenstion: it's heavy, it duplicate the pre-existing email protocol and softwares, and so on.

However, split/merge, multi group, xml sitemaps, and probably some others I don't remember right now, are basic feature of a web forum. As in, the Avatars feature is something more advanced and less used than these.

MattF · 2008-07-22 12:36:25

Jérémie wrote:

However, split/merge, multi group, xml sitemaps, and probably some others I don't remember right now, are basic feature of a web forum.

No, they are not. They are all features which someone thinks are necessary. They are not features which everyone thinks are necessary. The core is perfectly fine as it is. There is absolutely no need for those to be in core.

Simple point, in my personal opinion. I went with Pun originally for the simple fact that I *didn't* have to strip as much out as with other wares to get it small and, in my personal opinion, just right for enhancing to the level I wanted it. Pee on that minimal concept, and I know I for one would not exactly be impressed.

All of those mods above are easily official extensions if demand is deemed sufficient. There is absolutely no need to shoehorn them into the core.

Connor · 2008-07-22 13:09:19

Jérémie wrote:

Search with (to this day) unknown tools the repository and the forum for extensions that would fulfill your need, read them, read the user comments about them, check the author's rep, and select one
Check compatibility with your Flux version, and every extension you already have installed, by hand
Install it
Translate it, and hope the future updates of that extension doesn't erase or break your translation
Test it, to see if anything breaks, if the extension doesn't produce bad HTML, doesn't rely on heavy, slow, unnecessary javascript, and so on
Keep an eye on it for future updates

We have said lots of times, there will be an extension repository managed by us.

Jérémie · 2008-07-22 13:35:12

Connor wrote:

We have said lots of times, there will be an extension repository managed by us.

And translation, and dependencies, and compatibility (upward and downward), and keeping a board up to date away from your home (and your sftp client, passwords list, personal certificates, and every tools that might be needed), etc.? Nothing on that.

I'm not saying FluxBB will become the maintenance and update nightmare Textpattern or Sourcemod (to take another example) are, I'm saying that right now I don't know (and it's my job to know those kinds of things).

And until we test it in real life scenarios, I don't think anyone know. And that worries me, kinda.

Forums

#1 2008-07-22 01:54:35

What should be core?

#2 2008-07-22 01:59:19

Re: What should be core?

#3 2008-07-22 02:07:51

Re: What should be core?

#4 2008-07-22 02:24:46

Re: What should be core?

#5 2008-07-22 02:27:41

Re: What should be core?

#6 2008-07-22 02:31:01

Re: What should be core?

#7 2008-07-22 02:41:40

Re: What should be core?

#8 2008-07-22 02:56:03

Re: What should be core?

#9 2008-07-22 03:00:16

Re: What should be core?

#10 2008-07-22 03:04:53

Re: What should be core?

#11 2008-07-22 03:08:54

Re: What should be core?

#12 2008-07-22 03:16:44

Re: What should be core?

#13 2008-07-22 03:17:38

Re: What should be core?

#14 2008-07-22 03:22:22

Re: What should be core?

#15 2008-07-22 03:41:50

Re: What should be core?

#16 2008-07-22 07:08:00

Re: What should be core?

#17 2008-07-22 08:13:36

Re: What should be core?

#18 2008-07-22 08:42:44

Re: What should be core?

#19 2008-07-22 09:52:53

Re: What should be core?

#20 2008-07-22 10:13:23

Re: What should be core?

#21 2008-07-22 11:08:45

Re: What should be core?

#22 2008-07-22 12:24:46

Re: What should be core?

#23 2008-07-22 12:36:25

Re: What should be core?

#24 2008-07-22 13:09:19

Re: What should be core?

#25 2008-07-22 13:35:12

Re: What should be core?

Board footer