• Welcome to Simple Machines Community Forum. Please login or sign up.
September 21, 2021, 02:07:11 PM

News:

Want to get involved in developing SMF, then why not lend a hand on our github!


Were I given the keys to the kingdom...

Started by Arantor, September 22, 2013, 02:03:37 PM

Previous topic - Next topic

Arantor

September 22, 2013, 02:03:37 PM Last Edit: September 22, 2013, 02:39:10 PM by Arantor
Firstly, I'm well aware this should have been in 'Next SMF Discussion' but the board is set up so that I can't start a new topic there :/

Some people, it seems, are curious to know what I would make of SMF were I given the keys to the kingdom, especially as I have been critical of 2.1's developmental progress. (Please note that my issue with 2.1 is not the people doing it, because they're all good people. It's in the way the process has been handled)

And so, I found myself asking, what would I do? What would I like to see in SMF's future? We all know smCore is dead, and so we will ignore that, and instead focus on what SMF 3.0 - a pure forum software - should include, and the road to getting there.

Note that I am referring to building on top of current 2.1, and am only too mindful of what features were added in 2.1; things like auto-saved post drafts are already built in.

-- 2.1 --

I would declare 2.1 as it stands as feature frozen. Since no-one knows what should be in it or what it should be capable of, I'd just declare it feature frozen and concentrate on fixing up bugs. There are a number of bugs in it, as well as various theme issues especially related to some admin pages, where the theme just looks broken.

I would then concentrate 2.1 development on fixing the known issues and getting as much as possible just cleaned up with a view to a public beta by the end of the year. I do not see this as unrealistic given the manpower available to the SMF team should they actually get involved in trying to beta test things. As it stands, though, that kind of motivation is sadly lacking.

-- 3.0 --

I'd start 3.0 development on top of 2.1 while 2.1 is going through late beta/RC stages and use 2.1's build as basis.

There are a number of things I'd start changing. 2.1 has to retain a certain amount of backwards compatibility, 3.0 by definition does not (and smCore certainly wasn't going to in the first place)

1. I'd ditch the current converters. I'd build them into the admin panel as importers. This would encourage them being kept up to date more, and would actually be easier to use for users than the current organisation.

2. Facebook/XenForo like notifications system. It doesn't need to be as live as Facebook (and can't be done on shared hosting for the sheer server load issue) however the approach taken by XenForo and other systems would work well for this.

3. On the back of that, I'd move the Profile and Messages items out of the menu and place them up in the header area (much like other systems), and tie in the notifications to that.

4. I'd go UTF-8 only. Not needing multiple sets of language files, just one set of working language files. Everything gets a lot simpler under the hood, too.

5. I'd reorganise most of the current code into classes. I realise that going OOP is a big hurdle for modifications but for the core code it needn't be the headache that it sounds like it should be. Having the code organised by function, and more importantly organised in such a way that an autoloader will help you out is a big thing. Instead of loading a ton of code every page in case you might need it, load only what you need. You could save thousands of lines of code per page without any trouble doing that.

6. I'd overhaul the way news is currently handled. I'd move it to its own table, adding in visibility controls to it, making it clearer that bbc is allowed, and then I'd grab the entire thing, parse the bbc, then cache it. A surprising performance saving on every single page if news is in use. (Mostly like what I did for Wedge)

7. I'd rewrite the board index. All the board listing code relies on a quirk of MySQL that is no longer reliable. That needs rewriting to change the way it works.

8. I'd reimplement the ban system to be not based on triggers or groups, but simple linear bans. I would ditch the logging of bans and just flatten it out. (Like I did for Wedge)

9. I'd seriously give the warning system an overhaul (like I did for Wedge)

10. I'd phase out the Core Features page entirely. Half the items don't get any performance benefits from the actual page itself, so there's no real need for it.

11. I'd rearrange and streamline the admin panel.

There's plenty more but unfortunately I don't have time right now to deal with it.

EDIT TO ADD:

OK, let me go into more detail.

10. Core Features doesn't generate anything worthwhile, there is no performance saving in most cases because the individual 'core features' internally do other management themselves which the CF page just replicates.

11. There are a great many things I'd change in the admin panel. Lots of things in weird places; mostly historically 'where they would fit'. I already did a ton of rearrangement in Wedge for this, I'd replicate most of that.

In particular, I'd totally gut how Themes > Member Options works. Almost none of those actually needs to be a per-theme preference, it can just be a single preference with reset option, which would be a major improvement.

I'd also rejig how permissions work, I would be inclined to make them role-based rather than group-based (another reason for having the importer: by having even SMF 1.x and 2.x as imported rather than just straight upgraded)

12. I'd also seriously overhaul how plugins/addons work, I'd promote hooks to be even more significant, perhaps the only method by which add-ons can be used, getting rid of file edits. I did it for Wedge and I believe it's the way to do it - yes, it reduces flexibility. But on the other hand, you get a much more rugged and stable approach to add-ons, you get things like easy/safe removal of plugins by renaming a folder to disable them.

Some of the more complex arts can be mitigated by switching templates out of the current PHP approach and switching to something like Twig; plugins would be able to make changes to templates safely in that situation.

13. I'd also replace the jump-to system with one more flexible, like a popup of sorts. It would look nicer, and would work better with all the portals out there that currently have to fight with it.

14. I'd drop the facility whereby each post has its own subject, I'd move that all to the topic itself. Saves a ton of space, makes a surprising number of things faster.

15. Overhaul search to be able to handle other kinds of content as well as supporting Sphinx properly with real-time indexes, and also provide support for ElasticSearch. Ideally ship both of these with the core.

16. Paid Subs supporting other things (not just PayPal) and working with fewer issues in general.


I'm sure there's even more, but that's the immediate and more important stuff I'd do. I'd like to hear everyone else's thoughts.
No good deed goes unpunished
All helpful urges should be circumvented

emanuele

Quote from: Arantor on September 22, 2013, 02:03:37 PM
Firstly, I'm well aware this should have been in 'Next SMF Discussion' but the board is set up so that I can't start a new topic there :/
Yep, it is/was supposed to be a board where the devs (or others maybe) move things that are "accepted" to be in the "next SMF".

Quote from: Arantor on September 22, 2013, 02:03:37 PM
Some people, it seems, are curious to know what I would make of SMF were I given the keys to the kingdom,
You would run as fast as you can. :P ...unless you are crazy (and probably you are, like most of us... lol)

Quote from: Arantor on September 22, 2013, 02:03:37 PM
I would declare 2.1 as it stands as feature frozen. Since no-one knows what should be in it or what it should be capable of, I'd just declare it feature frozen and concentrate on fixing up bugs.
Actually it was declared feature frozen more than a year ago, except that add things it's easier than fix bugs.
And there were (yes, I know hidden board to the most) a very clear set of goals for 2.1.

...oh heck, here it is the post:
Quote from: xxxxx on February 22, 2012, 11:38:32 AM
One important note is that we are trying to not needlessly break the ecosystem that surrounds 2.0 (mods / themes / etc).  By this we are avoiding making changes in some areas unless there is a security or true performance improvement to be had. 

Basically this implies not doing pedantic changes or any large scale refactoring.  2.1 is really 2.0 "classic" and not a first step towards 3.0, since as we found that was really not going to be a step but a monumental change. 

All that said, yup as with any release we will run an excellent chance to break mods, thats the nature of mods that make source edits, its inherently fragile.  Simply adding a hook to a high traffic area will "break" things, but we should all know the benefits of that break.

You can go to the bug tracker and use a tag filter, 2.1 as the tag, across all projects to see what's being reviewed at this time.  There are bugs/features across 2.0, 2.1, 3.0 in the tracker, bit of a mess really, so again being this is a part time job :) I used 2.1 tags to track for now.

So with that brief introduction, below is a simplified list of things being worked on.  I felt this was an easier way to view things than providing any long list of bugs, who / what / where.  Also note a majority of the items on this list was derived from previous 2.1 wish list threads.



Core Improvements:
•   PHP 5 - This version will require a minimum version of php5.0  PHP 5 was first release in 2004, almost 8 years ago, its high time we moved to update our minimum support requirements.  (really I'd like this to be 5.1 which was late 2005)
•   IPv6 Support - Specifically in posts, ip tracking and ban management.
•   Smartphone theme - Smart phones have come a long way since SMF came out, and its time to provide a better basic interface than WAP for those users.  2.1 will have a new basic theme for viewing on smaller devices.
•   Jquery library support - Many mods require the use of Jquery so 2.1 will include it as part of the base.  You will have the option to load Jquery from the Google CDN or from your local install.  Where and how much we make use of Jquery ourselves is not clear ATM (menus, modal windows, autoselect are some thoughts)
•   Mod/admin sessions. - Today you can enable the need to login to gain access to the admin functions, new will be the setting to also require this for moderators as well.  Both will also come with the ability to end the admin or moderation "session".  This improves security to these more sensitive areas of the site.
•   Form Tokens - The results of adding tokens are improved security, such as basically no CSRF (Cross-site request forgery) can occur, even if an exploit successfully gained the admins session var and value they would still not gain access.
•   Code Cleaning - In some places the code has been cleaned up, mostly to remove older php 4 support and take advantage of new functions that were first made available in php5.  Additionally some basic functions have been refined to improve their performance.
•   HTTP only cookies - With this optional setting cookies won't be accessible by scripting languages, such as Javascript. This setting can help to reduce identity theft through XSS attacks

Modification support improvements:
•   Hooks Proliferation - There now almost 140 new hooks in 2.1, almost 4X the number in 2.0.  Hooks greatly reduce the need for mods to make direct code edits and allow for much easier mod installation and upgrades.
•   PHP Doc blocks - All of the code in 2.1 will have improved documentation, or at least easier to find documentation.  Instead of the files having all of the documentation at the top, things have been moved to each specific function.
•   Emulation Version dropdown - Allows for easy selection of other SMF versions when you need to install in emulation mode.

Theme support improvements:
•   PNG icons.  The old gif icons are nice, but PNG icons are better.
•   Cleanup of the markup/css.  There are many areas that are inconsistent from page to page, 2.1 will clean up a lot of these areas to provide a more consistent look and feel across the various pages.
•   Add ability to add javascipt/css via hooks
•   Move much of the inline javascript functions to script.js

Bug fixes:
•   We wrote them, you found them and reported them, now we are trying to fix them!  2.1 will look to squash as many bugs as possible, but we will not get to them all.  Currently 2.1 has addressed 80 bugs and there are many more being worked on.

Login/register improvements:
•   Allow the forum to be set up for the login username to only be the email address.

Package manager improvements:
•   <requires> tag, for mods that need to have another mod installed to operate, that mods package id can be specified to let the package manager check for compliance
•   <license> tag,  Will result in a notification under the readme, displaying the content of the license tag.
•   <hook> tag, allow the adding and removing of hooks from the package.info file.
•   <credits> tag, allow the adding the mods details under the credits area of the forum.
•   Improved zip file handling to reduce the number of invalid zip files or improperly expanded files.

Attachment improvements:
•   Auto manage attachments, will create new directory's as needed based on number of files set in the admin panel
•   Attachment Error Enhancements - Instead of the generic "you have a problem" message, you will now be notified which attachment has cause the problem and still be able to post with those attachments that did not fail.
•   Improve/fix image attachments checks. (i.e. reduce the number false infection positives on valid files)

Search improvements:
•   Search drop down, this topic / board / site
•   Improvements (bug fixes) to the custom index, text highlighting, searching on special characters like & > <, etc

Other improvements:
•   phpinfo() is now available form the Admin area
•   Recount users posts maintenance task will re-count and update the post counts for your users.

----------------------- CUT LINE / Scope creep ------------------------
These items should be in the core, but have not been claimed (just yet) ;)
"Smart emulate"  Let the system figure out the best version based on whats in the package info file and use of compareVersions functions
"message preview on hover"
"moved topic enhancements"
"Go Advanced" Option should be in Users Profile > Look and Layout not Admin panel
"signature preview" ... As a start, Perhaps anything that can be previewed, should have a preview
"categories in add/edit membergroups pages"
"Hooks admin area"
"Make the custom avatar directory functionality a default SMF feature"
"Option to "deny access" to a board"


Take a peek at what I'm doing! ;D




Hai bisogno di supporto in Italiano?

Aiutateci ad aiutarvi: spiegate bene il vostro problema: no, "non funziona" non è una spiegazione!!
1) Cosa fai,
2) cosa ti aspetti,
3) cosa ottieni.

Arantor

Interesting, then, that several people I've spoken to lately - including BOD members, didn't even know there WAS a roadmap of any kind. (Which is why they were talking to me in the hopes of trying to get me on board as a dev...)

And yes, in the 18 months since that post, pretty much everything has been implemented. In which case, perhaps it's time to call for a push on bugs by releasing a public beta...?
No good deed goes unpunished
All helpful urges should be circumvented

margarett

Good reading. Of course those are your personal perspectives but for most they make sense.

I would add (and I think you forgot) the need to have a responsive mobile theme.
Se forem conduzir, não bebam. Se forem beber... CHAMEM-ME!!!! :D

QuoteOver 90% of all computer problems can be traced back to the interface between the keyboard and the chair

Arantor

Yup, I forgot that, bearing in mind I was largely writing in a hurry :P
No good deed goes unpunished
All helpful urges should be circumvented

kat

Having decided against my usual Tl;dr with long posts, I took the plunge. :)

Obviously, with my severely limited talents, quite a bit of what you typed didn't make total sense. (My lack of knowledge, not your lack of explanation).

What I read, made me think "hmmm. what I would make of SMF were I given the keys to the kingdom... I like that analogy". :)

Then, I got down to your numbered points.

Pardon my use of the vernacular, here. But, I thought "****** my old boots!". They describe, almost word-for-word, what I happen to feel are major failings in the software, as it is. As most know, my forum's still on SMF v1, for most of the reasons you describe. The admin panel, in v1, is a bit illogical, in places. But, with v2, it got even more loopy, for reasons that I doubt I'll ever understand.

I don't know if the themes thing could be changed, on top of your other proposals, coz, again, the themes for v1 seem, to me, to be much more varied and more... well... well, some still have that "WOW!" factor, which not a single v2 theme ever has.

I guess that what I'm saying is...

emanuele

Quote from: Arantor on September 22, 2013, 02:42:58 PM
Interesting, then, that several people I've spoken to lately - including BOD members, didn't even know there WAS a roadmap of any kind. (Which is why they were talking to me in the hopes of trying to get me on board as a dev...)
It's nothing new that people don't read many boards here around and then complain about something that happened "somewhere else". :P
This post was firstly conceived in a dev-only board (about January), then posted in a team-only board (the above date), and finally posted in July last year (2012) in the dev board open to beta testers and friends.
ETA: and explained briefly in a blog post.

Quote from: Arantor on September 22, 2013, 02:42:58 PM
And yes, in the 18 months since that post, pretty much everything has been implemented. In which case, perhaps it's time to call for a push on bugs by releasing a public beta...?
Actually 6 months after that post everything was already ready (well, a new theme WIP came in because "SMF cannot be released without a new theme").
I tried to push a release twice, but my conditions (like have it live here) for it were never met, so now it's up to Oldiesmann. He mentioned it twice recently, so I suppose he wants to have a release.


Take a peek at what I'm doing! ;D




Hai bisogno di supporto in Italiano?

Aiutateci ad aiutarvi: spiegate bene il vostro problema: no, "non funziona" non è una spiegazione!!
1) Cosa fai,
2) cosa ti aspetti,
3) cosa ottieni.

Arantor

See, I can complain about things I don't know - the roadmap isn't public and thus as far as I knew...

Wait, so a full year ago it was basically ready to go?
No good deed goes unpunished
All helpful urges should be circumvented

emanuele

Except for what it was added outside the scope and at the time it was not ready (and as you pointed out is not yet still fully ready), yes, everything else was done and ready for testing.


Take a peek at what I'm doing! ;D




Hai bisogno di supporto in Italiano?

Aiutateci ad aiutarvi: spiegate bene il vostro problema: no, "non funziona" non è una spiegazione!!
1) Cosa fai,
2) cosa ti aspetti,
3) cosa ottieni.

Kindred

Hey Arantor... Not sure which members of the BoD were talking to you - but we've had a roadmap in place for 2.1 since before I resigned in 2012. Spuds and emanuele (and Suki for a little while) were working that while others started working SMCore.

And yes, technically it has been feature frozen (as emanuele points out) and we should have been working on fixing the bugs and getting the stuff out - especially since the alpha was released. (do note, I am not criticizing the devs who have stuck around, turn over has been a huge cause for delays)

I will say this:   Actually, the fact that BoD members don't know details of what is going on within the project is not specifically a problem or even concerning. The BoD should have nothing to do with the project on a day-to-day basis... Except as some members of the board overlap into the team (who should be aware - because, as emanuele says it has been posted to the team boards...) and, of course, the PM for the project is supposed to report "status of the project" to the board (which probably wouldn't include details, but should have included the fact that the roadmap exists).

As for getting you on board...  Come on in, the water's fine. :P (and the kool-aid is tasty)
Please do not PM, IM or Email me with support questions.  You will get better and faster responses in the support boards.  Thank you.

"Loki is not evil, although he is certainly not a force for good. Loki is... complicated."

Arantor

Well, it does seem there has been some crossed wires. Mind you, the people I asked, I didn't ask "is there a roadmap" because that's not a smart question.

The way it was phrased is much more relevant: does anyone know what is supposed to be in 2.1? The general consensus I had in answer to this question was no. The fact there is a roadmap, which was largely complete a year ago, still suggests there is a very large management problem on the project. (And not all the people I spoke to are 'merely' BOD but active project members too)

I'm getting a bit off track at this point. We agree that 2.1 has a roadmap and that it is largely complete, that save for bug fixes and some theme fixes, there is no real reason why it can't be moved onto the next phase of the process, i.e. an official beta release. (I wouldn't consider 2.1 RC status yet, but YMMV)

The real question I'm getting at in this topic is... once 2.1 actually rumbles along, does anyone have any actual idea what to do next? That's why I posted what I'd do were I leading 3.0 development. Call it a suggested roadmap for 3.0.

Note, it is with extreme trepidation that I even consider the possibility of getting involved. The fact that as evidenced in this topic, that there are quite significant communication issues, the fact that there's still a fair amount of mudslinging going on from what I hear, and you know I've always had concerns over my contributions actually getting taken on board. I find it hilarious that I fixed the wrong dates in the calendar a full month ago and despite producing a complete set of updated files for 1.1.x, 2.0, 2.1, the full upgraders and a mod package to fix it for interim users, not a single file was downloaded from the bug reports board. I realise the calendar is a thing that people don't really like but still... if a legitimate, albeit minor, bugfix is produced and not even *looked* at for a month, what hope does that give me about legitimate bugs? I am still very reluctant to get involved with another group project after the events of the last month and I need to be absolutely sure that it is the right thing for both me and SMF were I to properly get involved. (The fact that Git totally does not follow my workflow does significantly impede potential progress, too.)

That said, I have poked my nose in around Github a few times. I do not like what I see. I do not like legitimate usability concerns being ignored as 'I like the current behaviour' and that a legitimate solution to users who don't like how SMF 2.1 does things is that they should just edit the template. I'm sorry, that is not a legitimate solution to a viable criticism. (Mind you, I think 2.1's theme is ugly anyway and the use of popups on profile areas just doesn't work the way it's implemented)
No good deed goes unpunished
All helpful urges should be circumvented

Kindred

Personally, I get confused with github so I can't comment on the posts there, except when people have specifically linked me.

And yes, we should definitely be moving toward addressing the remiangn bugs and getting a beta release done.Not sure about your bug report/code....   I'm sure, now that you've pointed it out, emanuele or oldiesmann will take a look at it.

Yeah,there are always problems with communication... Less with mudslinging, IMO.  One of the things I am trying to work on for the next three months is improving the communication so the next pm to take over can have a smoother transition and a better motivated team. :)

as for beyond 2.1... That is also something that needs to be worked in. ;)

Please do not PM, IM or Email me with support questions.  You will get better and faster responses in the support boards.  Thank you.

"Loki is not evil, although he is certainly not a force for good. Loki is... complicated."

iamjanco

Thanks for starting this topic--as a very wet-behind-the-ears newbie, it's answered a lot of the questions I've had these past few days since joining the forum and installing 2.0.5 in my own little realm.

That said, I've still got questions galore, and one that stands out more than most of the others at this point in time has to do with trying to figure out what works and what doesn't, from a release perspective, without (e.g.) having to wade through the results of the searches through the mods sections (1.x.x and 2.x.x), and their respective forum topics. By that, I mean if a mod says it works with 2.0.5 in the mods section, is it safe to assume it was tested to at least some degree using that version?

Also, what might be handy someday (and I know the mod developers and/or others may or may not like this idea), would be a user ratings system of some sort [read: ratings provided by users in the know; yeah, I know not all users are in the know] that would allow newbies like me to quickly determine if I want to take an in-depth look at a mod and try it, or whether I might want to get myself a 10 foot stick. The way it stands right now, the only things a user has to go on are the comments of others in the forum itself, and/or the number of times the mod has been downloaded by others (keeping in mind that mods that have been around a long time will have a much wider user base). Oh, and then there are those mods made years back, perhaps updated through 2011 or even 2012, whose developers seem to have flown the coop and may have lost interest in them and since moved on to other things. Again, a number of those mod release notes at least imply that they're compatible with 2.0.5, but such leaves me wondering if that's truly the case.

Just some very subtle input from a new kid on the block, who's trying to cut some corners ;) , who really does appreciate the hard work that myriad of open-source developers contribute to the furthering of ::::cough:::: Al Gore's Internet.




mashby

QuoteI have tea, I have a cheese sandwich, I'm unstoppable  :D
I have a beer, but that's only because beer comes in a defined quantity. Maybe I should switch to tea. ;)
Always be a little kinder than necessary.
- James M. Barrie

Arantor

QuotePersonally, I get confused with github so I can't comment on the posts there, except when people have specifically linked me.

Github is very confusing, even to a seasoned developer, but maybe I'm doing it wrong. I do know there are latent issues with Git itself, and judging by the responses to my thread on it, I'm really not the only one. When you have a version control system where people are complaining that it 'messes up their repo', something is very wrong.

QuoteYeah,there are always problems with communication... Less with mudslinging, IMO.  One of the things I am trying to work on for the next three months is improving the communication so the next pm to take over can have a smoother transition and a better motivated team.

This is one of the major problems. The fact that no-one I spoke to seemed to be aware of the actual developmental state of 2.1 is very concerning. The mudslinging is far less of an issue than it used to be, but I'm very much in the 'twice bitten, thrice very shy' camp right now.

As far as beyond-2.1 is concerned, I've laid down what I'd do. The next steps are: 1) Is this what, generally, people would like to see SMF become? 2) Assuming it is, who wants to step up to do it? I know I could step up but I'm very worried about the situation I'd be stepping in to and the people I'd be working with, for reasons that I'd hope some could understand only too well.




QuoteThanks for starting this topic--as a very wet-behind-the-ears newbie, it's answered a lot of the questions I've had these past few days since joining the forum and installing 2.0.5 in my own little realm.

If you have questions, ask them. The odds are that someone here knows the answers.

QuoteIs it safe to assume it was tested to at least some degree using that version?

Yes and no. When a mod is first released, it is tested by the Customization Team, on the version(s) the mod author lists, and that is known to work. Subsequent updates are not reviewed, though.

What is certain is that there is a decent range of compatibility for most mods; most mods that work on 1.1.0 will work on 1.1.18 without any changes. For 2.0, almost any mod from 2.0 RC3 upwards will work on 2.0.5. Sometimes they need a little encouragement, as per How can I install a mod that doesn't work in my SMF version?

If in doubt, run a test site and try it. Mods are essentially just a list of find/replace codes. Sometimes the find/replace codes will trip up over each other.

A rating system was previously in use, and was abandoned due to inappropriate use, e.g. 'mod does not install 1/5' when there's nothing wrong with the mod at all, just that it clashed with something else already in place.

That said, mod development is not entirely something the core SMF developers should necessarily be thinking about, at least not for 2.1 (compatibility isn't bad actually, much better than I expected), but for 3.0 I'm a strong advocate of breaking all compatibility and building something more reliable and flexible that doesn't have the same historical issues; having something that generally allows mods to co-exist and be mostly version independent seems like a great way to go forward to me, and it is in that mindset I suggest the things I suggest for future development.

I should add, for those watching from the sidelines, the concept of a much more rugged plugin manager, along with a support system to handle it, is nothing new, and certainly not new to me. I have built such a system, and oversaw some of the support systems being built, on an SMF derivative and I know just how well it actually works in practice - much better than you might expect. But there is a very legitimate question about how much flexibility such a system provides and I know a number of people are very concerned about mods being given less flexibility; I'd personally rather have a system with better reliability and let flexibility be the price paid.

Big version numbers are the time to do backwards-incompatible changes, and I'd suggest 3.0 is the time to do it. There are plenty of backwards incompatible changes in 2.0 vs 1.1 and the knowledge base around SMF has grown - along with PHP's capabilities - to sufficiently enable a much stronger and more efficient methodology than what we have. There are also very legitimate security concerns that are raised with the current methodology, not to mention very large support requirements (between mod conflicts/installation errors, and permissions issues) and I know that can be avoided - but the price is flexibility.




Quote from: mashby on September 22, 2013, 10:18:39 PM
QuoteI have tea, I have a cheese sandwich, I'm unstoppable  :D
I have a beer, but that's only because beer comes in a defined quantity. Maybe I should switch to tea. ;)

Well, if you want to get picky, I have a mug of tea, I have a cheese sandwich and I'm unstoppable :P
No good deed goes unpunished
All helpful urges should be circumvented

iamjanco

Quote from: Arantor on September 22, 2013, 10:35:40 PM
QuoteA rating system was previously in use, and was abandoned due to inappropriate use, e.g. 'mod does not install 1/5' when there's nothing wrong with the mod at all, just that it clashed with something else already in place.

I'll buy that. I've been doing Wordpress sites for years for clients who refuse to ask the question before clicking Update. That said, again, lots of good info here. That's essentially coming from one segment of those sidelines you implied.

Arantor

Oh sure, the entire ecosystem of third party developers absolutely revolves around the decisions made by the core developers - and everything flows from that. Support systems for third party developments (like the mod site) are really a separate concern.

Most of the issues, as I see them, that are currently present are present because of the way SMF itself handles add-ons (which includes themes). I believe a complete upheaval of the add-on systems will allow for smoother development of 3PD and more importantly ensure reliability and stability for users. This to me is more of a goal than any amount of flexibility; WP has flexibility but it causes so much trouble.


EDIT: I should have also said that I'd drop XHTML and use HTML5 for the doctype, which would allow for all kinds of subtle goodness.
No good deed goes unpunished
All helpful urges should be circumvented

iamjanco

I noticed mention of a new, separate theme for smartphone users, in the roadmap above. With respect to moving forward into 3.x and beyond, has there been any discussion about the possibility of integrating what's needed for smaller screens into the default theme instead of creating a separate theme for (e.g.) smartphones? Only asking because of how other platforms deal with smaller screens, and I'm especially interested in what might need to be done as far as image resizing is concerned (I know what's done elsewhere which, depending on your POV, can have its caveats).

Arantor

Nope, there has not been any discussion of it as far as I know; most of the effort that has gone into it (which for SMF itself is not that much), was on a separate dedicated theme using jQuery Mobile.

There is so much more than just CSS juggling - in a mobile environment you potentially want to be doing more than just hiding elements, because if possible you want to avoid sending them to the user in the first place - lower bandwidth use for mobile is definitely an improvement.

Mind you, SMF ships with 3 oldschool mobile systems built in which are *extremely* low bandwidth.
No good deed goes unpunished
All helpful urges should be circumvented

Spuds

I must say I got a good chuckle out of this thread ....

QuoteSome people, it seems, are curious to know what I would make of SMF were I given the keys to the kingdom,
Translation: You need to get some sunblock on your nether regions, your going to get a nasty burn with all that sunshine :D  Not that there is anything wrong with some warm sunshine in the nether regions,  I must say it feels delicious.

QuoteInteresting, then, that several people I've spoken to lately - including BOD members, didn't even know there WAS a roadmap of any kind. (Which is why they were talking to me in the hopes of trying to get me on board as a dev...)
Ooopsie ... yes realizing there was a roadmap/goals thread would actually mean people would have to be involved in a positive sense, moving things forward, helping, supporting, however all of that is a less creative outlet :D ... Do note that those goals were generated from other very long threads where *gasp* the team was asked what they would like to see ... and then those long threads were culled down to what was requested and how often etc etc ... it was not a list pulled completely out of thin air.  Then posted for *gasp* feedback. 

QuoteI'd start 3.0 development on top of 2.1 while 2.1 is going through late beta/RC stages and use 2.1's build as basis.
Do note those 2.1 goals were developed when smCore (3.0 if you like) was alive and had a timeline, team, repo, etc  So yup that was (eventually) the plan arrived at, just with a bumpy road to get there.

QuoteWait, so a full year ago it was basically ready to go?
Indeed, but at that point most folks working on it had moved on for various reasons so it kind of wallowed around.  Also smCore died around that time so the future roadmap needed to be re-invented.
QuoteActually 6 months after that post everything was already ready (well, a new theme WIP came in because "SMF cannot be released without a new theme").
I tried to push a release twice, but my conditions (like have it live here) for it were never met, so now it's up to Oldiesmann. He mentioned it twice recently, so I suppose he wants to have a release.
We even got through most if not all of the cut line items .. several items/goals we specifically asked for help in order to get others involved, in the end one customizer came forward to lend a hand.   I even remember there were a couple of drama threads about WTF are we doing with 2.1, no one tells us :'(  cuz the developers all live in an tower and are evil (thats the nice version  O:)) ... then things are pointed to the thread of goals and help lol.  But thats just standard stuff that tends to go on in projects, you know all  the fun in dysfunctional distractions ...

Anyway yup, it was to a lets get a beta out mode a long time ago, meaning the bulk of the goal work had been put in place (not saying 100% complete nor bug free, just at a stage point).  There were also some good contributions early on that were included outside of said goals.

Quoteand the kool-aid is tasty
LOL .. don't worry kids, that really *is* lemon/lime kool-aid

Quotelots of items in the first post
Those goals are pretty much what you will find, in one form or another, in several of the forks but you know that ...  I would guess the forks are not necessary SMF ideals centric, not that I'm 100% certain what that means to the end user.  Its also rumored that another fork is on the horizon as well, but you know that too  :D

QuoteGithub is very confusing, even to a seasoned developer, but maybe I'm doing it wrong.
Your crazy ... GitHub rocks ... Damn now I need some sunshine :P

Advertisement: