Wedge

Revision: 1028
Author: arantor
Date: 13:13:12, 24 September 2011
Message:
! If the board or message index has a custom class/title specified for the left icon, make use of that. (Probably only of use to add-ons.) (Boards.template.php, MessageIndex.template.php)

! Limited ability to perform removal of add-ons and associated clean-up. It gets upset if it can't remove the files, though, because it doesn't have any ability to let you log in via FTP to fix them at present. (ManageAddons.php, ManageAddons.template.php, ManageAddons.english.php, addon_remove.png)

! Fix for undefined variable in addon-loader as reported by live627. (Load.php)
----
Modified : /trunk/Sources/Load.php
Modified : /trunk/Sources/ManageAddons.php
Modified : /trunk/Themes/default/Boards.template.php
Modified : /trunk/Themes/default/ManageAddons.template.php
Modified : /trunk/Themes/default/MessageIndex.template.php
Added : /trunk/Themes/default/images/admin/addon_remove.png
Modified : /trunk/Themes/default/languages/ManageAddons.english.php

I was avoiding this topic for the most part because I don't like fights. I've been in enough to know that little good comes of them, and invariably when I do join them, things get worse because I always call it as I see it, without much in the way of restraint.

That said, I do want to add my $0.02 here.


"I'm adamant that Pete and I will be the only developers on the first release."

We've never shied away from this being the case. We'll listen to anything people want to suggest, but the end of the day is that it will be our decision what gets implemented. Some things have been suggested and implemented, some things have even had code posted that I'm willing to accept in to the core once I've looked over them again and tested them to make sure there aren't any odd side effects, but the end of the day we're still acting as benevolent dictators.

Interestingly only yesterday I commented about this to a friend of mine, with the observation that every single open source project I've ever known to be actually actively developing has an inner core of benevolent dictators calling the final shots, but wise enough to accept contributions and ideas from anyone.Therein lies the first problem: why do people feel the need to be equal to contribute? Yes, we're not equal, but that's ultimately because we want to have the final say in what goes on. There are multiple goals for Wedge over its life, and some of those really rely on maintaining a single cohesive vision, and given the direction of some of those discussions, it's clear that that's a goal that's never going to be achieved.

You've indicated more than once about the desire (and even pretty much making it a *need*) for the software to have themes as the single dominant control on look and feel, but frankly, I don't think that holds up in real life as I tried to get across more than once. There have to be boundaries on creativity on both sides (between add-ons and themes) so that the two can co-exist nicely, and from the point we had that discussion, it seemed pretty clear to me that we were never going to work together too well, because we're at polar opposites and neither of us is prepared to give too much ground.

Yet, I don't regard anything that's happened as lost time. It's shown limitations and weaknesses in the status quo, and great things have happened as a consequence of all those discussions. Whether you realise it or not, or care or not, you have shaped Wedge in your own way, with the template skeleton stuff. That's huge stuff, it's completely changed how I will be writing mods, and you helped inspire that, so despite our mutual stubbornness, something good came of it. Maybe not everything you hoped for, but more than - I suspect - you expected. Much really as I thought when looking over the discussion about themes supporting add-ons. There isn't a way to handle both perfectly, so either you favour one or the other, or compromise somewhere in the middle, which is what I believe we're doing.

Specifically on the logo, I have to say I wasn't that impressed with the comment about the logo. You might find it a waste of time, but I don't, on different levels. Firstly, it's a relatively harmless way to blow off a bit of creative steam for Nao. Coding is a creative skill, yes, but it's a particular kind of creativity, and he has other creativity to let out, so to speak - so it allows for that.

Plus, it helps do something that we can't otherwise do at the moment: involve the community. I'm acutely aware that we have chosen, thus far, to keep the community at relative arms-length by keeping the code private and even keeping the demo site private. Thus for me, it's one of the few ways we can actually reach out to the community in a way that includes them in what we're doing.

I just want to answer Nao's last point, in a way that does kind of answer it on both sides, if that's going to make sense.Wedge has seen both Nao and myself with the freedom to do things that we've never done before, never been able to do before. It's a power, and it's sadly a little intoxicating. But with that power comes a responsibility.

There is one reason why we're conservative at times, it's because we know, deep down, we have to be. Yes, some things we can go nuts and really go to town on, like I'm doing with the admin panel and in particular add-ons. But I recognise that we can't just up-end things.

As much as I'd love to ignore the community and build our dream platform (which is, ultimately what Nao and I set out to do), we recognise that the community is going to use Wedge. That means it has to be usable, and at times it will have to fall back on the inconvenient truth that people don't always want what we want. They want what other software has. It means that we're bound by expectation to provide most of the same things as other packages, and it does restrict us from being too controversial particularly in the default look and feel area.

Therein lies the problem: you could design the single most innovative default theme ever, and we'd end up rejecting it or removing parts of it, not because it's you, not because it's not awesome: but simply because it's not what people expect, and it would reflect badly on all of us in that situation.

At the end of the day, though, the reason that Wedge is working the way it is, is because Nao and I have vision. It's mostly specific with some vagueness in places, but it's mostly a unified vision of where we want to take Wedge in the future, and the reality is that if someone wants to contribute something that doesn't really fit with that vision, no matter how good it is, the likelihood is that it's going to be kicked back, because that's life, and I know I've seen times where we've tweaked the vision to fit in with reality - but for the most part we have adhered to our vision.

Nao's right, when he says he's more conservative, but he doesn't give himself enough credit, because he didn't take the time to understand why he's more conservative, it's less about him wanting to be so and more because he knows, deep down, that he has to be.

As for being an innovator, I'd argue that by being prepared to buck the trend and go down roads that are less well travelled, or even untravelled, that does make you an innovator, Nao. I mean, what other system implements block management with the facility and ease of use that Wedge has?

---

The only reason I even came back to this topic today was that I was wondering if there was a bigger version of the current logo that I could pimp out my Twitter with. (Bigger version of the triangles, with the word underneath)

AeMe 2.10 was specifically the reason why I added ENUM support :P But it wasn't tested thoroughly, it just means that the code which deals with defaults is a shade buggy, not that it's a hellhole of support nightmare :P

Eeek, I'd better fix that! Thanks for the heads-up!

(I'll fix it in my local copy which has a bunch of related stuff in it and commit it all together.)

So, a progress report.

Enable/disable-time scripts now work, meaning that it's possible to install big and scary add-ons, and even work-around (badly) certain limitations in the add-on manager that will be phased out in due course.[1]

So, I already shared these behind closed doors, but I think I should share them here too.

Details

Details

Read the descriptions, there's more discussion there on the meat of the images.

Right now, there are three add-ons that I'm aware of for Wedge, one is a simple demo that proves what it does, one is WedgeDesk and there's a third, that was posted today as a WIP. I'm excited because it shows the potential of the system very neatly, but also because it's the first add-on that wasn't written by me.

It makes me feel warm inside because it's proof in some way that I'm not just living in a fantasy land of how wonderful writing Wedge add-ons will be,[2] and I have made some UI changes based on things that came out of that add-on, namely reminding me that certain things weren't present and making it easier for authors to provide contact details for their add-ons.

(Even though I have a certain amount of reputation for being an immovable force when it comes to what I think is best, I'm really not. Just have to give me a good enough reason for implementing something.)

1.	I plan to offer adding columns and indexes from the add-on manager itself so it can clean that up on uninstall automatically. Right now that can't be done, because adding new columns has to be done manually - so too would removing them.
2.	That's something I do regret a little about SimpleDesk. I loved the fact that it had a usable plugin system, but that it was built with the relative certainty that I'd be the only person writing for it, which always made me feel a little sad inside.

There's a link in the standard footer to the per-post RSS (RSS 1 format) feed, and has been for a while.

Neither RSS2 nor Atom feeds are linked, and neither is the new-topic feed, even though that last one is probably more useful.

Wedge puts it all in the sidebar:
* all-new-posts and all-new-topics every page
* new posts/new topics for a given board inside that board (including any topic inside that board)
* new replies to a given topic inside that topic

Revision: 1026
Author: arantor
Date: 17:02:55, 23 September 2011
Message:
! (SMF bug, thanks Dragooon) When registering users through instant activation, make sure the right variable is passed to the activate hook. (Register.php)
----
Modified : /trunk/Sources/Register.php


@ I did a quick look back through the code history in SMF. It looks like prior to RC4, a different integration hook was called entirely (and arguably that's also broken), but that in RC4 it changed and the way it changed made it break.