Cannot be changed in any way - archiving
-
Thanks for that. Argument aside, blaming a language or a DB engine on a perceived failure of a software is pointless.
I'm working on a web app that has:
- server side rendering for users with JS disabled
- client side rendering for users with JS enabled (preferred scenario since you can cache them clientside). Both server side and client side templates are using mustache templates and are actually the same file
- automatic routing to either JS driven links (hashchange event, dynamic loading) or plain old links if JS is not available
- automatic form submission using AJAX if JS available, regular GET/POST if not
- realtime notifications using websockets ("We don't need no stinkin' poll!")
And the backend (except for websockets) is PHP, fully OOP, namespaces, inheritance and all. So yeah, still some life in it, no?
-
Well, it can be relevant; misconfiguration of MySQL can cause all kinds of interesting and otherwise hard-to-diagnose conditions in the application upon it. But that doesn't necessarily mean it's either's fault.
Oh, and yes, there's certainly life in PHP/MySQL - speaking as someone that has been doing those for some years ;) Even if some other people around here consider it to be a herpes server stack or being part of the ghetto.
-
Well, it can be relevant; misconfiguration of MySQL can cause all kinds of interesting and otherwise hard-to-diagnose conditions in the application upon it.
So if you're doing it wrong, things will go wrong? Wow. MySQL logs do kinda suck though.
I stupidly started with MySQL with the thought of "frack it, I know this, get proof of concept going, then switch to something better"
Guess how that went.
-
The answer is badly? MySQL can be used well. It just frequently isn't. In many ways it is the ideal complement to PHP because that too can be used well but frequently isn't and both are the natural RWTF.
-
Well... it works? The thing is that the whole replacement bit is constantly getting pushed back because... well, let's just say that in addition to being the main developer on C++, PHP and JS parts of the application, I now have to tweak Asterisk as well since, well... Everyone else is out of their depth.
We kinda did things Asterisk wasn't meant to do. I mean it works, works great in fact, but unless you're up on your programming you're gonna get fucked. And most people who know Asterisk are telephony people, not programmers.
I REALLY have to try flexviews soon, lack of materialized views is killing me here.
-
Archived means "complete, forever, and not of any further use. May also be eventually deleted or moved to cold long term storage areas."
So moderators can't "unarchive" a thread?
@codinghorror said:Remember that closed topics allow pretty much everything except for adding new replies -- you can edit, like, delete (your own posts), and so forth.
Why would people be allowed to edit their own posts in a closed thread? It seems like Archived represents what everyone else considers closed, and Closed covers "sort of closed".
-
Posts can be unarchived, yes.
-
Let's forget the semantics angle for a minute because I think it's confusing you.
You have three states for a topic which are not entirely exclusive. We will, for the moment, also ignore that.
Let's see, by any definition of the behaviours applicable:
- Unlocked
- Pretend locked (no replies)
- Actually locked (no replies, likes, edits, deletes)
Why do you have two things that are the same thing? Just because you give the other one a different name, it doesn't suddenly change what it is.
-
- Unlocked
- Pretend locked (no replies)
- Actually locked (no replies, likes, edits, deletes)
You forgot "actually pretend locked" (no replies, no replies, likes, edits, deletes)
-
No I didn't. I said there were three states which were not entirely exclusive and that for the moment we would ignore it.
-
Pretend locked (no replies)
I actually covered this, we need this on BBS:
Without this mid level state we would probably have a lot less stuff here http://bbs.boingboing.net/badges/8/great-post
I think you are saying, get rid of that state and instead move topics to a category that does not allow replying, but I am not sure I would find that too appealing. You are forcing a taxonomy based on topic state.
I do get the semantic issue, and on places like http://meta.discourse.org or here this semantic subtlety means nothing and is a useless concept. But at least one of our customers needs this. I thing long term it may become an admin setting to disable lock or something.
-
I'm not saying that, actually. I'm simply saying: don't have the intermediate state.
If a given site needs to create that intermediate state - fine. If their intended procedures warrant such an intermediate state, sure, some method for it should be available. But it certainly shouldn't be baked into the core as such a fundamental concept.
Consider this: what would BoingBoing have done if the distinction wasn't available as it currently is? What alternative options are there?
-
2 options
- just archive stuff anyway (sucks)
- move every topic after a week to a readonly category
The moving thing could work (and in some ways has advantages), but has its own disadvantages cause now you force a new taxonomy.
Option 2 was not even available when we started BBS for perspective. I think eventually we will probably ship with a different default here, but really its not such a super major issue that we need to address right away ahead of v1, in my opinion. Mainly cause forum admins have complete control here. Nobody would even discover archive if moderators do not use it.
-
Bitcointalk should be Discourse's next target. That site is disgusting. And while you are at it talk to the Litecointalk.org folks. These two parties work pretty closely on changes to wallets security and general patches. Discourse would make an awesome fit for both.
You should also talk to Charlie Lee @ coinbase ((coblee) on either of those shitty forums) about this. He has some influence on the community and he is very approachable on stuff.
He would approve of 'anything' that would invigorate the cryptocoin people.
-
See, that in itself says to me that it's a case of YAGNI - what good is a feature that almost no-one uses? If not YAGNI then designing for edge cases.
I realise the taxonomic changes, but here Discourse is the outlier, not the mainstream.
As for the BCT people, there is precisely zero chance that you'll get them to convert. They don't want Web-2.0-ish by their own admission and Discourse qualifies in that category. That's why they're still running a forum software version that left beta in 2006.
-
See, that in itself says to me that it's a case of YAGNI - what good is a feature that almost no-one uses?
I just find that there are much more urgent YAGNI arguments out there, like the bookmark/favorite annoying split. Having 2 ways for an end user to save a topic for later viewing is something that is constantly impacting end users and needs cleaning up.
Even though I think we will eventually ship with a different archive vs lock default, I just don't feel its the right time to make changes here.
-
Fair enough on both counts. It's not a high priority item, but thank you for engaging in a meaningful discussion on the subject.
Filed under: unlike some other people we could name that would tell me I'm doing it wrong and then cite their blog as evidence for why they believe I'm doing it wrong.
-
Well that's only forever for a fairly limited definition of forever :)
-
Argument aside, blaming a language or a DB engine on a perceived failure of a software is pointless.
While that's indeed true, some software doesn't just pave the road to hell with good intentions, it actively builds a superhighway and the user community puts up flashing neon signs pointing down that way proclaiming it the best way to go. (Which is I suppose the real problem with PHP these days: vast amounts of bad community practice and precious little good, and not much of a push to try to fix things either.)
-
Close your eyes @Arantor and @faoileag because I'm going to link to my blog in an on-topic reply! Oh, the humanity!
The problem with PHP is that it's so badly designed at its core -- like a skyscraper built on a foundation of Silly Putty. You can't really fix that without tearing the whole thing down.
I'll be the first to tell you that computer "science" is mostly bullshit for day to day real world programming, it's craftsmanship for the most part, but when it comes to designing languages you really want your best honest to god Computer Scientists on that. Otherwise, you get PHP and all the pain that goes with it 10 or 20 years down the line.
-
PHP wasn't designed. It was cobbled together by someone for his personal home page. The fact it mutated and mutated over years, yes, that's why we have the pain we do. You think it's bad now? Good thing you weren't there in the PHP 3 days or even the early PHP 4 days.
The problem is that the current generation of PHP devs would be quite happy to ditch so much of the bad behaviour but the presence of PHP on the web makes it extremely difficult to do that. You can't suddenly upend how the language works without breaking large percentages of the web. Like it or not, PHP accounts for a frankly scary percentage of the web these days, and PHP is fucked over because it can't fix itself without significant amounts of pain. Coincidentally, WordPress has exactly the same problem, for all the same reasons.
-
The problem with PHP is that it's so badly designed at its core -- like a skyscraper built on a foundation of Silly Putty. You can't really fix that without tearing the whole thing down.
Quoted for truth.
Those who are boycotting Jeff's blog should still visit this: http://me.veekun.com/blog/2012/04/09/php-a-fractal-of-bad-design/
One of my favorite programming articles ever. Also, I believe, the original source of The PHP Hammer tm.
-
2 options
- just archive stuff anyway (sucks)
- move every topic after a week to a readonly category
Option 3: Build a plugin
-
(Here's a hint: when trying to better an existing product, understand what that product offers out of the box and then better it. They have years of experience on you.)
If you don't, this is the typical result.
Filed under: @ben_lubar's next language?
-
A forum that is predominantly Q&A is a fairly dull place to be, because it has no humanity. It's not a community at that point. I've been to many forums over the years - both for receiving and giving help in my own way - and none of them left me so cold as Stack Overflow did. It's just names on a screen with e-peen all around how much 'wisdom' one has.
And programmers.stackexchange is even worse because many of the questions there require discussion to adequately address, and the site's format and rules prohibit discussion, so those questions are simply closed instead of being addressed. What's left are questions with basically one obvious correct answer (and the first one to post it gets all the rep), and questions with many correct answers because the correct answer doesn't really matter. That's why I'm not there any more.
-
but when it comes to designing languages you really want your best honest to god Computer Scientists on that.
And this is one result. I like Haskell, but a lot of people have a problem with it for some reason.
-
Good thing you weren't there in the PHP 3 days or even the early PHP 4 days.
Oh sweet jesus.
One of my University's final projects ended up getting done in PHP3. At the time, I didn't know any PHP at all, and we let 1 guy do all the code work, and PHP3 was what he chose.15 years later, I wonder... WTF
(thanks for the bbcode/whatever bug that happens if the colon is directly after a word ends, making the motherfucking interrobang graphic damn near useless)
-
You have my deepest sympathies.
The PHP Singularity
