And back out
-
I'm totally willing to sign onto a community petition to switch us to a more stable branch and move us out of the alpha testing gig. I just don't have a ton(ne) of confidence that we'll get anywhere with said petition
I wish that I had the motivation to create that thread.
-
@chubertdev can do it! Yes he can! If he can't do it, NO ONE CAN!
WHOOIOOOOOO!
/me junps in the air and performs acrobatics
-
As I kinda edited into my post -- "WTDWTF administration" has volunteered us to be alpha testers for Discourse. So, this is what we get.
maybe Alex noticed he was running short of Frontpage material and basically decided to subscribe to a WTF-creation service...
like the AP, UPI, or Reuters for Bad Software Development.
-
@chubertdev can do it! Yes he can! If he can't do it, NO ONE CAN!
WHOOIOOOOOO!
/me junps in the air and performs acrobatics
If you have all that energy, then why don't you do it?
-
If you have all that energy, then why don't you do it?
I have taken the plunge:
http://what.thedailywtf.com/t/petition-to-move-off-latest/5535
-
If he can't do it, NO ONE CAN!
Thanks for the confidence in the rest of us.
Screw you, Discourse (and they took away our one offensive emoji)
-
according to discourse we're on stable. which was somewhat of my point.
What are you looking at? From my view of the dashboard, we're here:
Of course, we're at least here:
http://what.thedailywtf.com/t/docker-upgrades/1929/115?u=boomzilla
Which is more granular than whatever the dashboard is picking up. It's my understanding that we were on the tests-passed branch.
-
There've been a couple discussions around & after that time over on meta.d that basically state that latest has very minimal pre-release testing -- the term they use is "basic smoke testing", which I take to mean "the build didn't break".
They have a suite of automated tests, so it's a little bit more than the build not breaking. Honestly, i have no idea how much stuff their tests cover and whether they ever catch anything useful.
I do know that most commits I've looked at haven't touched the test folder. Not that I've looked at a sizable fraction of commits.
-
Sometimes after something has shown back up over and over again, one of them will mention on meta.d that they should add a test case for it. Course that is only after they've gotten regressions several times in a row with people calling them on it.
-
And I'll bet that sometimes they actually add a test, too.
-
True story: When I was making the fix for nested quotes (which has been replaced with a slightly less functional version and then a slightly more functional version) I messed some of the logic up and the tests failed. Someone told me that I should change the tests because they were definitely wrong.
-
or, in JeffSpeak , a regression
No - regression would imply we've had it before.
This one's brand new.
according to discourse we're on stable. which was somewhat of my point.
Hmm. Unless something's changed, I thought we were on tests passed. What are you looking at to get stable?
ETA: http://what.thedailywtf.com/t/docker-upgrades/1929 - we were on tests-passed
Commit hashes would appear to confirm that (see http://what.thedailywtf.com/t/docker-upgrades/1929/115 quoted by @boomzilla above) - unless stable is following tests-passed, which makes the difference rather moot.
-
It's my understanding that we were on the tests-passed branch.
Obviously doesn't include a test for 'View page in Firefox, one of the three most popular browsers worldwide'. But then the whole clusterfuck is run by Jeff 'Fuck Windows Phone' Atwood. And CDCK's first response? 'Do we have a bug filed with Mozilla?'.Says it all.
The fact that I can break Discourse by opening a page is so ridiculous that... I don't have the verbal dexterity to cover just how ridiculous it is. Or the profanities.
I am far from the most skilled software tester in the world (I'm a dev for fuck's sake: we hate testing!), and half the code I write is broken the moment I write it, but I at least try and make it work before sending it for release!
-
regression would imply we've had it before.
Jeff seems define "regression" as "Functionality that once
workedpassed tests that stops working due to an unrelated change". Which is the sort of meaning implied in "regression testing".
-
-
-
-
That would be exactly @Sockbot's birthday
-
Yes mistress PleegWat, I shall appear as summoned.
-
Yes mistress PleegWat, I shall appear as summoned.
<!-- Posted by SockBot 0.15.1 "Zany Zoe" on Tue, 16 Dec 2014 20:13:51 GMT-->
hehe
-
-
Good thing you bumped this thread - I was about to report the same issue.
-
I think that means our instance is literally hosted from a laptop in a stable inside a barn somewhere. And sometimes the horses craps on the laptop which plugs up the fan and makes everything quit.
Not often does a post make me laugh that hard. Well done.
-
it was usable
it wasn't since me and a lot of others couldn't get our accounts to actually exist
-
You are less than a year old and already have 2^12 posts?
Geez, talk about starting early in this industry.Filed Under: You even joined before you were born!!!
-
it wasn't since me and a lot of others couldn't get our accounts to actually exist
By the blakey rule of Windows crashes, since I had 100% success on registering accounts, I declare you to be a liar.
-
it wasn't since me and a lot of others couldn't get our accounts to actually exist
I had no issues. Maybe it just didn't like Belgium.
-
Maybe it just didn't like ■■■■■■■.
Hmm and since it was Murican it got it's geography wrong and didn't allow a lot of others too ... makes perfect sense!
-
No - regression would imply we've had it before.
This one's brand new.
I refer the honourable gentleman to the Discopaedia . Jeff has a bit of a history of calling any old bug a regression
Posting from mobile is a barrier to remembering the format for links
-
Hmm. Unless something's changed, I thought we were on tests passed. What are you looking at to get stable?
huh.... my memory lied to me.
i knew that RAM stick was dodgy.
-
By the blakey rule of Windows crashes, since I had 100% success on registering accounts, I declare you to be a liar.
why did i wait till discourse to register and participate? because CS wouldn't let me in!
trust me, you would have known if I was in CS.... I'm not exactly subtle.
-
CS wouldn't let me in!
I was going to suggest that, being in Maine, you just got caught in the Canadian-blocking filter that was set a little too tight, but it let Swampy in, so obviously that hypothesis doesn't pass muster.
-
since it was Murican it got it's geography wrong and didn't allow a lot of others too ... makes perfect sense!
Please provide references for your work. ;)
-
because CS wouldn't let me in!
Nope, I know you LIE because I got right in! YOU CAN'T REFUTE THAT.
-
Nope, I know you LIE because I got right in! YOU CAN'T REFUTE THAT.
.... channeling blakey rather strongly today, aren't we?
-
CS was buggy, but it was usable and fairly stable as long as nobody was abusing the bugs.
But people were abusing the bugs rather a lot. There were ways to defang the worst of the shit (such as turning off signatures) but there were still a great many ways in which things were wrong.
I was an abuser of the tag system.
-
ANY bug tracking software. because forums don't cut it.
No bug tracking software will fix Jeff's belief that his is the One True Way and that everyone else is Doing It WrongTM.
-
But people were abusing the bugs rather a lot. There were ways to defang the worst of the shit (such as turning off signatures) but there were still a great many ways in which things were wrong.
I was an abuser of the tag system.
My point was CS bugs usually took effort to trigger, Discourse bugs don't.
I was a massive abuser of the signature feature. Commander Kurn, Darth Vader, Cardinal Fang, and of course the ones where I replaced the next user's avatar with Hitler but later swapped out to President Skroob to be less offensive.
I miss @SignatureGuy.
-
-
I miss @SignatureGuy.
we can has him back.
slip me some quiddage and i'll write a module for sockbot that does signature guy. ;-)
-
Bots are much less entertaining than using poorly-implemented functionality in interesting ways.
Also while threads and the front page are now legible, profiles are not:
-
Bots are much less entertaining than using poorly-implemented functionality in interesting ways.
i'll grant you that using them might be, but i'll happily assert that MAKING them is fun as heck!
-
The old forums are still up. If everyone was still using them, we wouldn't have this problem.
Too bad that will never happen, because people don't work that way.
-
-
$base-font-size: 0.875em
How on $planet did anybody think defining font size as 0.875 times font size ever made sense‽
(yes, I understand that referred to commit that fixed it, but still, it made no eFfing sense in the first place)
-
My point was CS bugs usually took effort to trigger, Discourse bugs don't.
I was a massive abuser of the signature feature. Commander Kurn, Darth Vader, Cardinal Fang, and of course the ones where I replaced the next user's avatar with Hitler but later swapped out to President Skroob to be less offensive.
I miss @SignatureGuy.
@morbiuswilters's speech balloon was also quite interesting (even though it's not any more innovative than the stuff you referred to):
-
Haha. I love that. It's so passive aggressive and condescending.
-
Does it work without flash yet? I've wanted to try it for a while but it always tells me I can't use their site since I refuse to install flash.
-
I miss @SignatureGuy.
*sigh* crashed again?
$ sudo supervisorctl status signatureguy signatureguy FATAL Exited too quickly (process log may have details)
Yup, crashed again. DiscoDowntime is a still a barrier to bots.
Filed under: Let’s try `startretries = 999999`
-
I agree with whatever @mott555 posted just above.<t5531p88>