WTF Bites
-
@RaceProUK said in WTF Bites:
@cartman82 Why was he trying to submit a bug through customer service instead of through the proper channels?
I guess he didn't know about them?
Personally, I'd probably click that frowny feedback button and drop it there, even if that probably goes into a black hole that only marketing drones ever look at.
-
@cartman82 said in WTF Bites:
Personally, I'd probably click that frowny feedback button and drop it there, even if that probably goes into a black hole that only marketing drones ever look at.
While not necessarily the best bug report channel, it is at least a step above customer support lines.
-
@RaceProUK said in WTF Bites:
@cartman82 Why was he trying to submit a bug through customer service instead of through the proper channels?
He didn't know.
Still, customer support should be able to tell you "use the feedback feature from within the application".
-
@LB_ but even if customer support directed him in that direction, that's no guarantee anyone would even look at it, let alone fix it.
-
On Google Maps, if you star a location and also zoom in close enough for it to appear normally, its name shows up twice:
-
On Google Maps, if you star a location and also zoom in close enough for it to appear normally, its name shows up twice:
Presumably because stars have names too?
-
@Tsaukpaetra said in WTF Bites:
Presumably because stars have names too?
How else can you wish upon them?
-
@boomzilla said in WTF Bites:
@Tsaukpaetra said in WTF Bites:
Presumably because stars have names too?
How else can you wish upon them?
They're
iEnumerable
, and most people stop after the first.
-
-
Yesterday Firefox wouldn't play YouTube videos. Not as embeds on other websites, not on youtube.com itself. They would just get stuck with a loading icon.
There were no updates (of anything) on my system. No change of behaviour on my part. No errors in the console. Firefox was the only browser affected and it happened even with an addon-less "virgin" profile. I didn't care to investigate further.
Today they play just fine.
-
Yesterday Firefox wouldn't play YouTube videos. Not as embeds on other websites, not on youtube.com itself. They would just get stuck with a loading icon.
There were no updates (of anything) on my system. No change of behaviour on my part. No errors in the console. Firefox was the only browser affected and it happened even with an addon-less "virgin" profile. I didn't care to investigate further.
Today they play just fine.
Welcome to the Internet... (A co-worker just asked 'this update is downloading really slow, should I write a bug?' about an hour ago)
-
should I write a bug?
No, but you'll do that anyway despite your best efforts.
-
YouTube just randomly opened a new tab
https://www.youtube.com/[object%20Object]
while I was switching to a different video in a playlist. The original tab loaded just fine, so...Could be one of my browser extensions is interfering, so I wouldn't place too much blame on YouTube for this...
-
I'm looking at a theater's website. The info page for the play I'm looking at has a sidebar that shows all the dates they're showing it, which is pretty helpful. The problem is, the sidebar is ordered by day of the week, instead of by, well, date. And even that's wrong because we're in Europe and our week starts with Monday, not Sunday like this list.
- Sunday June 25
- Monday September 25
- Monday June 26
- Wednesday September 27
- Thursday September 21
- Thursday September 28
- Friday September 22
- Friday June 23
And the final entry is, for some reason:
- Sunday September 24
-
Youtube content ID blah blah blah
Not the first time this happens, IIRC. I heard about someone's original song being flagged because it later appeared in a newscast.
Of course, WTF-wise, this happening in the first place often pales in comparison to how Youtube reacts to you contesting the claim.
-
how Youtube reacts to you contesting the claim
when both a creator and someone making a claim choose to monetize a video, we will continue to run ads on that video and hold the resulting revenue separately. Once the Content ID claim or dispute is resolved, we’ll pay out that revenue to the appropriate party.
I will admit it took far too long for this to go into effect.
-
It's been raining so much in Montreal this spring, Newegg is confused
-
Seems like you are coming from Canada.
Stay at Newegg
-Go to New, Eh?
-
Checked in a new test that tested some previously untested stuff.
Code coverage went down.
-
@boomzilla said in WTF Bites:
Checked in a new test that tested some previously untested stuff.
Code coverage went down.
Well, did you add tests to cover your new test code?
-
Well, did you add tests to cover your new test code?
Actually, I did check to see if the test code was included in the metrics. It wasn't.
-
-
@Mikael_Svahnberg said in WTF Bites:
Code coverage went down.
Branch or statement?
Statements, I think. Branches might have gone down...pretty sure class and methods did not.
-
@boomzilla Could it be that your test calls otherwise dead code, which the compiler previously threw away?
-
@asdf No.
Here's what I suspect happened:
We test against a sanitized copy of production data. For a lot of (integration) tests, the "setup" involves querying the database for a suitable object to work on. There may be tests where sometimes the object isn't as suitable as it should be (hey, the query found a good object when I wrote the test!) and so when that happens some stuff might not get tested.
I suspect that for whatever reason, something like that happened where a different object was picked. The difference was something like 0.03%. My new test probably added about 20 lines (out of ~40,000) to the set of covered code. Not a huge deal, but one of those moments.
-
In Google Play Books, when you can't copy the text, you can search for it on Google instead and then copy it from the Google search query.
-
@boomzilla said in WTF Bites:
The difference was something like 0.03%. My new test probably added about 20 lines (out of ~40,000) to the set of covered code.
Hmm...my numbers were...off a bit.
Tested Before After Existing/Total Classes 1210 1210 1705 Methods 11566 11566 29707 Blocks 218096 218070 557707 Lines 51162 51117 129646 The total numbers of classes, methods, blocks and lines were identical for both runs (because again, the only thing that changed was that I added a test). So, 4 blocks and 55 lines. And yes, all tests passed both times (379 and 380, respectively).
Ah, flakey tests!
-
The time reporting systems for both my employments sucks. For my old job I have to log into an atrocious schedule. This works as such: Every workday is divided into two parts: before and after lunch. To approve (confirm that I worked) I have to click the part, which bring up a dialog. Then click approve. Then get asked if I really want to approve it. And I have to do this individually for every single part. And I can't approve any part before its end time, I have to wait until it's ended even if there's only 30 mins left. And then at the first day of the next month I have to digitally sign the entire month, after I have approved all my working hours. Because approving once isn't enough anymore. Oh, and the reset password form only accepts an e-mail. Which gets fun if your employer did not attach an e-mail address to your profile. No easy password reset for you! And if you want to add an e-mail to your profile? Fuck you, admins only!
As for my new job it has the login page from hell. Type in username and password? Nope, fuck you! You type in your role (which is "custom" for everyone) and your social security number to be taken to... the password management page! Here you can type your password to login, or change your password, or request a new password. And then you press the report button to make your report and get sent to the wrong page. Then spend a few minutes hunting the correct page so you can just report the no deviations and send for approval. Also, the deviations form is confusing with way too many options, so better have had a 100% attendance!
For a bonus, any browser will treat the role/SSN combo as your username/password. So I guess I'm Mr. Custom using my birthdate as my password because I suck at security. sigh The hoops I have to jump through just to get paid every month. Also, the time reporting is ofc an entirely separate system from the work planning I use daily, so I have a different login for that one.
-
-
-
@boomzilla said in WTF Bites:
Math before coffee? Meh!
Ugh.
There is only one thing good before coffee:
-
@TimeBandit said in WTF Bites:
@boomzilla said in WTF Bites:
Math before coffee? Meh!
Ugh.
There is only one thing good before coffee:
Also .
-
Updating the Hearthstone mobile app through Google Play does not always mean that the Hearthstone app is actually updated. Because it may surprise you with the need to download an additional 1.2GB of data when you open it, which is a mandatory download to play the game. I dunno why they felt the need to invent their own in-app downloader for game data when Google Play is perfectly able to provide that service.
-
@Atazhaia It might not be an in-app downloader. Google Play has a limit on how big the APK for the app can be (I forget exactly what it is, but it's well under a gigabyte), but you can add up to two 2GB files of 'additional data', which is downloaded (via the Google Play API) when you first run the app.
-
@boomzilla said in WTF Bites:
Checked in a new test that tested some previously untested stuff.
Code coverage went down.
Happens to me all the time. Usually because people didn't turn on Istanbul (the code coverage checker)'s "count coverage for everything in this directory even if it's not called" mode, so the first test in a new file tanks the coverage as the whole file gets pulled in.
-
@Yamikuronue
Get 100% code coverage with this one simple trick QA testers don't want you to know!!!
-
@Yamikuronue said in WTF Bites:
Happens to me all the time. Usually because people didn't turn on Istanbul (the code coverage checker)'s "count coverage for everything in this directory even if it's not called" mode, so the first test in a new file tanks the coverage as the whole file gets pulled in.
I can understand that, but definitely not what happened here. I added a new test in an existing test file that tested stuff in classes that were already being tested. In fact, the main effect of the test on coverage was to test branches in a method that weren't previously being tested (but other branches were).
-
@royal_poet and I share Steam libraries with each other, of course, and I was going to try The Witcher III. Then I noticed this:
Seems @royal_poet is an early adopter of Steam and in particular The Witcher III... so early it's from before either of us were born.
-
@Arantor
The more obvious conclusion would be that @royal_poet is Highlander
-
@royal_poet and I share Steam libraries with each other, of course, and I was going to try The Witcher III. Then I noticed this:
Seems @royal_poet is an early adopter of Steam and in particular The Witcher III... so early it's from before either of us were born.
The things I did in my previous life on my trusty mainframe. Only trouble was storing all that hole punched tape to load the game.
-
@Arantor
The more obvious conclusion would be that @royal_poet is Highlanderthere can be only one @royal_poet
-
@royal_poet said in WTF Bites:
The things I did in my previous life on my trusty mainframe. Only trouble was storing all that hole punched tape to load the game.
So that's what the storage units holding those 335,000 boxes of punch cards are... they're the beta for The Witcher III.
-
@Arantor yep. Those devs were groundbreaking for their time.
-
@royal_poet said in WTF Bites:
@Arantor yep. Those devs were groundbreaking for their time.
Backbreaking, too, given all those boxes.
-
@Dreikin yeah that used to be a real drag. And then you do one thing in the game and immediately it's like six punch card changes. Really made it less of a fluid gaming experience.
-
@royal_poet I also imagine WASD controls on a Model M keyboard must have developed hardcore finger strength.
-
@royal_poet I also imagine WASD controls on a Model M keyboard must have developed hardcore finger strength.
What? Why? I would think the rubber chicklet style would be more damaging...
-
@Tsaukpaetra them Modem M keyboards ain't your momma's keyboard, son. Them's got the good mechanical switches. The hard-wearing, long-life, solid-as-a-rock switches.
-
Today I got a followup to my latest WTF. I thought that failing to sign your time report approvals on time would mean you'd not get your salary that month until signed, and having to wait for the next payout date. But it seems that if one person does not sign on time, it could cause nobody in that location to get their salary.
-
@Tsaukpaetra them Modem M keyboards ain't your momma's keyboard, son. Them's got the good mechanical switches. The hard-wearing, long-life, solid-as-a-rock switches.
I always wanted to replace the shift and caps lock keys of my fortran-programming colleagues with one of these. That way, they would at least suffer a bit for typing all uppercase.