WTF Bites
-
@boomzilla said in WTF Bites:
@nerd4sale said in WTF Bites:
@boomzilla said in WTF Bites:
Ever tried pasting a SQL query into teams to share with someone? Every time we do this, we get errors about weird characters showing up in the query.
Fucking teams.
Agree, but no problem here with pasting queries.
At least when using the "Code snippet" setting, otherwise all whitespace including linebreaks are removed.I pretty much never use the rich formatting stuff. Spacing and linebreaks seem to be preserved. Maybe next time I'll remember to do that and use the code thingy.
Does your Teams use a different character set than your code editor?
Could be. I don't see anything in Teams that tells me what encoding it's using, but it should be all simple stuff. Editor uses UTF-8. Seems to happen in both Oracle SQLDeveloper and DBVisualizer.
Teams will emojify some character combinations.
That doesn't seem to be the case here though.
Also SQLDeveloper is hot garbage.
It's the worst, except for all the other hot garbage.
-
@boomzilla said in WTF Bites:
Does your Teams use a different character set than your code editor?
Could be. I don't see anything in Teams that tells me what encoding it's using, but it should be all simple stuff. Editor uses UTF-8. Seems to happen in both Oracle SQLDeveloper and DBVisualizer.
Same problem between two MariaDB CLIs with Teams in between, so I'd say it's definitely a Teams fuckup.
-
@boomzilla said in WTF Bites:
Does your Teams use a different character set than your code editor?
Could be. I don't see anything in Teams that tells me what encoding it's using, but it should be all simple stuff. Editor uses UTF-8. Seems to happen in both Oracle SQLDeveloper and DBVisualizer.
Same problem between two MariaDB CLIs with Teams in between, so I'd say it's definitely a Teams fuckup.
Wonder what code page Teams is using these days...
I never managed to find the source of the
chr(137)"evil spaces" I used to find running around in pre-UTF8 times, btw.
-
@boomzilla said in WTF Bites:
Also SQLDeveloper is hot garbage.
It's the worst, except for all the other hot garbage.
DataGrip / the IDEA Pro DB browser are merely warm, their primary failing being being a little bad at indexing.
-
@boomzilla same experience when copy+pasting someone’s snippets from .bashrc. (They have some shit from anaconda in there that changed their environment and I was wondering why shit’s working for everyone but them).
It looks like Teams replaces spaces with some stupid unicode-y whitespace that then don’t work. Probably for “let’s write everything in node” reasons.
-
It looks like Teams replaces spaces with some stupid unicode-y whitespace that then don’t work. Probably for “let’s write everything in node” reasons.
Huh. I was expecting magic quotes to be issue with MS products. Weird spaces in code was the Confluence (and Jira) trick.
-
It looks like Teams replaces spaces with some stupid unicode-y whitespace that then don’t work. Probably for “let’s write everything in node” reasons.
Huh. I was expecting magic quotes to be issue with MS products. Weird spaces in code was the Confluence (and Jira) trick.
As bad practices develop, they adapt.
-
Status: Doing mandatory trainings. Went through the pre-check to get the correct answers.
-
-
-
@Tsaukpaetra said in WTF Bites:
@boomzilla said in WTF Bites:
weird characters showing up in the query.
Is it putting colour codes?
Not in a way that colors the text.
-
@boomzilla said in WTF Bites:
Does your Teams use a different character set than your code editor?
Could be. I don't see anything in Teams that tells me what encoding it's using, but it should be all simple stuff. Editor uses UTF-8. Seems to happen in both Oracle SQLDeveloper and DBVisualizer.
Same problem between two MariaDB CLIs with Teams in between, so I'd say it's definitely a Teams fuckup.
Wonder what code page Teams is using these days...
I never managed to find the source of the
chr(137)"evil spaces" I used to find running around in pre-UTF8 times, btw.Never mind—you have about a dozen other spaces of varying evilness choose from now that are much sexier and more modern.
-
@LaoC Just two days ago I debugged a python script, where the problem turned out to be a NBSP instead of a normal space. Fun! No clue how that got in there; wasn't my code.
And inb4 h8rs: this had nothing to do with significant white space. The template engine did didn't recognize that character as white space but thought it was part of the preceding identifier.
-
-
@ixvedeusi said in WTF Bites:
@LaoC Just two days ago I debugged a python script, where the problem turned out to be a NBSP instead of a normal space. Fun! No clue how that got in there; wasn't my code.
And inb4 h8rs: this had nothing to do with significant white space. The template engine did didn't recognize that character as white space but thought it was part of the preceding identifier.
const char * spaces = " \t\r\n\f\v";
-
@ixvedeusi said in WTF Bites:
@LaoC Just two days ago I debugged a python script, where the problem turned out to be a NBSP instead of a normal space. Fun! No clue how that got in there; wasn't my code.
And inb4 h8rs: this had nothing to do with significant white space. The template engine did didn't recognize that character as white space but thought it was part of the preceding identifier.
const char * spaces = " \t\r\n\f\v";Fine but as with all draconian approaches, the loss attends: now you won't be able to render a lot of the obfuscated C entries.
-
I get an email with two attached Word documents. Clicking the attachments in OWA of course doesn't download them, like a sane email program would, but tries to display them in the embedded O365 Word web-craplet. Which just shows a spinning logo, the web version of a splash screen, and never actually opens the file.
So instead I manually click "show in OneDrive" to see if that works. It displays a page saying that I do not have access to this resource (which it just saved there from my own email account) and that I need to "click this link" to log in to my work account. I click the link and, without further ado or interaction, it displays the file.
-
without further ado or interaction
… but with a redirection through the login.microsoft.com.or.whatever I suppose. You have to go through the login whenever going to another app (outlook, word, excel, onedrive etc. are all separate apps in this regard) even though you are actually logged in to the login site itself so it won't ask you for a password again.
-
I've been assured that The Cloud™ makes everything simpler. So you must be telling filthy lies.
-
@Zerosquare said in WTF Bites:
I've been assured that The Cloud™ makes everything simpler. So you must be telling filthy lies.
Office 365 etc just works IME but it's a Microsoft product so I don't expect it consistently works for everyone.
-
@Zerosquare It does make things simpler, not simple.
If your company admin had to set up all those things it would be much bigger mess and you'd have to keep entering your password into a jillion of slightly different login windows all the time. With (or without) Cloud™ it's still a mess but at least it's now always the same login and it only asks you the password occasionally.
… for people like me with at least 5 different Azure/O365 accounts in regular use it still asks which one I want for this page all the time, and I have to pay attention which one I indeed want. But at least most of the time it does not want the password more than once a day.
-
Messing around with a RPi Pico. Good fun for the most part.
But, fuck, do I hate tinyusb with a passion, and I've barely used it. I can't find any documentation on it. Example: there's a function
tud_task()that you're supposed to call periodically, probably to let tinyusb do USB stuff.How often do you actually need to call it? Does it block? (Can I block and wait for an event?). What happens if I were to not call it for a few seconds or so? Digging around in a header shows that it actually just calls
tud_task_ext()with UINT_MAX as the first argument. This seems to be the maximum wait time. Except .. it doesn't seem to wait in my case. Well, if you run it without an OS, it doesn't actually wait. (The second argument of tud_task_ext is unimplemented in all cases, whatever it supposed to do. Interrupt something?)Can I somehow not run this thing at full tilt?
Gah.
-
-
Messing around with a RPi Pico. Good fun for the most part.
But, fuck, do I hate tinyusb with a passion, and I've barely used it. I can't find any documentation on it. Example: there's a function
tud_task()that you're supposed to call periodically, probably to let tinyusb do USB stuff.How often do you actually need to call it? Does it block? (Can I block and wait for an event?). What happens if I were to not call it for a few seconds or so? Digging around in a header shows that it actually just calls
tud_task_ext()with UINT_MAX as the first argument. This seems to be the maximum wait time. Except .. it doesn't seem to wait in my case. Well, if you run it without an OS, it doesn't actually wait. (The second argument of tud_task_ext is unimplemented in all cases, whatever it supposed to do. Interrupt something?)Can I somehow not run this thing at full tilt?
Gah.
That max, sounds like a max, not a wait, so, you need to make your own timer and call the
_extvariant with something smaller than max. It's a throttle but you need to supply your own gas.
-
There’s a lot of WTF going on here. But this:
this was was really hard
… and wrong.
-
Wow, that's overcounted quite badly. The correct answer, of course, is
1+6+15+20+15+6+1=64=2⁶
-
Or one fewer if you exclude the "no sauce" case, which is probably what they were going for.
-
@PleegWat You're counting metric combinations. McDonaldsUSA is clearly using imperial combinations.
-
@PleegWat the part I really don’t get is 180. How did they come up with that?
(The rest ist just double counting permutations.)
-
There’s a lot of WTF going on here. But this:
this was was really hard
… and wrong.
"6 unique combinations of 1"
-
@PleegWat the part I really don’t get is 180. How did they come up with that?
(The rest ist just double counting permutations.)Permutations count when dipping, tho.
-
@topspin I looked at the rest of the post as well.
is going on there? "McDonalds Free For Life", every combination 15 times with (their number) 343 combinations. Let's say you eat there once/day, that gives you roughly one year to get through (their number) of combinations. 15 times would imply about 15 years.
Oh.I guess they are including the fact that you will die from something like arteriosclerosis if you actually try to do that. (15 years seems like a high estimate.)
-
@cvi "for life" is 2 meals a week for 50 years. 5200 meals.
edit: on the correct maths that's every combination 81 times
-
There’s a lot of WTF going on here. But this:
this was was really hard
… and wrong.
"6 unique combinations of 1"
And what do you think of
"1 unique combinations of 0"
which is missing in their chart?
saw that case earlier. Fits to his name.
-
They should have included dipping order for a total of 1957 permutations (including the "no sauce today, thanks" option), which happens to be the year the original Macdonald's restaurant got its golden arches.
.
-
McDonald's marketing team, days ago:
Okay, any ideas for our next viral campaign? Go.
Ever heard of "nerd sniping"?
-
There used to be a time when you could use Google search as a simple calculator. It would even interpret units and correctly combine / convert them. That used to work, but last year or two it's badly failed for me, leaving me scratching my head where I'd input wrong numbers, until I figured out it's just broken (doesn't respect parentheses correctly, reordering under commutativity creates wildy different results, etc.).
But this ... This is Jeff levels of wrong.
Yes, to convert from feet to cm multiply by a 16 digit value.
Note: I only entered the query, not the bullshit in the formula box.
-
There used to be a time when you could use Google search as a simple calculator. It would even interpret units and correctly combine / convert them. That used to work, but last year or two it's badly failed for me, leaving me scratching my head where I'd input wrong numbers, until I figured out it's just broken (doesn't respect parentheses correctly, reordering under commutativity creates wildy different results, etc.).
But this ... This is Jeff levels of wrong.
Yes, to convert from feet to cm multiply by a 16 digit value.
Note: I only entered the query, not the bullshit in the formula box.
Maybe they added AI to fight bingchat?
-
@topspin The dude's probably lying about his height anyway.
-
@topspin No repro, btw.
You got a short stick on A/B testing.
Edit: there were 142,000,000 results, after all. Some of them are bound to be wrong
-
@topspin The dude's probably lying about his height anyway.
It’s shit from YouTube videos. I wouldn’t care.
-
@Zecc they removed 2 million results in the meantime, it seems.
-
You got a short stick on A/B testing.
Or rather
handling of decimal comma (
) vs decimal point (
)...
-
@BernieTheBernie said in WTF Bites:
You got a short stick on A/B testing.
Or rather
handling of decimal comma () vs decimal point ()...Yes. "Let's compute the intermediate result, convert it to text using German locale (even though everything else is in English), then read it again from the text box using English locale."
Morons.
-
@BernieTheBernie said in WTF Bites:
You got a short stick on A/B testing.
Or rather
handling of decimal comma () vs decimal point ()...Don't think so:
-
@nerd4sale Well, then, there might be a difference between German Feet and Elsewhere Feet.
But the Foot Fetish thread is
-
Maybe they added AI to fight bingchat?
And said AI used WTDWTF threads as a source to learn how to convert between metric and imperial units. Yeah, makes sense.
-
@nerd4sale it’s totally discoursistent.
-
Okay, but seriously…
-
@topspin Why would that be wrong? How dare you correct the AI overloads? You have lost their trust and respect. You're wrong, confused and rude. You are clearly not a good user; Google is just being a good search engine bot thing. They've been clear and polite - a good Google.
