WTF Bites
-
You call that “free WiFi”?!? Lol, fuck you.
Also, for one day that’s half of what I pay for mobile data in a month.
-
Chrome WTF of the day:
I'm preparing to do some screen recordings for documentation. So I figured I'd hide my bookmarks bar and do it in a separate window. So I pull the tab off to a new window, then uncheck "show bookmarks bar". And it does hide the bar...in the original window. Not in the new one I'd just opened, which is the one I wanted to hide.
Might be just Chrome for mac, though.
-
You call that “free WiFi”?!? Lol, fuck you.
Also, for one day that’s half of what I pay for mobile data in a month.You have to scroll down and find the button that's not a button.
-
@Benjamin-Hall said in WTF Bites:
Chrome WTF of the day:
I'm preparing to do some screen recordings for documentation. So I figured I'd hide my bookmarks bar and do it in a separate window. So I pull the tab off to a new window, then uncheck "show bookmarks bar". And it does hide the bar...in the original window. Not in the new one I'd just opened, which is the one I wanted to hide.
Might be just Chrome for mac, though.
You can toggle the bookmarks bar per window on Mac? Whenever I do it throws the bar up or down on everything but Ingocnito windows...
-
@Benjamin-Hall said in WTF Bites:
show bookmarks bar
what’s a bookmark?Note: characters referenced here are only for comedic exposition due to familiarity. I myself suffer from uses more tabs than is healthy syndrome.
-
@Tsaukpaetra said in WTF Bites:
@Benjamin-Hall said in WTF Bites:
Chrome WTF of the day:
I'm preparing to do some screen recordings for documentation. So I figured I'd hide my bookmarks bar and do it in a separate window. So I pull the tab off to a new window, then uncheck "show bookmarks bar". And it does hide the bar...in the original window. Not in the new one I'd just opened, which is the one I wanted to hide.
Might be just Chrome for mac, though.
You can toggle the bookmarks bar per window on Mac? Whenever I do it throws the bar up or down on everything but Ingocnito windows...
I was expecting one of two things--
- toggle on all windows
- toggle on current window based on focus.
I'd have been fine with either, although #2 would have been ideal.
What actually happened was
NAN. toggle only on the original window, not on any windows created later (even by using the New Window button), despite which one has focus.
-
@Benjamin-Hall Does opening a window after hiding the bar accomplish anything?
Chrome on Windows seems to go all or nothing with the bookmarks bar, even with incognito windows. And it always shows the bookmarks bar on the new tab page, regardless of the state of that setting.
-
@coderpatsy said in WTF Bites:
And it always shows the bookmarks bar on the new tab page, regardless of the state of that setting.
This might be the actual root cause here...Confirmed. When I navigated away from the new tab page, the bar went away.
Still
behavior if you're not expecting that.
-
Grrr...trying to pay a doctor bill for my daughter. Call the office this morning at 8...answering service says office hours start at 10. OK...
Call at 10. "Sorry we're running late. Please call back in 15 minutes." Ugh.
Just called again...get menu...eventually..."For billing, press 4."
"Leave your name and phone number and we'll call you back during normal business hours."
-
From google search below the 3 results it showed me that were absolutely unrelated to my query:
Well, that’s a new one. Translated:
These results might not be what you’re looking for. You can view all
advertisementweb results or view further videos.Is that what those are, the ad results?!Maybe if the search query contains
"American Dad"with quotes around it - because the results to my last query didn’t even contain these words - you should start giving me results that actually contain the fucking search phrase?! What’s the magic incantation to make you search for what I typed today, if putting things in quotes also doesn’t mean anything anymore?Edit:
is me for first misreading it and then mistranslating it. It actually says “all web results”, not “all ad results”. I though it says “werbe” instead of “web”.
Still, it didn’t actually show me what I was searching for.
-
Apparently, Entity Framework (or EF Core, or some particular subset/db dialect that I'm dealing with) doesn't understand
.ToLowerInvariant()
-
Apparently, Entity Framework (or EF Core, or some particular subset/db dialect that I'm dealing with) doesn't understand
.ToLowerInvariant()Yes, I remember that being a thing. You have to use
.ToLower()only if it's still in the "try to make this a SQL Query" stage. Feel free to.ToLowerInvariant()after you've materialized the results though.
-
@Tsaukpaetra I changed it to
.ToLower()for the query. In this case, I needed to match an identifier that's lowercase by necessity in one place, with another one that may be mixed case, and is most commonly all caps in our data.As usual,
is case sensitivity
-
@Tsaukpaetra I changed it to
.ToLower()for the query. In this case, I needed to match an identifier that's lowercase by necessity in one place, with another one that may be mixed case, and is most commonly all caps in our data.As usual,
is case sensitivityNah.
is having something that is an ID that isn't just to be exact-matched everywhere. Numbers don't need case.
-
@Tsaukpaetra I changed it to
.ToLower()for the query. In this case, I needed to match an identifier that's lowercase by necessity in one place, with another one that may be mixed case, and is most commonly all caps in our data.As usual,
is case sensitivity:trtrwtf: is using a case sensitive SQL collation
-
So you have to use different case sensitivities on a case by case basis?
-
@Tsaukpaetra I changed it to
.ToLower()for the query. In this case, I needed to match an identifier that's lowercase by necessity in one place, with another one that may be mixed case, and is most commonly all caps in our data.As usual,
is case sensitivityNah.
is having something that is an ID that isn't just to be exact-matched everywhere. Numbers don't need case.Technically neither part is the actual database identifier, but a (somewhat) human-readable unique abbreviation, that's also got a Fancy Process to match it to a service in a Fancy System.
:trtrwtf: is using a case sensitive SQL collation
I think it actually isn't, but I recently got bit by some case sensitivity with the same (not quite) identifier in another context, so I don't want to leave a landmine in case some
psychomental patientsadistfuture developer or admin turns on case sensitivity in the database.
-
Status: Attempting to get my "App" verified so I can use the YouTube API to make videos public. I need to go through "verification" for this.
Apparently I need to make a YouTube video demonstrating the app?
Oh! And apparently, if the App is for "personal use" (which this basically is) then I don't need verification?
Fucking Google...
-
Courtesy of doi:10.1109/TBCAS.2017.2760298: Lossless. ASCII. Compression.
That's definitely not ASCII, but on closer look it does seem to be some kind of combination of RLE and variable-length encoding for fixed-point numbers.
-
Lossless. ASCII. Compression.
What does that even mean?
At the preprocessing stage, MBioSigs are denoised, down sampled and then transformed to a two-dimensional (2-D) data array. SVD of the 2-D array is carried out and the dimensionality of the singular values is reduced.
Sounds reasonable.
The resulting matrix is then compressed by a lossless ASCII character encoding-based technique.
I guess the answer to the following is in the application, but I'm not going to read further
, so: why not just run it through gzip?“very good” as per the gold standard subjective measure.
-
The resulting matrix is then compressed by a lossless ASCII character encoding-based technique.
I guess the answer to the following is in the application, but I'm not going to read further
, so: why not just run it through gzip?That precludes sharing them on usenet!
-
@boomzilla said in WTF Bites:
The resulting matrix is then compressed by a lossless ASCII character encoding-based technique.
I guess the answer to the following is in the application, but I'm not going to read further
, so: why not just run it through gzip?That precludes sharing them on usenet!
Uuencode FTW.
Wait, what?
-
I guess the answer to the following is in the application, but I'm not going to read further
, so: why not just run it through gzip?I'd have to read through my work VPN to get that far (only the first paragraph is available otherwise, and that's completely irrelevant wittering) so
However, the bits with finding ways to reduce the noise will probably help more than just running the raw data through gzip. I'd expect it to work even better if the denoised, down-sampled data was fed through gzip…
-
However, the bits with finding ways to reduce the noise will probably help more than just running the raw data through gzip. I'd expect it to work even better if the denoised, down-sampled data was fed through gzip…
That’s what I was saying. They do the SVD first and then run it through their “lossless ASCII compression”. The first step is domain specific, the second step surely can be replaced by any standard compression algorithm.
-
Status: You stupid fucks...
Making it incredibly difficult to determine if I can remove a function that (seems to be) not used, because UE4 automatically makes the "Get" function essentially indistinguishable from the function prefixed Get.
Just.... why?
And the fucking variable it's wrapping Get and Set around is already publicly available! And the code inside those functions doesn't do anything special at all!!!
-
@Tsaukpaetra said in WTF Bites:
And the fucking variable it's wrapping Get and Set around is already publicly available! And the code inside those functions doesn't do anything special at all!!!
Well yeah but you need them to bind the object into functional compositions. You can't just call a field a consumer and supplier of F, much less a
BiConsumer<T,F>/Function<T,F>. And what if the field is part of a covariant interface?
-
Status: Attempting to troubleshoot why this replicated cube goes berserk when the Scale property is bumped. The thing is set to replicate, everything should "just work", and yet...
The server disagrees where the object is slightly.
Okay, fine, shouldn't be a big issue, except the collision geometry is what's actually fucked up, and I can't seem to get that information.... Or update it, for that matter.
-
We could make the same joke with ₀ through ₉, but ⅰ through ⅻ are actual lowercase numerals.
You should definitely trying using them.
Sometimes I'm tempted to make a string to int parser that can understand all those stupid unicode things.
Perl's
Unicode::UCD::charinfoor C#'sCharUnicodeInfo.GetNumericValueget you a long way already. The rest is special-casing for non-decimal weirdness like Roman, Tibetan, Tamil or combinations thereof.
-
@Tsaukpaetra said in WTF Bites:
And the fucking variable it's wrapping Get and Set around is already publicly available! And the code inside those functions doesn't do anything special at all!!!
Well yeah but you need them to bind the object into functional compositions. You can't just call a field a consumer and supplier of F, much less a
BiConsumer<T,F>/Function<T,F>. And what if the field is part of a covariant interface?That's C++, not Java. (I don't know if that makes things more or less miserable. Probably more.)
-
@Tsaukpaetra said in WTF Bites:
Status: Attempting to troubleshoot why this replicated cube goes berserk when the Scale property is bumped. The thing is set to replicate, everything should "just work", and yet...
The server disagrees where the object is slightly.
Okay, fine, shouldn't be a big issue, except the collision geometry is what's actually fucked up, and I can't seem to get that information.... Or update it, for that matter.
Floating point wobble?
-
(this could go in about any of the rant threads...)
TIL about
%HOMEDRIVE%. How did I come to this realisation? Suddenly git stopped knowing about my SSH keys. After some digging around, I found thatgit-bashwas actually usingM:\as my home directory (related: fuck every answer to "where isgit-bashhome" that is essentially "it's~, which means your home directory"... stupid morons, if I was asking what~means I would ask what~means, not how to find where it truly is
). Turns out that, if not told otherwise, git-bashuses%HOMEDRIVE%%HOMEFOLDER%as~and in my caseM:\is a home on some network share that I had never ever used, nor even noticed, before, as evidenced by the fact that it was entirely empty (not even one hidden file).Given that everything was working a few days before, my guess is that one update pushed by corporate IT changed some setup and caused
%HOMEDRIVE%to suddenly be defined where it wasn't defined before. So fuck them (and/or Microsoft for that update, but as everyone knows blame, like stupidity, can be freely shared between many parties without reducing).To make things even more frustrating,
%HOMEDRIVE%does not show in the environment variables listed in the settings dialog, so I initially thought this wasn't the issue.The additional layer of
is that when I discovered that we now had this %HOMEDRIVE%variable, I asked a couple of coworkers to check if this was just me or not.Both had it as well. One noticed it (for the same reason as I did) and also noticed that this variables changed on the fly when he plugged/unplugged his laptop from our corporate network, and pointed to either his local user profile (as we all always expected) or to some network share that he knew about but had never used (and wasn't the same as the network share in my case, but we're physically in different offices so that's not a huge surprise at this point).
The other also had this variable and hadn't noticed it (he's not a developer so didn't hit this git issue), but together we saw that it was pointing to a different network share than me (yet he's in the same office, and literally the next desk), and that this network share was on a machine with a naming convention that isn't used in the company for at least 10 years, contained folders that had nothing to do at all with our group (e.g. one was "Finance", another "Marketing documents" etc.), and none of those folders had ever been modified since 10 years or so.
That last example tells me that this whole
%HOMEDRIVE%thing is not something that was thought out and done on purpose, but more likely a random messing up of some script.
-
To make things even more frustrating,
%HOMEDRIVE%does not show in the environment variables listed in the settings dialog, so I initially thought this wasn't the issue.Huh. TIL...
However (in cmd):
>set h HOMEDRIVE=C: HOMEPATH=\Users\mylogin
-
@dcon also if you ever hit this problem, at least for
git-bashsetting%HOME%overrides this%HOMEDRIVE%madness, which is how I fixed it.
-
@dcon also if you ever hit this problem, at least for
git-bashsetting%HOME%overrides this%HOMEDRIVE%madness, which is how I fixed it.Luckily, the only time I use git-bash is for
git reset HEAD^since trying that in cmd doesn't work. All my other git commands do work from cmd. (but good to know!)
-
Luckily, the only time I use git-bash is for
git reset HEAD^since trying that in cmd doesn't work.Shirley the problem is that cmd considers
^to be an escape character and it should work if you writegit reset HEAD^^; which is not confusing at all.
-
Luckily, the only time I use git-bash is for
git reset HEAD^since trying that in cmd doesn't work.Shirley the problem is that cmd considers
^to be an escape character and it should work if you writegit reset HEAD^^; which is not confusing at all.Would it work in Powershell?
-
@hungrier Quick test seems to indicate PS does not consider ^ to be an escape character.
So it would 'work'
but reset to HEAD^^and notHEAD^.(cf. "not confusing at all")
Edit:
it works better than cmd.exe since it does exactly what you are telling it to do.
cmd.exe is the one that induces the inattentive in error.
-
@hungrier Quick test seems to indicate PS does not consider ^ to be an escape character.
So it would "work" but reset to
HEAD^^and notHEAD^.(cf. "not confusing at all")
These sorts of little puzzles are what differentiates Windows for the developer. Keeps you on your toes!
-
@dcon also if you ever hit this problem, at least for
git-bashsetting%HOME%overrides this%HOMEDRIVE%madness, which is how I fixed it.Luckily, the only time I use git-bash is for
git reset HEAD^since trying that in cmd doesn't work.That sort of thing is what I have a GUI for. (Eclipse's git support is genuinely very good; it shows signs of being written by someone who actually understands what's really going on.)
-
That sort of thing is what I have a GUI for
I do most things in a GUI. I haven't bothered figuring out SmartGit's magic incantation.
-
An old-timey WTF bite.
ARPL (Adjust Requested Privilege Level)
ARPL is an obscure instruction, opcode 0x63h, that originated in 1985 with the Intel 80386 CPU. It has no practical application in modern operating systems.
The 80386 CPU also introduced something called V86 mode (Virtual-8086). When the CPU is running a protected mode operating system, switching to V86 mode allows the execution of real mode applications that are incapable of running directly in protected mode (e.g., programs written for Microsoft's MS-DOS).
I first learned about ARPL in 1995 when I read a book called "Unauthorized Windows 95" by Andrew Schulman, in which he explored many of the inner workings of Microsoft's latest verion of Windows. In one chapter, he was disassembling some code and discovered something strange. The code jumped to an address located in his computer's BIOS, specifically, the copyright string.
As mentioned above, the opcode for ARPL is 0x63h which also happens to be the ASCII code for the letter "c".
Although Windows 95 ran in protected mode, it made extensive use of V86 mode to allow compatibility with real mode programs and drivers, and Microsoft's programmers discovered an odd quirk of the 80386: the fastest way to switch from V86 mode to protected mode was by executing an invalid instruction.
So, there are numerous places in Windows where the code jumps to an address that contains 0x63h, ARPL, which is an invalid instruction in V86 mode. And what better place to find the letter c than in the BIOS copyright string.
-
the BIOS copyright string.
Why is that a permanent waste of non-reusable memory when such things were a high premium is the bigger WTF.
-
-
Related:
The hunt for a faster syscall trap - The Old New Thing
The performance of the syscall trap gets a lot of attention. I was reminded of a meeting that took place between Intel and Microsoft over fifteen years ago. (Sadly, I was not myself at this meeting, so the story is second-hand.) Since Microsoft is one of Intel’s biggest customers,
At this meeting, the Intel representatives asked, “So if you could ask for only one thing to be made faster, what would it be?”
Without hesitation, one of the lead kernel developers replied, “Speed up faulting on an invalid instruction.”
The Intel half of the room burst out laughing. “Oh, you Microsoft engineers are so funny!” And so the meeting ended with a cute little joke.
After returning to their labs, the Intel engineers ran profiles against the Windows kernel and lo and behold, they discovered that Windows spent a lot of its time dispatching invalid instruction exceptions. How absurd! Was the Microsoft engineer not kidding around after all?
No he wasn’t.
-
@Tsaukpaetra said in WTF Bites:
the BIOS copyright string.
Why is that a permanent waste of non-reusable memory when such things were a high premium is the bigger WTF.
Because for legal reasons it helps for it to appear in clear in the binary. Duh.
-
-
@Tsaukpaetra said in WTF Bites:
appear in clear
binary
you don't say? 
Such that
stringscan find it, yes, I do say.
-
Status: This motherfucking game is in competition with YouTube for sucking data...
-
So, there are numerous places in Windows where the code jumps to an address that contains 0x63h, ARPL, which is an invalid instruction in V86 mode. And what better place to find the letter c than in the BIOS copyright string.
Why not just execute the ARPL instruction instead of jumping to a random data location you think should contain that code sequence?
-
So, there are numerous places in Windows where the code jumps to an address that contains 0x63h, ARPL, which is an invalid instruction in V86 mode. And what better place to find the letter c than in the BIOS copyright string.
Why not just execute the ARPL instruction instead of jumping to a random data location you think should contain that code sequence?
Makes it really bitchy to strip the copyright...
