|
Protocol7 posted:Passwords and whatnot changed for obvious reasons. There's two PowerShell scripts on our company network shares that have hardcoded credentials as well. They no longer work, yes I've tried them, but that they were ever there horrifies me.
|
#
?
Feb 1, 2022 18:16
|
|
|
|
| # ? May 23, 2024 14:35 |
|
|
Protocol7 posted:Passwords and whatnot changed for obvious reasons. At least it's using === to do it.
|
|
#
?
Feb 1, 2022 18:19
|
|
|
OddObserver posted:At least it's using === to do it. Added layer of security. 
|
|
#
?
Feb 1, 2022 18:27
|
|
|
Having had to write an API Gateway in AWS using a lambda function to handle auth to an S3 bucket because credentials in client side JS is bad wrong this made my eye twitch. Edit: having to do babbys first what is MVC lecture to a group of non technical people but have sign off on the technical design of a project. BigPaddy fucked around with this message at 18:56 on Feb 1, 2022 |
|
#
?
Feb 1, 2022 18:32
|
|
|
I also recently removed a hard coded token from JS that was working.
|
|
#
?
Feb 1, 2022 22:57
|
|
|
Protocol7 posted:Passwords and whatnot changed for obvious reasons. That is one way to interpret 2 factor authentication
|
|
#
?
Feb 2, 2022 21:06
|
|
|
Votlook posted:That is one way to interpret 2 factor authentication Something you have (lovely code) and something you are (a lovely coder)... checks out
|
|
#
?
Feb 2, 2022 21:10
|
|
|
https://twitter.com/nilscaspar/status/1488966680378109957
|
|
#
?
Feb 2, 2022 23:29
|
|
|
Criptobros code is a spaguetti ball of bad code. https://twitter.com/marcan42/status/1489116645440507904
|
|
#
?
Feb 3, 2022 08:51
|
|
|
Rest of the thread is amazing too: https://twitter.com/marcan42/status/1489120184392908801
|
|
#
?
Feb 3, 2022 09:31
|
|
|
Meh, I've gotten attached to it now. Might keep it.
|
|
#
?
Feb 3, 2022 14:51
|
|
|
C++ code:
|
|
#
?
Feb 3, 2022 20:44
|
|
|
Perfectly legal as long as that is not in a header. It is in a header isn't it.
|
|
#
?
Feb 4, 2022 14:41
|
|
|
Yes, that's the official API for gettext. It's gross but marking translateable strings kinda is something that merits a single-character prefix.
|
|
#
?
Feb 4, 2022 19:28
|
|
|
Plorkyeran posted:Yes, that's the official API for gettext. It's gross but marking translateable strings kinda is something that merits a single-character prefix. It's also not that likely to collide with other stuff, unlike, say, Xlib.h or windows.h #define's
|
|
#
?
Feb 4, 2022 19:31
|
|
|
Plorkyeran posted:Yes, that's the official API for gettext. It's gross but marking translateable strings kinda is something that merits a single-character prefix. Should be updated to use a UDL.
|
|
#
?
Feb 4, 2022 19:42
|
|
|
Plorkyeran posted:Yes, that's the official API for gettext. It's gross but marking translateable strings kinda is something that merits a single-character prefix. No. Also afaik it is technically reserved identifier when global, like a macro. OddObserver posted:It's also not that likely to collide with other stuff, unlike, say, Xlib.h or windows.h #define's Unless you try to write C++ like it was a saner language C++ code:
|
|
#
?
Feb 4, 2022 20:20
|
|
|
There was a proposal to use _ as a wildcard for pattern matching in C++23 (or whenever), but apparently that has now been superseded by __.
|
|
#
?
Feb 4, 2022 20:41
|
|
|
Zopotantor posted:Should be updated to use a UDL. Well given that it's a C library... But yes, a UDL would be a a much better way to do something similar in C++.
|
|
#
?
Feb 4, 2022 20:46
|
|
|
Zopotantor posted:There was a proposal to use _ as a wildcard for pattern matching in C++23 (or whenever), but apparently that has now been superseded by __. I'm still amazed they ever got rid of trigraphs
|
|
#
?
Feb 5, 2022 09:56
|
|
|
IBM fought long and hard to keep them. It makes me wonder about their code.
|
|
#
?
Feb 5, 2022 10:47
|
|
|
Nalin posted:IBM fought long and hard to keep them. It makes me wonder about their code. There's no mystery: several IBM customers -- notably banks, and I suspect also military clients -- still use EBCDIC instead of ASCII in various codebases that have been around since the mainframe days. You can read IBM's reply to the committee's "how many backwards yokels still use this crap?" inquiry here, if you want to briefly live in a world where IBM-only codepages are still a valuable contribution. They're also not wrong, entirely. The standard should strive very hard to be backwards compatible, you shouldn't need to worry that a feature you rely on will be going away just because it is niche. Trigraphs were just sufficiently ruinous to be worth removing anyhow.
|
|
#
?
Feb 5, 2022 16:16
|
|
|
I love that PDF and the sheet level of three-piece suit it exudes. The part I couldn't figure out is: what kind of old code needs both to retain trigraph support *and* to get updated to C++17? Shouldn't extremely legacy software be entrusted to extremely legacy language versions?
|
|
#
?
Feb 5, 2022 16:32
|
|
|
Xerophyte posted:There's no mystery: several IBM customers -- notably banks, and I suspect also military clients -- still use EBCDIC instead of ASCII in various codebases that have been around since the mainframe days. You can read IBM's reply to the committee's "how many backwards yokels still use this crap?" inquiry here, if you want to briefly live in a world where IBM-only codepages are still a valuable contribution. EBCDIC does seem to have an _, though, so that's not it in this case. And yeah, IBMs ability to keep running 50 year old software is actually pretty admirable.
|
|
#
?
Feb 5, 2022 16:35
|
|
|
NihilCredo posted:The part I couldn't figure out is: what kind of old code needs both to retain trigraph support *and* to get updated to C++17? Shouldn't extremely legacy software be entrusted to extremely legacy language versions? It's not uncommon for very old code and very new code to live side-by-side in that sort of still-used legacy system. You might reasonably want to keep the battle-tested, bug-free ancient C code intact while using C++17 in the shiny new code that calls it from one file over, and that's a lot easier if the same compiler and settings mostly works for both because the language hasn't had any major breaking changes. That sort of compatibility is also helpful if you're trying to modernize legacy code, since then you can more easily do so incrementally. I can understand IBM's position. It's just that, unfortunately for them, trigraphs are the worst.
|
|
#
?
Feb 5, 2022 17:11
|
|
|
Some might say that the worst trigraph was "IBM" all along
|
|
#
?
Feb 5, 2022 17:15
|
|
|
CPColin posted:Some might say that the worst trigraph was "IBM" all along "C++" is up there too.
|
|
#
?
Feb 5, 2022 17:18
|
|
|
NihilCredo posted:I love that PDF and the sheet level of three-piece suit it exudes. also the sheer level of bad writing
|
|
#
?
Feb 5, 2022 17:45
|
|
|
Xerophyte posted:You might reasonably want to keep the battle-tested, bug-free ancient C code intact while using C++17 in the shiny new code that calls it from one file over, and that's a lot easier if the same compiler and settings mostly works for both because the language hasn't had any major breaking changes. Xerophyte posted:That sort of compatibility is also helpful if you're trying to modernize legacy code, since then you can more easily do so incrementally. ExcessBLarg! fucked around with this message at 02:33 on Feb 6, 2022 |
|
#
?
Feb 6, 2022 02:22
|
|
|
Well, you see, IBM provides perfect backwards compatibility for mainframes. This means that your environment will always be perfectly the same, including the old keyboards that led to creation of trigraphs.
|
|
#
?
Feb 6, 2022 15:39
|
|
|
poo poo, you think those keyboards are capacitive buckling spring? Where can I buy new?
|
|
#
?
Feb 6, 2022 16:10
|
|
|
You get one free with the purchase of a mainframe.
|
|
#
?
Feb 6, 2022 16:28
|
|
|
ExcessBLarg! posted:poo poo, you think those keyboards are capacitive buckling spring? Where can I buy new? This was a joke question, but there's a company called Unicomp that bought IBM's old keyboard factory and offers updated designs with the same mechanism.
|
|
#
?
Feb 6, 2022 20:01
|
|
|
Yeah but those are membrane buckling spring. Close. I have two Model Ms I've been using for 20 years, but the Model Fs are where it's at.
|
|
#
?
Feb 6, 2022 20:57
|
|
|
https://twitter.com/bgluckman/status/1491576221649358848 Car have a media system integrated. The radio subsystem download files from HD radio stations and try to interpret them based o the extension [!]. Local radio uses a image withouth extension. This file enters the cache of the system, and causes a unexpected exception, that reboot the device... forever. Has a result, a type of vehicle gets frozen to only play this radio station forever.
|
|
#
?
Feb 10, 2022 08:54
|
|
|
Mazda just loves putting these little easter eggs in their car radios, don't they. https://99percentinvisible.org/episode/the-roman-mars-mazda-virus/
|
|
#
?
Feb 10, 2022 15:22
|
|
|
https://www.bbc.co.uk/news/uk-england-tyne-60369098quote:An energy firm has apologised after 74 customers hit by power cuts during Storm Arwen accidentally received compensation cheques for trillions of pounds. Smells like a software issue that wasn't identified before the cheques went out, although who knows. Wonder what their system was doing.
|
|
#
?
Feb 13, 2022 19:00
|
|
|
Hammerite posted:https://www.bbc.co.uk/news/uk-england-tyne-60369098 In freedom land they're accountable for the value on the check. Not sure if that's also true in the UK, but lol if they go bankrupt because they wrote bad checks.
|
|
#
?
Feb 13, 2022 19:47
|
|
|
leper khan posted:Not sure if that's also true in the UK, but lol if they go bankrupt because they wrote bad checks. It isn’t and they won’t.
|
|
#
?
Feb 13, 2022 19:53
|
|
|
|
| # ? May 23, 2024 14:35 |
|
|
DoctorTristan posted:It isn’t and they won’t. tyranny
|
|
#
?
Feb 13, 2022 20:21
|
|