|
https://twitter.com/kennwhite/status/1067133581435305984
|
# ? Nov 26, 2018 21:06 |
|
|
# ? May 17, 2024 19:28 |
|
A Reg article on same that explains a bit more: https://www.theregister.co.uk/2018/11/26/npm_repo_bitcoin_stealer/
|
# ? Nov 26, 2018 22:59 |
|
Ever wanted a PHP implementation written in Go? Neither did I. https://github.com/MagicalTux/goro
|
# ? Nov 27, 2018 00:11 |
|
bigmandan posted:Ever wanted a PHP implementation written in Go? Neither did I. mark continues to do nothing wrong
|
# ? Nov 27, 2018 01:03 |
|
looks like he's using his japanese prison time well
|
# ? Nov 27, 2018 02:19 |
|
bigmandan posted:Ever wanted a PHP implementation written in Go? Neither did I. https://www.youtube.com/watch?v=kY-pUxKQMUE
|
# ? Nov 27, 2018 04:29 |
|
bigmandan posted:Ever wanted a PHP implementation written in Go? Neither did I. now all we need is a go implementation written in php
|
# ? Nov 27, 2018 05:32 |
php...finds a way
|
|
# ? Nov 27, 2018 05:33 |
|
https://twitter.com/felixfbecker/status/1067117685375422464 choice quote: quote:Felix,thank you for comments.The software includes the legal notice of use of MIT license which fully capacitates extendors to make use of it.Please keep in mind the extension code is currently about 100K lines.Friendly reminder,reverse engineering is a violation of license terms
|
# ? Nov 27, 2018 18:19 |
|
if you don't want people to re-wrap a thing and sell it wholesale, don't use MIT as a license.
|
# ? Nov 27, 2018 19:30 |
|
I think the complaint is they didn't include the copyright notice, the only thing the license requires them to do
|
# ? Nov 27, 2018 19:36 |
|
https://twitter.com/DevsenseCorp/status/1067410510289747968 Lol. Their stable release is going to suck then and a loooot of people are going to ask for their money back Guess someone snatched the twitter account away from the intern who'd read too many Wendy's tweets NtotheTC fucked around with this message at 19:56 on Nov 27, 2018 |
# ? Nov 27, 2018 19:53 |
|
Huh I'm surprised even that happened. Around ten years ago I found a half dozen for-profit "security" products that were using WireShark unattributed and reported it to the Wireshark team. I lost track of them after a few years but nothing changed despite a strongly worded "Thanks we're going after these guys hard" email from WireShark. And that's GPL not MIT.
|
# ? Nov 27, 2018 20:00 |
|
nielsm posted:You are developing a Windows application for syncing user photos between a PC and your cloud service. Where would you store a local cache of files? I was pretty sure iCloud syncs your photo library to %USERPROFILE%\Pictures\iCloud Photos. Are you sure about this?
|
# ? Nov 28, 2018 06:39 |
Simulated posted:I was pretty sure iCloud syncs your photo library to %USERPROFILE%\Pictures\iCloud Photos. Are you sure about this? I found this on a user's machine that had disk full. After telling the iCloud client to log out (where it warned this would remove all local copies of things) it removed all data in that C:\Windows\System32\iCloud Photos\ folder.
|
|
# ? Nov 28, 2018 08:49 |
|
RE the strncpy stuff from last page, I'm still amazed how often I come across people, supposedly with lots of C experience, using it as a "safe" copy for null-terminated strings...
|
# ? Nov 28, 2018 09:40 |
|
null terminated strings were a mistake and any use of any c stdlib str* functions is a code smell. please use one of the many libraries that provide sane strings implemented using structs that store the length and capacity alongside the buffer.
|
# ? Nov 28, 2018 09:50 |
|
megalodong posted:RE the strncpy stuff from last page, I'm still amazed how often I come across people, supposedly with lots of C experience, using it as a "safe" copy for null-terminated strings... But I guess from that perspective so are while loops looking for null terminators...
|
# ? Nov 28, 2018 12:35 |
|
bigmandan posted:Ever wanted a PHP implementation written in Go? Neither did I. I like this part of the readme: quote:Why?
|
# ? Nov 28, 2018 12:52 |
|
Soricidus posted:null terminated strings were a mistake and any use of any c stdlib str* functions is a code smell. please use one of the many libraries that provide sane strings implemented using structs that store the length and capacity alongside the buffer.
|
# ? Nov 28, 2018 16:37 |
|
They're probably cool if it's 1850 and you have 3 whole bits of RAM to rub together or embedded systems i guess, but why the hell are embedded systems working with strings
|
# ? Nov 28, 2018 18:25 |
|
RPATDO_LAMD posted:They're probably cool if it's 1850 and you have 3 whole bits of RAM to rub together How else is the thermostat supposed to scroll go pro now at just $7.99/mo to unlock temps above 68F
|
# ? Nov 28, 2018 18:57 |
|
Ola posted:How else is the thermostat supposed to scroll go pro now at just $7.99/mo to unlock temps above 68F You cannot beat the executive explanation: marketing data has shown a 10% subscribers increase since we have added that text. Done and over.
|
# ? Nov 28, 2018 19:28 |
|
68° ought to be enough for anybody
|
# ? Nov 28, 2018 19:58 |
|
Munkeymon posted:68° ought to be enough for anybody 68° ought to be enough for any body.
|
# ? Nov 28, 2018 20:07 |
|
69 would be nice
|
# ? Nov 28, 2018 22:27 |
|
Munkeymon posted:68° ought to be enough for anybody I looked at this post on a bad screen and I thought you said 68" and I was wondering what you were talking about.
|
# ? Nov 28, 2018 22:41 |
|
nielsm posted:I found this on a user's machine that had disk full. After telling the iCloud client to log out (where it warned this would remove all local copies of things) it removed all data in that C:\Windows\System32\iCloud Photos\ folder. Are you sure the user didn't set that directory themselves? I've never, ever seen iCloud store photos in anything other than in the user's Pictures\iCloud Photos directory.
|
# ? Nov 29, 2018 02:13 |
|
Searching it all I can find reference to is the %USERPROFILE%\Pictures\iCloud Photos\ directory, which appears to be the default. I assume it being in System32 was entirely user error.
|
# ? Nov 29, 2018 05:09 |
|
"ASCII text over serial" is a fairly common protocol for talking to equipment from embedded stuff that adds string handling to things. I like to do strings as unterminated fixed size arrays in structs i.e. code:
|
# ? Nov 29, 2018 05:57 |
DELETE CASCADE posted:69 would be nice
|
|
# ? Nov 29, 2018 15:50 |
|
RPATDO_LAMD posted:They're probably cool if it's 1850 and you have 3 whole bits of RAM to rub together What do you think "embedded" covers these days?
|
# ? Nov 29, 2018 18:01 |
I dunno if this is a "horror", per se, but it's a completely ridiculous thing I learned at a recent talk at work. Apparently, CPython doesn't have any concept of primitive types - everything is spun up into a struct with a bunch of ancillary type information. Naturally, creating and maintaining structs is more expensive than just putting a value in a memory location. At some point, someone said "I bet we could improve performance if only there was some way to reduce the amount of cycles we spend creating and garbage-collecting all these structs for frequently used values." I believe this person then proceeded to drink heavily while thinking about the problem, because at some point on the far side of the ballmer peak they decided that the best course of action in this instance would be to preconstruct a set of "small number" objects and return references to those any time an object with that value was needed rather than creating a new object. Our talk didn't cover the exact range (it may vary depending on system?), but the tl;dr is: code:
Whatever makes python faster, I guess? e: thousand != 100 ChickenWing fucked around with this message at 21:21 on Nov 29, 2018 |
|
# ? Nov 29, 2018 18:21 |
|
ChickenWing posted:I dunno if this is a "horror", per se, but it's a completely ridiculous thing I learned at a recent talk at work. this kind of thing is commoner than you might think. e.g. java does the same thing for instances of the Integer class with small values. and the optimisations some languages do to reduce memory usage of short strings are quite impressive. i'm surprised python bothers though given how little thought they normally give to performance
|
# ? Nov 29, 2018 18:52 |
|
CPython refuses to do big structural changes for the sake of perf, but that's a pretty easy localized optimization that doesn't really impact the rest of the code base and has big benefits.
|
# ? Nov 29, 2018 19:08 |
|
I assume that is supposed to say "thousand = 1000" but yeah, this is very common and just a thing you do if you want performance. For example, the .NET async/await system works by wrapping all sorts of results in a Task structure, which is a real waste if the task is short-lived or, even worse, executed synchronously so it just goes to waste right from the start. So it keeps a cache of things like "task that returned with true". Things have improved in the next version, though, where you can skip some of these allocations altogether.
|
# ? Nov 29, 2018 19:15 |
|
EssOEss posted:I assume that is supposed to say "thousand = 1000" It's the difference between "same value" and "same object" so it was intentional.
|
# ? Nov 29, 2018 20:22 |
|
UraniumAnchor posted:It's the difference between "same value" and "same object" so it was intentional. Not the last line, the one that initializes "thousand" needs to be fixed for the demonstration to work. code:
|
# ? Nov 29, 2018 20:42 |
|
Oh and for bonus points if you put that in a file (or compress it to one line) you get a different result, because all instances of "1000" compiled at the same time will share the same object.
|
# ? Nov 29, 2018 20:52 |
|
|
# ? May 17, 2024 19:28 |
|
ChickenWing posted:I dunno if this is a "horror", per se, but it's a completely ridiculous thing I learned at a recent talk at work. As someone already said, Java does this too. What is extremely fun is that IIRC you can use reflection to gently caress around with the conversion slots, so that int with value 2 converts to Integer that has value 4... I mean if you do it, it is your own drat fault, but it is still pretty drat funny.
|
# ? Nov 29, 2018 20:57 |