|
I'm all for a nice belittling tirade. It's the actual writing that makes me want to punch that guy in the dick.
|
# ? May 31, 2013 20:03 |
|
|
# ? Jun 6, 2024 01:47 |
|
lol zeromq
b0lt fucked around with this message at 04:22 on Jun 1, 2013 |
# ? May 31, 2013 21:29 |
|
What is wrong with you people, attacking the tone of his bug report? Go post on HN or something.
|
# ? May 31, 2013 21:54 |
Yeah, hopefully you won't need long property values while debugging.
|
|
# ? May 31, 2013 21:56 |
|
Volmarias posted:
I'd like to thank the academy. I actually had a whole series planned out but I quit my job instead haha.
|
# ? May 31, 2013 22:04 |
|
nielsm posted:Yeah, hopefully you won't need long property values while debugging. zmq_assert is a bespoke implementation that doesn't disappear in release mode. 2GB limit though, no idea. According to the spec it is supposed to be 2^63-1 not 2^31-1. MrMoo fucked around with this message at 22:17 on May 31, 2013 |
# ? May 31, 2013 22:15 |
|
MrMoo posted:zmq_assert is a bespoke implementation that doesn't disappear in release mode. 2^31-1 == 28
|
# ? May 31, 2013 22:39 |
|
edit: nevermind that's the xor operator
|
# ? May 31, 2013 22:44 |
|
aaah, it doesn't stop. https://github.com/zeromq/libzmq/pull/568 MrMoo fucked around with this message at 23:53 on May 31, 2013 |
# ? May 31, 2013 22:44 |
|
https://github.com/mongodb/mongo-java-driver/blob/master/src/main/com/mongodb/ConnectionStatus.java#L212-215
|
# ? May 31, 2013 23:39 |
|
FateFree posted:I was cleaning out a development folder and I found a file called MP.java. I had no recollection until I opened it and it all came rushing back to me. I suppose this falls under the category of code that makes you laugh. Okay. That was cool.
|
# ? Jun 1, 2013 00:09 |
|
Suspicious Dish posted:https://github.com/mongodb/mongo-java-driver/blob/master/src/main/com/mongodb/ConnectionStatus.java#L212-215 is that a loving joke
|
# ? Jun 1, 2013 00:10 |
|
hieronymus posted:is that a loving joke MongoDB is the joke.
|
# ? Jun 1, 2013 00:40 |
|
hieronymus posted:is that a loving joke It was something about not spamming the logs everytime that exception was thrown, so they only do it sometimes
|
# ? Jun 1, 2013 00:44 |
|
shrughes posted:What is wrong with you people, attacking the tone of his bug report? Go post on HN or something. on behalf of every developer ever: All we want are the facts, ma'am.
|
# ? Jun 1, 2013 04:14 |
|
hieronymus posted:is that a loving joke "heh, look at this hack i just implemented, guys. it's so tight"
|
# ? Jun 1, 2013 05:24 |
|
Slanderer posted:"heh, look at this hack i just implemented, guys. it's so tight" i don't want to imagine the kind of mind that even could think of doing it this way the idea is so bad.
|
# ? Jun 1, 2013 06:25 |
|
shrughes posted:What is wrong with you people, attacking the tone of his bug report? Go post on HN or something. I blame tumblr.
|
# ? Jun 1, 2013 14:35 |
|
Suspicious Dish posted:https://github.com/mongodb/mongo-java-driver/blob/master/src/main/com/mongodb/ConnectionStatus.java#L212-215 Ahahahaha holy gently caress there is no way this actually is in production. (and yet it is )
|
# ? Jun 2, 2013 06:55 |
|
hieronymus posted:i don't want to imagine the kind of mind that even could think of doing it this way the idea is so bad. Come on. Don't pretend you haven't done something like that before when you're coding at 4am and you just need to get it out the door dammit. Toekutr fucked around with this message at 07:03 on Jun 2, 2013 |
# ? Jun 2, 2013 07:00 |
|
Zombywuf posted:I blame tumblr. I mean, I guess the tone of your bug report doesn't matter if you don't care about it actually being fixed and just want sweet sweet internet points. Where did this weird trend of blaming Tumblr for everything bad come from?
|
# ? Jun 2, 2013 09:59 |
|
Hughlander posted:Probably so that you don't spam the log files thousands of times a second with the same error, now it'll only spam HUNDREDS of times a second! The best part about this is that it spams 90% of the time instead of 10% of the time as originally intended, because someone refactored the code without stopping and going "wtf is going on here".
|
# ? Jun 2, 2013 10:06 |
|
yaoi prophet posted:I mean, I guess the tone of your bug report doesn't matter if you don't care about it actually being fixed and just want sweet sweet internet points. Where did this weird trend of blaming Tumblr for everything bad come from? Overly-sensitive social justice warriors who constantly try to one-up each other about how totally inclusive and equality-minded they are while simultaneously attacking everyone else they think beneath them, or the ones who see signs of *insert issue here* even in everyday mundane actions. For some reason this stuff has really manifested itself in certain corners of Tumblr. You non-PoC cis-privileged scumbag.
|
# ? Jun 2, 2013 15:23 |
|
Ender.uNF posted:Overly-sensitive social justice warriors who constantly try to one-up each other about how totally inclusive and equality-minded they are while simultaneously attacking everyone else they think beneath them, or the ones who see signs of *insert issue here* even in everyday mundane actions. For some reason this stuff has really manifested itself in certain corners of Tumblr. Someone explain this post to me.
|
# ? Jun 2, 2013 17:21 |
|
Vanadium posted:Someone explain this post to me. non-PoC => non-Person-of-Color => average white person who has what is termed "white privilege" cis-privileged => cis is the opposite of trans (in the sense of transsexual), so it refers to someone with what is considered average or normal gender/sexuality/orientation privilege => having the benefit of being more accepted or better placed in society, i.e. "normal" white heterosexual male vs. flamboyant black-asian transsexual person. Much of society cuts the former a lot more slack in lots of things tumblr apparently have people who don't just take the above seriously, but plunge into a sort of insanity in pointing out how privileged people are in some way. (note that this doesn't reflect my opinion as I choose not to have one about tumblr)
|
# ? Jun 2, 2013 17:43 |
|
http://www.somethingawful.com/d/news/tumblr-privilege-check.php
|
# ? Jun 2, 2013 17:49 |
|
Oh so it's some sort of reverse-racism thing?
|
# ? Jun 2, 2013 22:20 |
|
Are we seriously going to have an 'is privilege real' discussion in the thread where we make fun of people who are awful programmers? People have been absurdly sensitive on the Internet since it loving began and picking out one particular website and saying 'yes, this, this is the cause' is short-sighted. Tumblr as a site is pretty badly-coded though! Like when one of their engineers was editing the production version of the dashboard page and inserted an i before the <? php, with predictable results.
|
# ? Jun 2, 2013 22:20 |
|
That seems more like a failure in their deployment process than anything else. Then again I have never fathomed why any production web service would display internal errors to the user. evensevenone fucked around with this message at 22:30 on Jun 2, 2013 |
# ? Jun 2, 2013 22:28 |
|
^^^^e: exactlyyaoi prophet posted:Tumblr as a site is pretty badly-coded though! Like when one of their engineers was editing the production version of the dashboard page and inserted an i before the <? php, with predictable results. that site posted:The only solution I see to this is pre-commit syntax checking for committed PHP files. Um...no, the only solution is not changing stuff directly on the production system without testing it first?
|
# ? Jun 2, 2013 22:31 |
|
oh PHP http://blog.kevburnsjr.com/how-to-prevent-a-leak-like-tumblrs posted:[UDPATE] Turns out PHP returns 200 when it encounters a fatal error. Inconceivable.
|
# ? Jun 2, 2013 22:36 |
|
Tamba posted:Um...no, the only solution is not changing stuff directly on the production system without testing it first? None of the suggestions in that article are reliable ways to detect that error. Syntax checkers would miss it. Without a valid processing instruction, they would just assume that the code is supposed to be plain text. In a typical PHP application (including Tumblr's, apparently), the raw code would have been streamed to the browser before an error even occurred. Integration testing might have been the only way to catch it. Lysidas posted:oh PHP The solution he suggests to this problem wouldn't work, either. He registers a shutdown function to change the response code to 500, but by the time it would run, it's too late: the headers have already been sent. You could make it work with output buffering, but Tumblr doesn't seem to be using it. DaTroof fucked around with this message at 22:57 on Jun 2, 2013 |
# ? Jun 2, 2013 22:54 |
|
If what you're trying to solve is specifically leaking the password, wouldn't including that from an external PHP-only config file that you can never accidentally publicly leak be the way forward?
|
# ? Jun 2, 2013 23:21 |
DaTroof posted:None of the suggestions in that article are reliable ways to detect that error. Syntax checkers would miss it. Without a valid processing instruction, they would just assume that the code is supposed to be plain text. In a typical PHP application (including Tumblr's, apparently), the raw code would have been streamed to the browser before an error even occurred. Integration testing might have been the only way to catch it. You could make a requirement that all PHP source files must begin with <?php, that would catch that single type of error. Any serious application (haha php serious haha... right working on such a thing myself) wouldn't have a PHP file that starts out by outputting text anyway. Jonnty posted:If what you're trying to solve is specifically leaking the password, wouldn't including that from an external PHP-only config file that you can never accidentally publicly leak be the way forward? Not really, due to the way PHP includes work. But you could put the database credentials etc. in a real config file which isn't supposed to be read by a Turing-complete interpreter.
|
|
# ? Jun 2, 2013 23:22 |
|
nielsm posted:You could make a requirement that all PHP source files must begin with <?php, that would catch that single type of error. Any serious application (haha php serious haha... right working on such a thing myself) wouldn't have a PHP file that starts out by outputting text anyway. I agree totally, but I also wouldn't be surprised if a shitload of scripts with inline HTML make that assertion scream bloody murder.
|
# ? Jun 2, 2013 23:35 |
|
Jonnty posted:If what you're trying to solve is specifically leaking the password, wouldn't including that from an external PHP-only config file that you can never accidentally publicly leak be the way forward? The normal way would be to include it from a file that a) is outside of webroot, b) has the .php extension. That way, two things have to go wrong for the password to become visible to users: not only does what happened in the Tumblr case have to happen (to the file containing the password), but also the server has to be misconfigured so that things outside of what-is-supposed-to-be-webroot are getting served. As an aside, Firefox's spelling dictionary has "misconfiguration" but not "misconfigured".
|
# ? Jun 2, 2013 23:37 |
|
Jonnty posted:If what you're trying to solve is specifically leaking the password, wouldn't including that from an external PHP-only config file that you can never accidentally publicly leak be the way forward? That's exactly what they did. You can see where they test where /config.php is. But because PHP includes also output raw text content, the /config.php file gave output instead.
|
# ? Jun 2, 2013 23:42 |
|
What about having staging and like, one behavioral test with selenium webdriver?
|
# ? Jun 3, 2013 00:33 |
|
Dren posted:What about having staging and like, one behavioral test with selenium webdriver? This. Why the crap would you ever deploy directly to prod, let alone edit prod directly?
|
# ? Jun 3, 2013 02:18 |
|
|
# ? Jun 6, 2024 01:47 |
|
Lets be serious, the answer is because it's quicker every time, unless you gently caress it up.
|
# ? Jun 3, 2013 02:37 |