|
Suspicious Dish posted:No. I will call you when C++ isn't a trash language for idiot babies. (So never) k but text me when you get here my parents are gone 
|
#
?
Nov 26, 2014 23:29
|
|
|
|
| # ? Jun 7, 2024 16:04 |
|
|
Oh God. About a year ago I showed a friend how to use interfaces in vb6. He wanted to create an data import routine that got the same data from files with different formats. So he had something like code:code:
|
|
#
?
Nov 27, 2014 01:42
|
|
|
Skuto posted:Java Date is millisecond accurate, so ignoring day/month is not the problem. It's just running the loop one time too much. Maybe I work in Korea, you guys  I'm not copying/pasting direct code, so it's just a quick, possibly incorrect summary from memory. The loop was the funny part, don't get too hung up on the details.
|
|
#
?
Nov 27, 2014 05:17
|
|
|
GCC 5.0 now has Cilk Plus support. I could rant for hours how Cilk "Plus" is a sign of everything that's wrong with (people talking about parallel programming in) this world, but I'll just offer this academic paper that attempts to patch up Cilk "Plus" to the basic functionality level Cilk already had TWENTY years ago. http://dspace.mit.edu/bitstream/handle/1721.1/77032/825818523.pdf?sequence=1 "This implementation will be included with this thesis on DSpace@MIT, and maintained at http://github.com/rubenmp8." "Unfortunately, at the time of writing, this approach currently causes segmentation faults for unknown reasons." MIT (and Harvard) in 2012, people.
|
|
#
?
Nov 27, 2014 10:09
|
|
|
Skuto posted:MIT (and Harvard) in 2012, people. It's a bachelors thesis.
|
|
#
?
Nov 27, 2014 15:05
|
|
|
fritz posted:It's a bachelors thesis. A Masters, actually. Not that it matters, it's just more of the same "we'll tackle the simple problem everyone already solved and ignore the one that's actually interesting". My gripe is with Intel, not the poor student who got to deal with the consequences. He could've actually put up the code on his github, though. But who wants reproducible results anyway.
|
|
#
?
Nov 27, 2014 15:16
|
|
|
A project manager posted a bug on our bug tracker, saying "why isn't this in bold?!?!?!?!?!" A CSS developer was confused by things not appearing in bold - and decided to escalate it to myself, an omnipotent developer. Come on, seriously. EDIT: So this is really intentional. Westie fucked around with this message at 17:55 on Nov 28, 2014 |
|
#
?
Nov 28, 2014 17:37
|
|
|
Skuto posted:A Masters, actually. Not that it matters, it's just more of the same "we'll tackle the simple problem everyone already solved and ignore the one that's actually interesting". My gripe is with Intel, not the poor student who got to deal with the consequences. OK I'm confused about the cilk vs cilkplus history, it looks like MIT maintained cilk up thru about 2010, and meanwhile "cilk arts" licensed something and got bought out by intel who is now doing ???? with it.
|
|
#
?
Nov 28, 2014 17:54
|
|
|
fritz posted:OK I'm confused about the cilk vs cilkplus history, it looks like MIT maintained cilk up thru about 2010, and meanwhile "cilk arts" licensed something and got bought out by intel who is now doing ???? with it. Cilk Arts was an MIT startup to commercialize it. http://www.mit.edu/~jfc/Intel.html BTW: "18. I don't like Microsoft Windows. I was warned before I joined that that attitude was 'career-limiting.'" I'm totally a Unix guy myself but people who complain about having to support the platform(s) on which your paycheck gets made get no sympathy. Edit: Given that I brought up this rant I'll elaborate: Cilk combined a work-stealing scheduler with abortable, hierarchical parallel tasks. This is a very powerful combination for problems that aren't purely data-parallel and need some speculative work dispatching, and allows you to parallelize a bunch of problems that aren't embarrassingly (data-) parallel and that stuff like OpenMP is less suitable for. It also had some nice features such as analyzing the ideal possible speedup and critical path of parallel algorithm. Unfortunately, they had some problems with those features in C++ and instead of solving the problem they gave up and now have something that solves the same problem everyone else already solved. Hiowf fucked around with this message at 18:19 on Nov 28, 2014 |
|
#
?
Nov 28, 2014 18:09
|
|
|
Genuinely surprised that isn't "Micro$oft Windoze".
|
|
#
?
Nov 28, 2014 18:17
|
|
|
2010 and can't manage VNC, the worst unixer.
|
|
#
?
Nov 28, 2014 18:19
|
|
|
Seeing people talking about node.js things reminds me of this from the NodeOS (seriously) documentation:
|
|
#
?
Nov 29, 2014 18:28
|
|
|
pretty much the entirety of nodeOS is a horror and that's coming from a guy who actually uses node and enjoys it
|
|
|
#
?
Nov 29, 2014 18:29
|
|
|
Is the real horror using an insecure protocol when you genuinely believe it's insecure?
|
|
#
?
Nov 30, 2014 01:35
|
|
|
xtal posted:Is the real horror using an insecure protocol when you genuinely believe it's insecure? Maybe if there were other choices.
|
|
#
?
Nov 30, 2014 01:44
|
|
|
xtal posted:Is the real horror using an insecure protocol when you genuinely believe it's insecure? Depends on what you use it for and whether it's possible to use an alternative.
|
|
#
?
Dec 1, 2014 09:58
|
|
|
xtal posted:Is the real horror using an insecure protocol when you genuinely believe it's insecure? Not really. Knowing something is insecure and in what ways allows you to build in other security to compensate if its necessary for the project. If you just go around assuming the security of various things then you're liable to get screwed over real hard when it turns out the thing you were putting the entire burden of security on isn't actually secure. "Why does our wireless network need a password? Our network shares already have passwords on them!" "Why do our network shares need passwords? Our wi-fi points have passwords on them!" Also, discussing whether or not certain standards like HTTP are secure is silly considering that many times security breaches happen through the implementation of protocols and security standards. The Heartbleed vulnerability was an example of this. Oh , don't worry about, SSL is secure! Meanwhile, the actual implementation left large swaths of the internet unprotected for months on end. Even if the standards themselves are theoretically secure you never know when one piece of the network stack is going to fail at implementing a protocol in a secure way. ErIog fucked around with this message at 10:36 on Dec 1, 2014 |
|
#
?
Dec 1, 2014 10:32
|
|
|
It's worth noting that that section was updated very recently: https://github.com/NodeOS/NodeOS/wiki/asgard#authenticationquote:HTTP + SSL is secure. If you think otherwise... It might have been a typo or a markdown flub or something. Then again, it could be node devs trying to make their own linux are dumb as hell.
|
|
#
?
Dec 1, 2014 23:59
|
|
|
Munkeymon posted:Then again, it could be node devs trying to make their own linux are dumb as hell. Well, yeah, last time they thought /usr/bin/env was a more integral component to the system than the libc. Currently they've decided to take all the commands they run and append `|| exit $exit_code`, which is an extraordinary waste of effort, especially since if they re-order commands it's going to be a bugger to keep up to date. Also, they're using exit codes over 255, which means some will be aliasing, since exit codes are modulo 256. They've at least discovered that they need to build the compiler and libc, though they don't seem to be making any particular effort to actually build any of their components with the new compiler, nor link their binaries against the libc they build, so it's actually building against what they've got on the host, which just happens to work because he's using the same versions. Frankly this is no more useful than the previous version, where they copied the required files out of the host system. I'd try to explain what they're doing wrong, but the mountain of mistakes implies they wouldn't understand. I'm not sure why I spent an hour looking through this project to see what it actually does
|
|
#
?
Dec 2, 2014 01:09
|
|
only
|
Munkeymon posted:HTTP + SSL is secure. If you think otherwise... 2 out of 3 of the major browsers seem to think otherwise, yes.
|
|
#
?
Dec 2, 2014 02:37
|
|
|
NodeOS threw me for a loop there for a second since I was mistaking it for JNode which is similar in that it's an OS written using a questionable language at best. It's different in that the authors obviously know what something like an interrupt handler is since they wrote it, but they wrote it in Java so it runs like a beached whale OD'd on Russian bathtub heroin.
|
|
#
?
Dec 2, 2014 04:31
|
|
|
Skuto posted:2 out of 3 of the major browsers seem to think otherwise, yes. I'm at least willing to assume it means SSL/TLS/whatever the gently caress makes data in motion safe since SSL 3 has only somewhat recently been totally broken and not everyone has updated their jargon mappings yet.
|
|
#
?
Dec 2, 2014 08:55
|
|
|
Munkeymon posted:It's worth noting that that section was updated very recently: https://github.com/NodeOS/NodeOS/wiki/asgard#authentication Nope!
|
|
#
?
Dec 2, 2014 09:47
|
|
|
Munkeymon posted:I'm at least willing to assume it means SSL/TLS/whatever the gently caress makes data in motion safe since SSL 3 has only somewhat recently been totally broken and not everyone has updated their jargon mappings yet. Encryption is (reasonably) fine. Authentication, though, is totally hosed and no-one has any idea how to fix it.
|
|
#
?
Dec 2, 2014 10:25
|
|
|
Does Mathematica count? http://mathematica.stackexchange.com/questions/66538/how-do-i-draw-a-pair-of-buttocks code:
|
|
#
?
Dec 2, 2014 19:43
|
|
|
php:<? // more horrors above insertdata(); retrievedata(); print_tables($content, $periods, $total); echo "</pre>"; exit(1); ?> </BODY> </excel-tab> ?>
|
|
#
?
Dec 3, 2014 17:36
|
|
|
</excel-tab> wtf is that
|
|
|
#
?
Dec 3, 2014 17:54
|
|
|
down with slavery posted:</excel-tab> wtf is that Some internal HTML-esque poo poo we use to convert reports into excel spreadsheets. I dare not look into that abyss. Also, following the above garbage: php:<?
function insertdata()
{
global $database;
// ...
}
function retrievedata()
{
global $content, $periods, $database, $total, $filters_array, $section_list;
// ...
}
function print_tables($content, $periods, $total)
{
global $section_list;
// ...
}
?>
IT BEGINS fucked around with this message at 18:37 on Dec 3, 2014 |
|
#
?
Dec 3, 2014 18:14
|
|
|
Dear awful CRM system, you won't internally convert an int32 to a decimal and I hate you for it.
|
|
#
?
Dec 4, 2014 18:38
|
|
|
Are all java developers poo poo at anything and everything involving databases or is it just the ones I deal with? Today's entertainment was reformatting a slow query to remove delights like left joins hidden in the where clause on yet another string being executed directly because apparently stored procs are against nature or something.
|
|
#
?
Dec 4, 2014 19:52
|
|
|
A disappointingly large amount of developers are terrible at databases, regardless of language.
|
|
#
?
Dec 4, 2014 19:59
|
|
|
Java is poo poo, full stop. 
|
|
#
?
Dec 4, 2014 21:28
|
|
|
TinTower posted:Java is poo poo, full stop. Java is actually a decent language. It's just that a lot of people using it are awful.
|
|
#
?
Dec 4, 2014 21:36
|
|
|
Apparently the reason for strings instead of procs is "java allegedly handles transactions better" but that was from a .net dev who said she didnt know what the gently caress they were doing either so I'm more inclined to believe thatTinTower posted:Java is poo poo, full stop.
|
|
#
?
Dec 4, 2014 21:37
|
|
|
Powerful Two-Hander posted:Are all developers poo poo? Yes
|
|
#
?
Dec 4, 2014 22:27
|
|
|
Just had to update about 200 ASP files that broke because the Access DBs they reference (directly) were moved to a different drive.
|
|
#
?
Dec 4, 2014 23:21
|
|
|
That would be an excellent forum sub title. "Cavern of COBOL: Are all developers poo poo? Yes"
|
|
#
?
Dec 5, 2014 00:23
|
|
|
today i called a method called 'check transaction history' and it started generating images
|
|
#
?
Dec 5, 2014 00:28
|
|
|
MALE SHOEGAZE posted:today i called a method called 'check transaction history' and it started generating images How else are you going to check the transaction history? You have to look at the images!
|
|
#
?
Dec 5, 2014 00:32
|
|
|
|
| # ? Jun 7, 2024 16:04 |
|
|
Powerful Two-Hander posted:Are all java developers poo poo at anything and everything involving databases or is it just the ones I deal with? Speaking as a Java developer, what's a stored procedure? We just use whatever Hibernate generates. If the database is slow, just throw more hardware at it. (No, seriously, I don't think I've ever worked on a project that used stored procedures. Or involved a DBA in development at all.)
|
|
#
?
Dec 5, 2014 01:31
|
|