|
Sapozhnik posted:i want a standard ML that actually has some sort of ecosystem and isn't tied to the JVM or CLR or whatever time to apply to jane street then
|
# ? Jan 19, 2018 08:59 |
|
|
# ? Jun 1, 2024 20:21 |
|
DELETE CASCADE posted:the problem with doing "immutable everything everywhere possible" in java isn't with the language per se, but the garbage collector. java's gc is not designed for the use case of many small immutable objects with short lifetimes whose "changes" can be optimized into mutations (when the previous version of the immutable object is no longer visible). sure you can design a gc for this, but then it wouldn't work so well with traditional imperative java code where large objects have longer lifetimes and get mutated all over the place odd, given how e.g. strings work in java. I guess they special cased a bunch of stuff for them instead of finding general solutions?
|
# ? Jan 19, 2018 11:34 |
|
redleader posted:christ, you're not wrong. as bad as powershell is (and it's bad), batch scripts are absolutely worse is there a better shell script than powershell? actual question
|
# ? Jan 19, 2018 11:51 |
|
NihilCredo posted:is there a better shell script than powershell? actual question
|
# ? Jan 19, 2018 13:11 |
|
JawnV6 posted:yah im p clever like that the problem with thinking of computers as giant state machines that transition instantenously between wholely known states is that abstraction does not hold when you have two points of view that is, when you've got more than one processor looking at the 'curent' state, special relativity says they are going to be different. there's nothing you could do about either, without a synchronization pulse that would basically make using more that one processor pointless this is why programming languages now have memory models, with terms like 'happens before'; assuming changes apply instantaneously and uniformly will burn you
|
# ? Jan 19, 2018 13:21 |
|
the continued inertia of bash poo poo I think is some insane cargo culting it's an entire community built on doing: code:
code:
|
# ? Jan 19, 2018 13:27 |
|
redleader posted:there's a bit of a mentality that if it's not from microsoft, it should be written in-house. hence a lot of stuff is reinvented (badly) or not done at all the sole exception is parsing json, which Microsoft have failed to implement after three (?) tries and now just points everyone at the third party lib
|
# ? Jan 19, 2018 14:00 |
|
MALE SHOEGAZE posted:i think i'm going to try to reboot CGI except instead of perl we'll use go it's callled serverless
|
# ? Jan 19, 2018 15:34 |
|
NihilCredo posted:is there a better shell script than powershell? actual question Theres scripting w/ c# these days but its not common. Powershell is the least bad shell but its still bad.
|
# ? Jan 19, 2018 15:49 |
|
comedyblissoption posted:The primary reason code in Java/C#/C++/javascript/whatever devolves into spaghetti is unnecessary pervasive use of global mutable state. i'm sure this used to be true but now people actually use the alternatives i can confidently say they're sprawling messes too
|
# ? Jan 19, 2018 16:05 |
|
tef posted:i'm sure this used to be true that is unfair, true functional programming without pervasive use of global mutable state has not been properly implemented anywhere
|
# ? Jan 19, 2018 16:19 |
|
Just like communism, right (And this is from a guy who has a repo with planned economy poo poo in it right now, lol)
|
# ? Jan 19, 2018 16:28 |
|
full-stack ReasonML is the future
|
# ? Jan 19, 2018 17:24 |
|
bob dobbs is dead posted:Just like communism, right Hey, it's gotta be better than late-stage OOP, right?
|
# ? Jan 19, 2018 17:27 |
|
lmfao my manager moved my 1-on-1 again... only until this afternoon... if he does it again I'm just going to go home for the day. Oh and in this morning's standup, the engineering VP wanted us to be really pumped that he was able to "get us 6 weeks" of time dedicated to "making this product into what we really wish it could be", i.e. stable and not prone to crashing half an AWS stack. I brought up that 6 weeks seems kind of arbitrary and that we'd still be expected to work on bugs, QA kickbacks, and anything else that came from support. "6 weeks was all I could get and really that should be plenty of time. If we can't get it figured out by that time then we'll need to figure something else out entirely." ...I've started looking for jerbs again
|
# ? Jan 19, 2018 18:09 |
|
Soricidus posted:odd, given how e.g. strings work in java. I guess they special cased a bunch of stuff for them instead of finding general solutions? yes and no. string concatenation in java actually does suck, like if i write a loop that does accumulator += more_string then it will perform really bad, and you're supposed to use a stringbuilder object in that case. in fact it's so bad that in a constant string expression like "x" + "y" + "z" the compiler will optimize it to use a stringbuilder lol Night Shade posted:i was under the impression recent JREs had new-generation GCs designed around short-lived immutable object spam, especially given the rise in popularity of stuff like clojure, scala and more recently kotlin yeah i need to read more about / play around with g1gc
|
# ? Jan 19, 2018 18:49 |
|
if project valhalla ever gets off the ground and java gets real value objects ill be happy. i also hope they get the kotlin compiler to use them when that happens.
|
# ? Jan 19, 2018 18:51 |
|
Brain Candy posted:the problem with thinking of computers as giant state machines that transition instantenously between wholely known states is that abstraction does not hold when you have two points of view there's state in the DRAM, there's state in the L3/L2/L1I$/L1D$, there's the MESI protocol determining what accesses are allowed by each agent, PTE's determining WB/WT behavior. these things aren't the unknowable mystic wonders you're pitching here. i've worked on multicore systems, do you really think once there's 2 processors it's impossible to stop the clocks and dredge out a consistent state? how on earth would you debug a chip like that? at a 4ghz clock you have single-cycle access to anything within 250 picoseconds, or 7.495 cm. i've generally thought this is a particularly asinine way to look at digital systems with electrical signals in silicon, but your "specialty relativity" argument is bunk
|
# ? Jan 19, 2018 19:23 |
|
JawnV6 posted:there's state in the DRAM, there's state in the L3/L2/L1I$/L1D$, there's the MESI protocol determining what accesses are allowed by each agent, PTE's determining WB/WT behavior. these things aren't the unknowable mystic wonders you're pitching here. i've worked on multicore systems, do you really think once there's 2 processors it's impossible to stop the clocks and dredge out a consistent state? how on earth would you debug a chip like that? no? what did you think a synchronization pulse would do? doing work means mutating state; I hope it’s obvious you can’t to do work while everything is stopped.
|
# ? Jan 19, 2018 20:35 |
|
today i accidentally saw a job posting for a 'react native engineer' and now i have brain lesions
|
# ? Jan 19, 2018 20:37 |
|
I just read a quick and dirty article comparing react native and swift performance and some how RN beat swift out in 2 out of three categories e; here it is: https://medium.com/the-react-native-log/comparing-the-performance-between-native-ios-swift-and-react-native-7b5490d363e2
|
# ? Jan 19, 2018 20:38 |
|
cinci zoo sniper posted:a lot Luigi Thirty posted:you are already dead gently caress. what makes groovy so bad?
|
# ? Jan 19, 2018 21:09 |
|
Brain Candy posted:no? what did you think a synchronization pulse would do? im confused by "that is, when you've got more than one processor looking at the 'curent' state, special relativity says they are going to be different." because two cores on a single die have solved this problem deterministically and you're handwaving about how impossible it is
|
# ? Jan 19, 2018 21:23 |
|
i mean, i don't know groovy outside of jenkinsfile scripts but this always amazed me:code:
|
# ? Jan 19, 2018 21:28 |
|
dont get me staretd on jenkinsfiles
|
# ? Jan 19, 2018 21:52 |
|
HoboMan posted:gently caress. what makes groovy so bad? quote:Groovy's syntax permits omitting parentheses and dots in some situations. The following groovy code
|
# ? Jan 19, 2018 22:05 |
|
Valeyard posted:dont get me staretd on jenkinsfiles related: why do so many loving tools (both meanings) need to have their own special snowflake configuration file formats? jenkinsfiles, dockerfiles, nginx.conf, caddyfiles, poo poo.gradle, it goes on and on i give a pass to tools that are two decades old or more, but for anything modern, between xml, json, yaml and toml you can absolutely find a standard format that can describe your options, you'll be able to use an out-of-the-box parser, and you won't make your users waste time on trivial bullshit like braces and punctuation and string escaping
|
# ? Jan 19, 2018 22:13 |
|
unix grognards are doomed to continuously reinvent the registry, badly
|
# ? Jan 19, 2018 22:16 |
|
things do seem to be gradually congealing around yaml at least also groovy is so bad that even its creator has disowned it, saying something to the effect that if he'd known scala existed when he started groovy then he wouldn't have started the project to begin with.
|
# ? Jan 19, 2018 22:16 |
|
NihilCredo posted:related: they probably don't want you to be able to use configs between different things to lock you into using their product or something like that
|
# ? Jan 19, 2018 22:16 |
|
Sapozhnik posted:things do seem to be gradually congealing around yaml at least please no Sapozhnik posted:also groovy is so bad that even its creator has disowned it, saying something to the effect that if he'd known scala existed when he started groovy then he wouldn't have started the project to begin with. please no
|
# ? Jan 19, 2018 22:24 |
|
FAT32 SHAMER posted:they probably don't want you to be able to use configs between different things to lock you into using their product or something like that you can probably write a converter from whatever bespoke horseshit config format to json/xml/whatever without too much trouble if you really need it, so i don't think it would be effective as a lock-in strategy my theory is that for early project prototypes, some quick and dirty config format gets cobbled together, then no one changes it as the rest of the project evolves, then there's enough people using the existing format that changing it is impractical. plus ripping out the config system and replacing it with something else is boring as poo poo to code, so no one wants to do it anyhow the only flaw with this theory is that grabbing some serialization library is less work than writing your own parsing garbage, so i'm not sure why the custom config format would appear in the first place
|
# ? Jan 19, 2018 22:25 |
|
pokeyman posted:the sole exception is parsing json, which Microsoft have failed to implement after three (?) tries and now just points everyone at the third party lib hah, yeah. and even then, that's only because it comes with asp.net out of the box. i don't think it json.net would have gotten much traction otherwise despite our product including json.net for 4 years, i still see people using the others occasionally 3rd party logging libs and di/ioc containers seem to get a bit of use, but only because ms don't (or didn't) have an equivalent nunit is somehow still a thing, despite the fact that all testing frameworks are effectively identical
|
# ? Jan 19, 2018 22:26 |
|
HoboMan posted:gently caress. what makes groovy so bad? it's the new hipste
|
# ? Jan 19, 2018 22:45 |
|
tef posted:i'm sure this used to be true if it's inside a global singleton dependency-injected class as a non-static, then it's not a global variable!
|
# ? Jan 19, 2018 23:00 |
|
you have a really weird axe to grind over dependency injection.
|
# ? Jan 19, 2018 23:03 |
|
NihilCredo posted:related: because for line in lines: line.split(=) is good enough to begin with quote:i give a pass to tools that are two decades old or more, but for anything modern, between xml, json, yaml and toml yaml has a 'safe_load' option because well, it's not safe to use. or interoperable json is well, painful for config formats because of the lack of things toml and yaml provide toml eh? well i guess if you want a holier than thou ini format well and xml? billion lols, namespaces, xsds. if i was writing a svg like format, i.e an interoperable document format, it would be a great choice. oh i guess there's protobufs that don't have real collections and uh thrift that does but um it seems flakey and why does finagle use a different version
|
# ? Jan 19, 2018 23:06 |
|
these can be valid complaints but if they follow up with "and that's why dockerfiles use their own special format" it's like bitching that the 2-michelin-star cook oversalted your salmon steak so you'd rather scavenge a dead trout from the riverbed after a drought
|
# ? Jan 19, 2018 23:17 |
|
i think i made a syntax error somewhere when i manually added several hundred lines to this 1600-line XML config file. no automated tool detects the problem, but the function that is supposed to parse it is silently failing after my additions i hate every single word of that paragraph.
|
# ? Jan 19, 2018 23:17 |
|
|
# ? Jun 1, 2024 20:21 |
|
how often are you editing configs that the syntax of json or xml is a significant concern?
|
# ? Jan 19, 2018 23:32 |