|
Luigi Thirty posted:oh my god my new goal is to get a Spectre GCR I know a couple people with the Spectrum hardware, people without the GCR model had to do weird things to talk Macs into formatting disks as 720KB MFM with HFS and copy software onto them of course the cracks for the Spectrum software that let you use it with ROM images were very popular as well there were also a couple equivalents for Amiga, but I think the Amiga drives could at least read and write Mac disks
|
#
?
Oct 6, 2016 07:55
|
|
|
|
| # ? May 11, 2024 14:27 |
|
|
tef posted:then you get to "EDSLS" unless you're talking about implementing DSLs with macros in Common Lisp or Scheme of course
|
|
#
?
Oct 6, 2016 07:58
|
|
|
gradle would've been fine it it was just maven with nicer syntax instead, it is: • a convoluted dsl based on a language nobody uses or understands • a great way to end up with a build system that's more complex than your business logic • hot garbage it's really bad java optionals are too little too late. they don't do much to help because the main source of problematic nulls is other people's code, which doesn't use optionals. they also manage to have an even more cumbersome syntax than @Nonnull which is quite a feat!
|
|
#
?
Oct 6, 2016 08:50
|
|
|
HoboMan posted:lol same with the strings You can specify the parameter direction in c# as .output and then in the proc itself have the matching parameter as an output (@fart bool OUTPUT in the proc declarations), when you execute the value of that var in the proc will be returned to the c# object it is way neater than using a data table for one value and it also means you can reuse the proc in other procs as you can pass variables to and from procs in sql but can't easily take a table output from a proc as input into another proc you do have to cast the parameter.value explicitly to its type in c# though as sql doesn't return a defined type automatically because it's stupid edit: you can actually make a parameter an input and an output, i guess to confuse the poo poo out of everyone that has to work out why @Butt both returns and accepts values Powerful Two-Hander fucked around with this message at 09:40 on Oct 6, 2016 |
|
#
?
Oct 6, 2016 09:03
|
|
|
@ExitOnly Butt
|
|
#
?
Oct 6, 2016 10:05
|
|
|
HoboMan posted:can someone recommend a book on SQL development? like optimizing queries, best practices, how to think about reads and writes? everything i've seen looks like it's for someone who doesn't even know what an RDBMS is. (i am working on SQL server 2014) if you wanna huff database performance farts, check out this site which goes into pretty deep detail on db indexes. it's database-agnostic in that it covers all of the major dbs, and points out implementation quirks/weirdnesses and terminology differences. also available in book form as sql server in particular gets your rocks off, you'll at some point need to dig into some execution plans. pick up this and read through it over a few days/weeks/months. note that execution plans are obtuse as gently caress and difficult to figure out, especially if you only need to look at them occasionally red gate actually publish a buttload of sql server (and related) ebooks, so you could have a browse and see if anything interests you there SET STATISTICS IO, TIME ON and execution plans are your sql server perfect troubleshooting bread and butter special mention goes to sql sentry plan explorer, which is a much nicer and more powerful tool for reading/interpreting execution plans general advice: watch out when writing functions in join conditions/WHERE clauses - this can easily turn a quick index seek into a full table scan, and can happen easily if you're not looking. an annoying variant on this is the 'surprise! your column data types don't match and there's now an implicit conversion running on every row in the table!' also try to simplify your queries, by rewriting or by breaking an operation up into multiple queries using temp tables (WITH clauses and views don't count!). simpler queries are easier for the query optimizer to work on and are easier to get good statistics on, meaning the optimizer can check more potential plans with more appropriate row estimates, possibly giving a quicker plan. i recently had a statement with an OR condition that was performing a table scan. by breaking it up on the OR into two separate statements joined by a UNION, i was able to get the optimizer to do a couple of seeks on some sensible indexes - the alternative would have been to introduce a really gross, highly specific, specialized index to the base table, which i wasn't going to do because that would be stupid i like databases 
|
|
#
?
Oct 6, 2016 11:31
|
|
|
idk if you do firefighting or on call or whatever, but if you do you will wanna consider learning this stuff on the clock - it's not something you're likely to pick up easily when a client is on the phone complaining about their slow site in this situation you'll want to know how to quickly figure out if any indexes have blown up and how to rebuild them, because indexes are like 90% of the solution to a slow db. again, maybe practice this at work sometime?
|
|
#
?
Oct 6, 2016 11:39
|
|
|
https://medium.com/friendship-dot-js/i-peeked-into-my-node-modules-directory-and-you-wont-believe-what-happened-next-b89f63d21558#.uw7554xnh has this been posted yet
|
|
#
?
Oct 6, 2016 14:08
|
|
|
I love poes law
|
|
#
?
Oct 6, 2016 14:13
|
|
|
The Pantone bloggers again disrespecting the poor rouge node.js developers.
|
|
#
?
Oct 6, 2016 14:50
|
|
|
holy gently caress two stupid things today: 1. a major system that process hundreds of thousands of records per day has no keys on any db tables and every column is varchar because the devs "didn't want to think about what type something should be". I actually shouted 'are you loving kidding me' out loud 2. i sent the new dev off to rewrite the file download component for an app, two weeks later they've come back with a half finished upload component that duplicates existing functionality wtf
|
|
#
?
Oct 6, 2016 14:59
|
|
|
Sapozhnik posted:yeah that's cool that doesn't help when your function takes a parameter and it should not ever be null.
|
|
#
?
Oct 6, 2016 14:59
|
|
|
Lutha Mahtin posted:https://medium.com/friendship-dot-js/i-peeked-into-my-node-modules-directory-and-you-wont-believe-what-happened-next-b89f63d21558#.uw7554xnh jfc just burn it all down at this point
|
|
#
?
Oct 6, 2016 15:01
|
|
|
Lutha Mahtin posted:https://medium.com/friendship-dot-js/i-peeked-into-my-node-modules-directory-and-you-wont-believe-what-happened-next-b89f63d21558#.uw7554xnh holy lmao
|
|
#
?
Oct 6, 2016 15:01
|
|
|
Lutha Mahtin posted:https://medium.com/friendship-dot-js/i-peeked-into-my-node-modules-directory-and-you-wont-believe-what-happened-next-b89f63d21558#.uw7554xnh https://i.imgur.com/3R4d0cZ.png 
|
|
#
?
Oct 6, 2016 15:07
|
|
|
eschaton posted:I wanna hear the shaggarpinion on coderush ive never used it
|
|
#
?
Oct 6, 2016 15:07
|
|
|
im the guy fieri image in a text document
|
|
#
?
Oct 6, 2016 15:13
|
|
|
ok we have a table that has a terminal id a user id and a login timestamp. i need to get all the times someone logged in more than once on the same day on the same terminal. i have some garbage that i tossed together, but it's got dupes and poo poo in it and i just generally can't conceptualize this in sql properly help?
|
|
#
?
Oct 6, 2016 15:14
|
|
|
Lutha Mahtin posted:https://medium.com/friendship-dot-js/i-peeked-into-my-node-modules-directory-and-you-wont-believe-what-happened-next-b89f63d21558#.uw7554xnh javascript is the worst and web "development" is the worst and Microsoft was trying to protect us from this for years. Then a bunch of retards decided to write failfox and chome and led us to this horrible state of affairs.
|
|
#
?
Oct 6, 2016 15:15
|
|
|
if you ever got mad at IE not doing something its litterrallly your fault that web 'development" is such poo poo
|
|
#
?
Oct 6, 2016 15:16
|
|
|
Finster Dexter posted:Does MUMPS count as a DSL? i guess if you take the "domain specific" part literally, yes, cuz it was designed basically for healthcare databases and it's really good at that and nothing else. but its a proper full-blown language and not a lang-on-top like the stuff you typically call a "dsl".
|
|
#
?
Oct 6, 2016 15:16
|
|
|
i wanted to rant at length a bit more about DSLs specifically with respect to hardware design languages the incumbent HDLs (VHDL, verilog, systemverilog) are widely accepted as absolute trash. they are uselessly verbose and full of all sorts of exciting gotchas, non-features, and sharp corners. this is not particularly surprising, as they are old languages designed by hardware engineers. it's also worth noting that they were originally defined as hardware description languages, rather than synthesis languages. as a result, huge portions (maybe the majority of?) each language is strictly for simulation - if you use more than like 10% of the language, your synthesis tool will fail, sometimes silently. further, even in that remaining 10% of the language, you're pretty much stuck to predetermined patterns of writing code, because anything even remotely unusual will probably confuse the synthesis tool, leading to wildly unexpected garbage. the always-brilliant dan luu wrote a bunch more about verilog weirdness here: http://danluu.com/why-hardware-development-is-hard/ so our languages are bad, and this is widely recognized. enter many people trying to solve that problem. unfortunately, inexplicably, (nearly?) everybody chooses to implement their new HDL as a DSL in something else. such nightmares include: * tools that synthesize a subset of C++ (vivado HLS, among others) * a python DSL (MyHDL) * a scala DSL (Chisel) * a million haskell DSLs (Lava, Hydra, CLaSH) the common thread here is the idea of "what if we bolt on synthesis features to an existing language" which is literally the root problem with verilog and vhdl. here is a language, now only use an arbitrary subset of it. now you also get the added benefit of more translation layers between your design and your horrorshow synthesis results! (because of the exciting way in which fpga development works, all of these languages have to compile to some minimal set of verilog or vhdl to actually be synthesized, because synthesis is extremely closed-source and terrible. vivado HLS may synthesize to some proprietary pile of rear end, because it is a first-party tool) there are many other aborted hdls (like JHDL, a java synthesis language!), but they are all abandonware if you're ever wondering what computer programming would've looked like without open source, just look at hardware development
|
|
#
?
Oct 6, 2016 15:35
|
|
|
probably the most common hdl other than verilog or vhdl is "pile of ad-hoc perl scripts that spew verilog/vhdl"
|
|
#
?
Oct 6, 2016 15:36
|
|
|
Powerful Two-Hander posted:holy gently caress two stupid things today: So a MongoDB developer?
|
|
#
?
Oct 6, 2016 15:38
|
|
|
Bloody posted:probably the most common hdl other than verilog or vhdl is "pile of ad-hoc perl scripts that spew verilog/vhdl" Itym tcl scripts
|
|
#
?
Oct 6, 2016 15:40
|
|
|
MrMoo posted:So a MongoDB developer? They should add a chapter to SQL Antipatterns about MongoDB. Chapter 83: MongoDB
|
|
#
?
Oct 6, 2016 15:43
|
|
|
Bloody posted:probably the most common hdl other than verilog or vhdl is "pile of ad-hoc perl scripts that spew verilog/vhdl" i did a lot of vhdl stuff in college and the language was fine but the tooling was terrible.
|
|
#
?
Oct 6, 2016 15:44
|
|
|
Lutha Mahtin posted:https://medium.com/friendship-dot-js/i-peeked-into-my-node-modules-directory-and-you-wont-believe-what-happened-next-b89f63d21558#.uw7554xnh is this real or a joke i can't tell
|
|
#
?
Oct 6, 2016 15:46
|
|
|
CRIP EATIN BREAD posted:im the guy fieri image in a text document I dunno it's not that farfetched 
|
|
#
?
Oct 6, 2016 15:47
|
|
|
Mr SuperAwesome posted:is this real or a joke i can't tell It's somebody's lame rear end joke
|
|
#
?
Oct 6, 2016 15:49
|
|
|
it's a joke you lunatics.
|
|
#
?
Oct 6, 2016 15:49
|
|
|
https://github.com/search?q=http.re...&utf8=%E2%9C%93 only result is a russian(?) translation of op
|
|
#
?
Oct 6, 2016 15:50
|
|
|
re ide and gitchat just use git at the command line and vim, its p great
|
|
#
?
Oct 6, 2016 15:50
|
|
|
Mr SuperAwesome posted:re ide and gitchat just use emacs and magit, its p great
|
|
#
?
Oct 6, 2016 15:57
|
|
|
Mr SuperAwesome posted:re ide and gitchat just use git at the command line and vim, its p great anything else is just wrong
|
|
#
?
Oct 6, 2016 16:02
|
|
|
Mr SuperAwesome posted:re ide and gitchat just use git at the command line and vim, its p great
|
|
#
?
Oct 6, 2016 16:06
|
|
|
Mr SuperAwesome posted:re ide and gitchat just use git at the command line and vim, its p great lol
|
|
#
?
Oct 6, 2016 16:07
|
|
|
Lutha Mahtin posted:https://medium.com/friendship-dot-js/i-peeked-into-my-node-modules-directory-and-you-wont-believe-what-happened-next-b89f63d21558#.uw7554xnh https://twitter.com/steveklabnik/status/783321843793010688
|
|
#
?
Oct 6, 2016 16:11
|
|
|
tef posted:just one thing: don't use auto increment, use uuids yeah we got guids for most primary keys (except the tings our last dba built out is all autoincrementing ints since they hated guids because they are annoying to type out) Finster Dexter posted:It's worth noting that you if you're using GUID primary keys you should still have an autoincrement clustering key to avoid bonkers fragmentation. huh?
|
|
#
?
Oct 6, 2016 16:12
|
|
|
|
| # ? May 11, 2024 14:27 |
|
|
Mr SuperAwesome posted:re ide and gitchat just use git at the command line and vim, its p great
|
|
#
?
Oct 6, 2016 16:13
|
|