Carbon dioxide posted:Uh, just learn Japanese guys. It's not that complicated for 99% of usages. Yeah occasionally you'll need to do something complicated, but the idea that so many of you don't understand how to read/write Japanese is surprising to me, I thought that was a pretty rudimentary part of being a weeaboo. これ、でも風刺じゃありません VikingofRock fucked around with this message at 00:17 on Nov 16, 2017 |
|
# ? Nov 15, 2017 23:08 |
|
|
# ? May 25, 2024 15:29 |
|
Tomie knows me posted:Now you have 2 problems. code:
PhantomOfTheCopier fucked around with this message at 00:47 on Nov 16, 2017 |
# ? Nov 16, 2017 00:43 |
|
Dylan16807 posted:Writing good SQL is often more an exercise in working around a really cranky JIT than code quality. Two queries that are fundamentally stating the same actions can have massive performance differences based in picky little things the query planner recognizes. Or just what the DB optimizes for. I've fixed n+1 queries on MySQL and had performance drop as a result since MySQL decided to go ahead and optimize around an anti-pattern.
|
# ? Nov 16, 2017 00:57 |
|
MisterZimbu posted:In my experience the DBA is either a goddamn genius or an incompetent buffoon who should have all their computers taken away. There's no in-between. I can attest that when I finally tried to bone up on relational algebra and join algorithms, I immediately decided that there was no way in hell I was ever going to be good enough to make good databases.
|
# ? Nov 16, 2017 01:58 |
|
Dylan16807 posted:Writing good SQL is often more an exercise in working around a really cranky JIT than code quality. Two queries that are fundamentally stating the same actions can have massive performance differences based in picky little things the query planner recognizes. Look up this thread. In no other language would we accept "but the compiler generated bytecode that did what I want" as justification for otherwise terrible code. Rubellavator posted:I once got really annoyed trying to figure out why my query wasn't using some index to join two tables and that was when I discovered that the optimizer actually uses different plans based on table statistics. For example, the table I was joining on only had 3 rows of test data so the optimizer ignores the index. Isn't that exactly what you want though? It's like using an array rather than a hash set for a three element collection because if x in y is still faster to linearly search through an array than to compute the hash function even if asymptotically constant time is faster than linear.
|
# ? Nov 16, 2017 05:15 |
|
KernelSlanders posted:Isn't that exactly what you want though? It's like using an array rather than a hash set for a three element collection because if x in y is still faster to linearly search through an array than to compute the hash function even if asymptotically constant time is faster than linear. Yeah it is, he's just saying he didn't see it coming.
|
# ? Nov 16, 2017 05:19 |
|
Carbon dioxide posted:Uh, just learn Japanese guys. It's not that complicated for 99% of usages. Yeah occasionally you'll need to do something complicated, but the idea that so many of you don't understand how to read/write Japanese is surprising to me, I thought that was a pretty rudimentary part of being a weeaboo. If you run an Import/Export business that specializes in stuff from Japan, that's good advice, yes. If 1/4 of your job is talking to Japanese people who don't know English, it's still good advice if you're at all interested in doing your job better.
|
# ? Nov 16, 2017 15:54 |
|
Munkeymon posted:If 1/4 of your job is talking to Japanese people who don't know English, it's still good advice if you're at all interested in doing your job better. What if your company has a translator and you can just give him the plain English version and he will translate it for the Japanese people and 99.9% of the time this is fine but just occasionally you need the translations urgently and so have to write it out yourself? :itisanORManalogy:
|
# ? Nov 16, 2017 16:01 |
|
NtotheTC posted:What if your company has a translator and you can just give him the plain English version and he will translate it for the Japanese people and 99.9% of the time this is fine but just occasionally you need the translations urgently and so have to write it out yourself? Congratulations on your good fortune because translators that don't accidentally offend people constantly are hard to find.
|
# ? Nov 16, 2017 16:11 |
|
NtotheTC posted:What if your company has a translator and you can just give him the plain English version and he will translate it for the Japanese people and 99.9% of the time this is fine but just occasionally you need the translations urgently and so have to write it out yourself? Unfortunately then you’ll need to start over and rewrite the book he was translating from scratch when he starts taking months to translate when he translated in a few hours before. :itsanORManalogy:
|
# ? Nov 16, 2017 18:06 |
|
If it's a translator you know and work with a lot you know when to have them do the translation or if you need to do it yourself or if you need to get someone else to do it.
|
# ? Nov 16, 2017 18:12 |
|
Hughlander posted:Unfortunately then you’ll need to start over and rewrite the book he was translating from scratch when he starts taking months to translate when he translated in a few hours before. Wait is this really a thing? To be fair I've only worked with one translator before and ok gently caress the analogy is dumb as hell. But yeah I work with the django ORM and it's pretty much solid for day to day use, on the rare occasions where I do need to do a certain query in raw SQL it makes it pretty easy to drop down for that one query and you don't have to throw out the entire ORM...
|
# ? Nov 16, 2017 21:52 |
|
https://twitter.com/jetserge/status/852536709757710337 Oh my sweet Jesus
|
# ? Nov 16, 2017 22:44 |
|
Zaphod42 posted:https://twitter.com/jetserge/status/852536709757710337 That's uh... Jarring.
|
# ? Nov 16, 2017 22:49 |
|
iospace posted:That's uh...
|
# ? Nov 16, 2017 23:18 |
|
Zaphod42 posted:https://twitter.com/jetserge/status/852536709757710337 ...how the hell does that even happen? Including a metric poo poo ton of media in the jar file? Which is undoubtedly in their VC as well...
|
# ? Nov 17, 2017 04:34 |
|
The Phlegmatist posted:Which is undoubtedly in their VC as well... Hahahaha that was my next thought
|
# ? Nov 17, 2017 04:44 |
|
At that point, I'd assume the jars themselves are checked in.
|
# ? Nov 17, 2017 05:20 |
|
Can we agree that the DSLs that are just SQL are the real coding horror?code:
code:
|
# ? Nov 17, 2017 10:12 |
|
NtotheTC posted:Wait is this really a thing? To be fair I've only worked with one translator before and ok gently caress the analogy is dumb as hell. SQL is no different from any technology a dev comes into contact with but doesn't necessarily directly develop in - if you don't at least take an interest in understanding it, you are bad. Horse Clocks posted:Can we agree that the DSLs that are just SQL are the real coding horror?
|
# ? Nov 17, 2017 11:43 |
|
Query builders reduce the temptation to build SQL queries dynamically using string operations.
|
# ? Nov 17, 2017 12:03 |
|
The Phlegmatist posted:...how the hell does that even happen? Including a metric poo poo ton of media in the jar file? Which is undoubtedly in their VC as well... It happen one jar at a time my friend, one jar at a time.
|
# ? Nov 17, 2017 13:31 |
|
Doom Mathematic posted:Query builders reduce the temptation to build SQL queries dynamically using string operations. This. I enjoy writing SQL, but I'll put up with a lot if it gets me compile-time syntactic and semantic checking.
|
# ? Nov 17, 2017 13:37 |
|
NtotheTC posted:Wait is this really a thing? To be fair I've only worked with one translator before and ok gently caress the analogy is dumb as hell. No I was being factious with the normal reasoning as to why you want to avoid an ORM to begin with. And honestly in a fashion I never experienced. The only time I would chose an ORM is purely to save the speed of development time on project that isn't core to what the team is doing. IE: One where performance characteristics of it will never matter. But then again most of my usages of databases or the past decade or so has either been massive de-noramlized NoSQL stores or tooling around ETL/BVT systems. For the former I don't know of an ORM that could handle it, for the later, I've used multiple ORMs.
|
# ? Nov 17, 2017 15:29 |
|
Horse Clocks posted:Can we agree that the DSLs that are just SQL are the real coding horror? If the later is typesafe it's probably an improvements, but to the extent that code:
|
# ? Nov 17, 2017 15:55 |
|
Hughlander posted:No I was being factious with the normal reasoning as to why you want to avoid an ORM to begin with. And honestly in a fashion I never experienced. The only time I would chose an ORM is purely to save the speed of development time on project that isn't core to what the team is doing. IE: One where performance characteristics of it will never matter. This is why it's good to know SQL to some extent AND to know your ORM. I use the Django ORM extensively because it's so easy, but Django makes it easy to fall back to SQL when it makes sense...and they encourage you to do so. In doing so you can get the speed of development benefits along with the speed of execution benefits. To me the most useful part of an ORM isn't to avoid writing SQL, it's all the stuff around that like mapping to objects and managing life cycles.
|
# ? Nov 17, 2017 16:21 |
|
Doom Mathematic posted:Query builders reduce the temptation to build SQL queries dynamically using string operations. so do prepared statements, and without having to translate your sql into some other framework's "native syntax" that almost does what you want but not quite
|
# ? Nov 17, 2017 22:19 |
DELETE CASCADE posted:so do prepared statements, and without having to translate your sql into some other framework's "native syntax" that almost does what you want but not quite How would you write a prepared statement that can check for any combination of 20 different search parameters, each of which may have various types of criteria configured (e.g. prefix match vs. present in a given set)?
|
|
# ? Nov 18, 2017 09:21 |
|
nielsm posted:How would you write a prepared statement that can check for any combination of 20 different search parameters, each of which may have various types of criteria configured (e.g. prefix match vs. present in a given set)? Theoretical answer: SQL is Turing compete so there's guaranteed to be a way. Practical answer: optional search parameters can be represented as (@searchParam IS NULL OR [column] =@searchParam). Multiple possible patterns, same, just use OR and make the unused patterns false through additional parameters.
|
# ? Nov 18, 2017 14:11 |
|
nielsm posted:How would you write a prepared statement that can check for any combination of 20 different search parameters, each of which may have various types of criteria configured (e.g. prefix match vs. present in a given set)? This looks like something you hire a lawyer for. Edit: https://twitter.com/fasterthanlime/status/931452702458081280 Absurd Alhazred fucked around with this message at 17:39 on Nov 18, 2017 |
# ? Nov 18, 2017 15:54 |
|
So much wrong in single tweet
|
# ? Nov 18, 2017 18:36 |
|
I once heard someone say that the entire history of C++ is the history of an attempt to get a string type to work.
|
# ? Nov 18, 2017 19:06 |
|
lifg posted:I once heard someone say that the entire history of C++ is the history of an attempt to get a string type to work. http://en.cppreference.com/w/cpp/string/basic_string http://icu-project.org/apiref/icu4c/classicu_1_1UnicodeString.html http://doc.qt.io/qt-5/qstring.html http://static.cegui.org.uk/docs/0.8.7/classCEGUI_1_1String.html https://docs.unrealengine.com/latest/INT/Programming/UnrealArchitecture/StringHandling/FString/ ... The search continues.
|
# ? Nov 18, 2017 19:59 |
|
In a project at work, I have three string types (and conversions between the three) in the same Objective C++ source file: NSString, std::string and a counted string type from a third party library, plus of course std::string_view
|
# ? Nov 18, 2017 23:42 |
|
the russians used a pascal
|
# ? Nov 19, 2017 01:15 |
|
nielsm posted:How would you write a prepared statement that can check for any combination of 20 different search parameters, each of which may have various types of criteria configured (e.g. prefix match vs. present in a given set)? As a user, I would take myself elsewhere. "You mean I have to fill out all this Every Time? But I just want to X!" Good products, and, by introspection, good programs and databases are built with the proper balance of well-placed optimizations. Rarely is the proper decision to do it all in the database. Rarely is the proper decision to do it all in the program.
|
# ? Nov 19, 2017 15:07 |
|
Soricidus posted:the russians used a pascal
|
# ? Nov 19, 2017 17:05 |
|
Soricidus posted:the russians used a pascal http://wiki.freepascal.org/Character_and_string_types
|
# ? Nov 19, 2017 18:26 |
|
Soricidus posted:the russians used a pascal
|
# ? Nov 19, 2017 20:08 |
|
|
# ? May 25, 2024 15:29 |
|
PhantomOfTheCopier posted:wtf are you even doing? What the spec asked for quote:As a user, I would take myself elsewhere. "You mean I have to fill out all this Every Time? But I just want to X!" Part of the spec is to remember the last search problem solved
|
# ? Nov 20, 2017 17:10 |