Conversation

Notices

1. Embed this notice
   Alex Gleason (alex@gleasonator.com)'s status on Thursday, 22-Sep-2022 06:47:04 JST Alex Gleason

   Overall I love TypeScript. But sometimes it’s so frustrating…

   No overload matches this call. Overload 1 of 2, '(queryKey: QueryKey, updater: Updater<InfiniteQueryObserverLoadingErrorResult<Record<{ account: EmbeddedEntity<Account | ReducerAccount>; id: string; unread: number; last_message: string | null; updated_at: Date; }> & Readonly<...>, unknown> | InfiniteQueryObserverLoadingResult<...> | InfiniteQueryObserverRefetchErrorResult<...> | InfiniteQueryObserverSuccessResult<...> | undefined, InfiniteQueryObserverLoadingErrorResult<...> | ... 3 more ... | undefined>, options?: SetDataOptions | undefined): [...][]', gave the following error. Argument of type '(chatPages: UseInfiniteQueryResult<Chat>) => { pages: any[]; pageParams: unknown[]; } | undefined' is not assignable to parameter of type 'Updater<InfiniteQueryObserverLoadingErrorResult<Record<{ account: EmbeddedEntity<Account | ReducerAccount>; id: string; unread: number; last_message: string | null; updated_at: Date; }> & Readonly<...>, unknown> | InfiniteQueryObserverLoadingResult<...> | InfiniteQueryObserverRefetchErrorResult<...> | InfiniteQueryO...'. Type '(chatPages: UseInfiniteQueryResult<Chat>) => { pages: any[]; pageParams: unknown[]; } | undefined' is not assignable to type 'DataUpdateFunction<InfiniteQueryObserverLoadingErrorResult<Record<{ account: EmbeddedEntity<Account | ReducerAccount>; id: string; unread: number; last_message: string | null; updated_at: Date; }> & Readonly<...>, unknown> | InfiniteQueryObserverLoadingResult<...> | InfiniteQueryObserverRefetchErrorResult<...> | Inf...'. Types of parameters 'chatPages' and 'input' are incompatible. Type 'InfiniteQueryObserverLoadingErrorResult<Record<{ account: EmbeddedEntity<Account | ReducerAccount>; id: string; unread: number; last_message: string | null; updated_at: Date; }> & Readonly<...>, unknown> | InfiniteQueryObserverLoadingResult<...> | InfiniteQueryObserverRefetchErrorResult<...> | InfiniteQueryObserverS...' is not assignable to type 'UseInfiniteQueryResult<Record<{ account: EmbeddedEntity<Account | ReducerAccount>; id: string; unread: number; last_message: string | null; updated_at: Date; }> & Readonly<...>, unknown>'. Type 'undefined' is not assignable to type 'UseInfiniteQueryResult<Record<{ account: EmbeddedEntity<Account | ReducerAccount>; id: string; unread: number; last_message: string | null; updated_at: Date; }> & Readonly<...>, unknown>'. Overload 2 of 2, '(filters: QueryFilters, updater: Updater<InfiniteQueryObserverLoadingErrorResult<Record<{ account: EmbeddedEntity<Account | ReducerAccount>; id: string; unread: number; last_message: string | null; updated_at: Date; }> & Readonly<...>, unknown> | InfiniteQueryObserverLoadingResult<...> | InfiniteQueryObserverRefetchErrorResult<...> | InfiniteQueryObserverSuccessResult<...> | undefined, InfiniteQueryObserverLoadingErrorResult<...> | ... 3 more ... | undefined>, options?: SetDataOptions | undefined): [...][]', gave the following error. Type 'string[]' has no properties in common with type 'QueryFilters'.ts(2769)
   • Embed this notice
     Alex Gleason (alex@gleasonator.com)'s status on Thursday, 22-Sep-2022 11:38:18 JST Alex Gleason

     in reply to

     • Josh Adams
     I expect a stack trace. I don't expect a query plan. It needs one of those websites where you can paste the TypeScript error in and it tells you in English what's going on.
     
     Anyway I fixed it, the two objects were indeed different. But it's like I Spy in that message!
   • Embed this notice
     Josh Adams (josh@social.rebased.network)'s status on Thursday, 22-Sep-2022 11:38:19 JST Josh Adams

     in reply to
     It doesn't look inscrutable, just like it's telling you a couple things it could have been but how far off each was
   • Embed this notice
     ew (e@masochi.st)'s status on Thursday, 22-Sep-2022 12:14:56 JST ew

     in reply to

     • NEETzsche
     • Josh Adams
     @NEETzsche @alex @josh productivity is better due to the tooling
     Alex Gleason likes this.
   • Embed this notice
     NEETzsche (neetzsche@iddqd.social)'s status on Thursday, 22-Sep-2022 12:14:57 JST NEETzsche

     in reply to

     • ew
     • Josh Adams
     @e @alex @josh As though JavaScript weren't soyware enough.
   • Embed this notice
     ew (e@masochi.st)'s status on Thursday, 22-Sep-2022 12:14:58 JST ew

     in reply to

     • NEETzsche
     • Josh Adams
     @NEETzsche @alex @josh bonk
   • Embed this notice
     NEETzsche (neetzsche@iddqd.social)'s status on Thursday, 22-Sep-2022 12:14:59 JST NEETzsche

     in reply to

     • Josh Adams
     @alex @josh typescript is a meme
   • Embed this notice
     ew (e@masochi.st)'s status on Thursday, 22-Sep-2022 12:15:00 JST ew

     in reply to

     • NEETzsche
     • Josh Adams
     @NEETzsche @alex @josh
     
     auto-generating your own admin interface
     https://adminjs.co/
     auto-generating your own api calls
     https://www.the-guild.dev/graphql/codegen
     auto-generating your own api layer
     https://www.graphile.org/postgraphile/
     auto-generating your own validation layers
     https://www.npmjs.com/package/typescript-json-schema
     auto-generating typed sql calls
     https://github.com/adelsz/pgtyped
     
     I can keep going lol
     Alex Gleason likes this.
   • Embed this notice
     NEETzsche (neetzsche@iddqd.social)'s status on Thursday, 22-Sep-2022 12:15:01 JST NEETzsche

     in reply to

     • ew
     • Josh Adams
     @e @alex @josh it really isn't lmfao
     Alex Gleason repeated this.
   • Embed this notice
     Alex Gleason (alex@gleasonator.com)'s status on Thursday, 22-Sep-2022 12:15:52 JST Alex Gleason

     in reply to

     • ew
     • NEETzsche
     • Josh Adams
     I want to see your repo of what you're building.
   • Embed this notice
     ew (e@masochi.st)'s status on Thursday, 22-Sep-2022 12:18:48 JST ew

     in reply to

     • NEETzsche
     • Josh Adams
     @alex @NEETzsche @josh I'll get a repo up this weekend
     Alex Gleason likes this.
   • Embed this notice
     Alex Gleason (alex@gleasonator.com)'s status on Thursday, 22-Sep-2022 12:20:21 JST Alex Gleason

     in reply to

     • ew
     • NEETzsche
     • Josh Adams
     Yeah that's after you're already running the code. TS tells you inside the text editor with hover tooltips. It really is pretty amazing.
   • Embed this notice
     NEETzsche (neetzsche@iddqd.social)'s status on Thursday, 22-Sep-2022 12:20:22 JST NEETzsche

     in reply to

     • ew
     • Josh Adams
     @e @alex @josh You know what else will bitch at you if you try to select a column that doesn't exist? The database.
   • Embed this notice
     NEETzsche (neetzsche@iddqd.social)'s status on Thursday, 22-Sep-2022 12:20:23 JST NEETzsche

     in reply to

     • ew
     • Josh Adams
     @e @alex @josh "my implementation of X uses/is built for typescript therefore typescript is a necessary productivity increaser"
     
     Should I start listing my favorite Ruby gems to prove how much more productive I am for using it
   • Embed this notice
     ew (e@masochi.st)'s status on Thursday, 22-Sep-2022 12:20:23 JST ew

     in reply to

     • NEETzsche
     • Josh Adams
     @NEETzsche @alex @josh due to typescript and graphql, if I code a graphql query that's like `{ users { name }` which is effectively the same as the sql query `select name from users` and the name column doesn't exist on my database my linter will bitch at me that the field is invalid.
     
     I basically get to write database calls on my frontend and get them validated like sql ones directly in a sql command line
   • Embed this notice
     Josh Adams (josh@social.rebased.network)'s status on Thursday, 22-Sep-2022 12:28:59 JST Josh Adams

     in reply to
     Look dude I want people writing elm you're preaching to the choir
     Alex Gleason likes this.
   • Embed this notice
     Caek Islove ? ❤️ (caekislove@gleasonator.com)'s status on Thursday, 22-Sep-2022 12:29:25 JST Caek Islove ? ❤️

     in reply to

     • NEETzsche
     • Josh Adams
     Exactly. Typescript is a way for Junior Developers to automatically elevate their bugs to Senior Developers who actually know JavaScript.
     Alex Gleason likes this.
   • Embed this notice
     Caek Islove ? ❤️ (caekislove@gleasonator.com)'s status on Thursday, 22-Sep-2022 12:33:05 JST Caek Islove ? ❤️

     in reply to
     Real men code in pure JavaScript! ?
   • Embed this notice
     Alex Gleason (alex@gleasonator.com)'s status on Thursday, 22-Sep-2022 12:33:05 JST Alex Gleason

     in reply to

     • Caek Islove ? ❤️
     Real men prefer to console.log 40 times instead of the text editor just telling them what's in that variable.
   • Embed this notice
     Alex Gleason (alex@gleasonator.com)'s status on Thursday, 22-Sep-2022 13:58:46 JST Alex Gleason

     in reply to

     • ew
     • NEETzsche
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     Yeah, but the developer experience is substantially better. You don't have to memorize so much anymore. That frees your brain to think about other things.
   • Embed this notice
     ew (e@masochi.st)'s status on Thursday, 22-Sep-2022 13:58:48 JST ew

     in reply to

     • NEETzsche
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @NEETzsche @roboneko @alex @josh you're deliberating ignoring the question of scale. it's pretty easy to do "type shit in, other people can see it" when you're dealing with a few hundred network requests
   • Embed this notice
     NEETzsche (neetzsche@iddqd.social)'s status on Thursday, 22-Sep-2022 13:58:48 JST NEETzsche

     in reply to

     • ew
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @e It isn't a question of scale. You and Alex just got done justifying your bloat on things like a GraphQL linter and a better IDE experience. It's about having six gorillion times as much system resources, but lacking the vision to do something with it other than make the task of writing software already written easier.
     
     It's bloat soyware. Straight up.
     
     @roboneko @alex @josh
   • Embed this notice
     NEETzsche (neetzsche@iddqd.social)'s status on Thursday, 22-Sep-2022 13:58:49 JST NEETzsche

     in reply to

     • ew
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @e Consider the following: the point that you're trying so desperately to dodge is that the entire program, all of it, the whole stack, the entire thing, needs to either do something a lot more significant than a variation on "type shit in, other people see it," or, and get this, it needs to confirm to 1990s standards of system resource consumption.
     
     If it doesn't?
     
     Anything in excess of that is definitionally bloat.
     
     @roboneko @alex @josh
   • Embed this notice
     ew (e@masochi.st)'s status on Thursday, 22-Sep-2022 13:58:52 JST ew

     in reply to

     • ew
     • NEETzsche
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @NEETzsche @alex @josh @roboneko err unused ram* my dyslexia kicking in
   • Embed this notice
     ew (e@masochi.st)'s status on Thursday, 22-Sep-2022 13:58:53 JST ew

     in reply to

     • NEETzsche
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @NEETzsche @roboneko @alex @josh unwasted ram is unwasted ram that just sits there doing nothing costing money
   • Embed this notice
     ew (e@masochi.st)'s status on Thursday, 22-Sep-2022 13:58:53 JST ew

     in reply to

     • ew
     • NEETzsche
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @NEETzsche @alex @josh @roboneko wasted ram*
   • Embed this notice
     NEETzsche (neetzsche@iddqd.social)'s status on Thursday, 22-Sep-2022 13:58:54 JST NEETzsche

     in reply to

     • ew
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @e We... literally just covered this.
     
     @roboneko @alex @josh
   • Embed this notice
     ew (e@masochi.st)'s status on Thursday, 22-Sep-2022 13:58:55 JST ew

     in reply to

     • NEETzsche
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @NEETzsche @roboneko @alex @josh all that ram usage on the server is taken up by the database indexes not the programming language
   • Embed this notice
     ew (e@masochi.st)'s status on Thursday, 22-Sep-2022 13:58:56 JST ew

     in reply to

     • NEETzsche
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @NEETzsche @roboneko @alex @josh we are not using 16gb to run a web browser tab
   • Embed this notice
     NEETzsche (neetzsche@iddqd.social)'s status on Thursday, 22-Sep-2022 13:58:56 JST NEETzsche

     in reply to

     • ew
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @e We are using much more than that to run the servers that process those messages.
     
     Something that's just as bad.
     
     
     Considering
     
     We
     
     Had
     
     Servers
     
     With
     
     Less
     
     Resources
     
     Than
     
     Our
     
     Phones
     
     Today
     
     By
     
     Multiple
     
     Orders
     
     Of
     
     Magnitude
     
     And
     
     They
     
     Worked
     
     Just
     
     Fine
     
     Are you... starting to comprehend the issue yet?
     
     @roboneko @alex @josh
   • Embed this notice
     NEETzsche (neetzsche@iddqd.social)'s status on Thursday, 22-Sep-2022 13:58:57 JST NEETzsche

     in reply to

     • ew
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @e Exactly. And yet here we are.
     
     @roboneko @alex @josh
   • Embed this notice
     ew (e@masochi.st)'s status on Thursday, 22-Sep-2022 13:58:58 JST ew

     in reply to

     • NEETzsche
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @NEETzsche @roboneko @alex @josh it doesn't take 16gb of ram to do that
   • Embed this notice
     ew (e@masochi.st)'s status on Thursday, 22-Sep-2022 13:58:59 JST ew

     in reply to

     • NEETzsche
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @NEETzsche @roboneko @alex @josh phones have 16gb of ram these days
   • Embed this notice
     NEETzsche (neetzsche@iddqd.social)'s status on Thursday, 22-Sep-2022 13:58:59 JST NEETzsche

     in reply to

     • ew
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @e k
     
     That's not an argument for using up 16 GB of RAM on a variation of "type shit in, other people see it."
     
     @roboneko @alex @josh
   • Embed this notice
     NEETzsche (neetzsche@iddqd.social)'s status on Thursday, 22-Sep-2022 13:59:00 JST NEETzsche

     in reply to

     • ew
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @e Not really. The yardstick was actually set in the 1990s. All system resource usage in excess of past implementations that worked is by definition bloat. The use case is "type your shit in, other people see it." We had that in the 1990s.
     
     We use up multiple orders of magnitude more system resources to accomplish this simple task today than we did then, and all we're doing is the same thing: typing shit in so others can see it.
     
     @roboneko @alex @josh
   • Embed this notice
     ew (e@masochi.st)'s status on Thursday, 22-Sep-2022 13:59:01 JST ew

     in reply to

     • NEETzsche
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @NEETzsche @roboneko @alex @josh >So you're saying that we need to start getting rid of databases so that we can reduce the system resource requirements for running a server that handles "type shit in, other people see it" back to what they were in the 1990s?
     ok you're just trolling me
   • Embed this notice
     ew (e@masochi.st)'s status on Thursday, 22-Sep-2022 13:59:02 JST ew

     in reply to

     • NEETzsche
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @NEETzsche @roboneko @alex @josh dude you have to be trolling. javascript is fast enough to do api calls. what's not fast is doing a bunch of relational table joins that you wouldn't have to do otherwise
   • Embed this notice
     NEETzsche (neetzsche@iddqd.social)'s status on Thursday, 22-Sep-2022 13:59:02 JST NEETzsche

     in reply to

     • ew
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @e "it's not the scripting language, it's the db" really is the weakest argument ever when it comes to this topic. So you're saying that we need to start getting rid of databases so that we can reduce the system resource requirements for running a server that handles "type shit in, other people see it" back to what they were in the 1990s? Is that your big contention? Because that's a really bad contention.
     
     Okay so I'll give you -3% on my final grade for failing to explicitly mention databases in your soy bloatware stack. Final grade: 97% or A+.
     
     In conclusion, these tools do not improve performance or productivity because the system requirements yardstick for "type shit in, other people see it" and variants was established in the 1990s and is significantly less than what you can get for a $5/mo VPS.
     
     @roboneko @alex @josh
   • Embed this notice
     ew (e@masochi.st)'s status on Thursday, 22-Sep-2022 13:59:04 JST ew

     in reply to

     • NEETzsche
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @NEETzsche @roboneko @alex @josh if you want to talk about protocols, graphql is leaner and what I've been trying to tell you but you dismiss it even if it saves a ton of db calls
   • Embed this notice
     NEETzsche (neetzsche@iddqd.social)'s status on Thursday, 22-Sep-2022 13:59:04 JST NEETzsche

     in reply to

     • ew
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @e >my bloatware is using up a ton of system resources to do something that was done on a Pentium II no problem
     >maybe if I add a little soyware to my bloatware it'll make it run slightly faster...
     
     @roboneko @alex @josh
   • Embed this notice
     NEETzsche (neetzsche@iddqd.social)'s status on Thursday, 22-Sep-2022 13:59:05 JST NEETzsche

     in reply to

     • ew
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @e "or"
     
     This got addressed before you responded.
     
     @roboneko @alex @josh
   • Embed this notice
     NEETzsche (neetzsche@iddqd.social)'s status on Thursday, 22-Sep-2022 13:59:06 JST NEETzsche

     in reply to

     • ew
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @e This... doesn't even approach refuting my point.
     
     @roboneko @alex @josh
   • Embed this notice
     ew (e@masochi.st)'s status on Thursday, 22-Sep-2022 13:59:06 JST ew

     in reply to

     • NEETzsche
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @NEETzsche @roboneko @alex @josh you brought up 1kb payloads like they don't matter. they do matter, it has nothing to do with string parsing performance, it has to do with the db calls to generate that 1kb payload.
   • Embed this notice
     ew (e@masochi.st)'s status on Thursday, 22-Sep-2022 13:59:07 JST ew

     in reply to

     • NEETzsche
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @NEETzsche @roboneko @alex @josh it's more than likely it's the db calls that prevent scaling
   • Embed this notice
     ew (e@masochi.st)'s status on Thursday, 22-Sep-2022 13:59:08 JST ew

     in reply to

     • NEETzsche
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @NEETzsche @roboneko @alex @josh >under 1kb
     that's not the issue. the issue is the unnecessary db calls to generate that response data
   • Embed this notice
     NEETzsche (neetzsche@iddqd.social)'s status on Thursday, 22-Sep-2022 13:59:08 JST NEETzsche

     in reply to

     • ew
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @e "or the soyware scripting language your server runs requires a supercomputer to do string manipulation"
     
     @roboneko @alex @josh
   • Embed this notice
     NEETzsche (neetzsche@iddqd.social)'s status on Thursday, 22-Sep-2022 13:59:09 JST NEETzsche

     in reply to

     • ew
     • Neko McCatface v2023 :verified::makemeneko:
     • Josh Adams
     @roboneko Except unironically. This message content I'm sending you is under 1KB so if it takes more than 1s to transfer to you on my 14400 baud modern it's because your protocol is bloated or the soyware scripting language your server runs requires a supercomputer to do string manipulation.
     
     @alex @e @josh
   • Embed this notice
     ew (e@masochi.st)'s status on Thursday, 22-Sep-2022 13:59:10 JST ew

     in reply to

     • NEETzsche
     • Josh Adams
     @NEETzsche @alex @josh ok well enjoy hunting down bugs. if I change a datatype in my code I don't have to go hunt down every changed instance my linter will just error
   • Embed this notice
     Neko McCatface v2023 :verified::makemeneko: (roboneko@bae.st)'s status on Thursday, 22-Sep-2022 13:59:10 JST Neko McCatface v2023 :verified::makemeneko:

     in reply to

     • ew
     • NEETzsche
     • Josh Adams
     @e @alex @NEETzsche @josh no man u don't get it should do everything in a combination of C89 and forth the way God intended. if you haven't rolled your own https implementation in assembler then you really need to git gud
   • Embed this notice
     NEETzsche (neetzsche@iddqd.social)'s status on Thursday, 22-Sep-2022 13:59:11 JST NEETzsche

     in reply to

     • ew
     • Josh Adams

     @e Because I don’t need to. There are two ways to argue for “technologies” that actually matter:

     • Show me something it can do that previous languages/frameworks can’t (rearranging “patterns” and data/object structures doesn’t count)
     • Improve performance in terms of CPU usage and memory footprint

     Basically everything else is bullshit

     @alex @josh

   • Embed this notice
     ew (e@masochi.st)'s status on Thursday, 22-Sep-2022 13:59:12 JST ew

     in reply to

     • NEETzsche
     • Josh Adams
     • Marine ?
     @NEETzsche @alex @marine @josh >a problem that has been thoroughly solved for two decades at a minimum.
     you have a lot of opinions but you're not even willing to look into this stuff
     
     https://relay.dev/docs/principles-and-architecture/thinking-in-graphql/
     https://relay.dev/docs/principles-and-architecture/thinking-in-relay/
   • Embed this notice
     ew (e@masochi.st)'s status on Thursday, 22-Sep-2022 13:59:13 JST ew

     in reply to

     • NEETzsche
     • Josh Adams
     @NEETzsche @alex @josh actually I linked to two of those that are typescript based but I thought we were also talking about javascript
   • Embed this notice
     NEETzsche (neetzsche@iddqd.social)'s status on Thursday, 22-Sep-2022 13:59:13 JST NEETzsche

     in reply to

     • ew
     • Josh Adams
     • Marine ?
     @e @alex @josh My point is that all of this soyware that purports to increase productivity has a rather flimsy argument for it actually being the case. Especially since you can make the same arguments with things that are no longer the new hotness and haven't been in years, like Rails. It's just a variation on the framework treadmill from a decade ago, except this time it's programmers trying to get ever more abstracted and removed from the metal without really implementing anything new.
     
     I just got out of a conversation with @marine about an idea of hers that actually qualifies as new and none of these "technologies" are going to help her with it because all these "technologies" do is permute on CRUD, a problem that has been thoroughly solved for two decades at a minimum.
   • Embed this notice
     NEETzsche (neetzsche@iddqd.social)'s status on Thursday, 22-Sep-2022 13:59:14 JST NEETzsche

     in reply to

     • ew
     • Josh Adams
     @e @alex @josh none of that requires typescript lol