| 00:00:03 | FromDiscord | <Invisible> wtf |
| 00:00:19 | FromDiscord | <graveflo> what? |
| 00:01:50 | * | derpydoo joined #nim |
| 00:04:11 | FromDiscord | <Invisible> not all errors are catchable? |
| 00:05:30 | FromDiscord | <graveflo> again errors are not exceptions. When you are programming in a lower level language like nim you have to consider how things are implemented more so then a high level language. Any error handling / exceptions system needs to be implemented somewhere. They can be subverted or ignored its just how it is. |
| 00:06:08 | FromDiscord | <graveflo> You dont have to think too much about it if you dont want to but something like this might have more complexity then you would think |
| 00:08:47 | FromDiscord | <Elegantbeef> Think about nil dereferencing |
| 00:09:16 | FromDiscord | <graveflo> are you trying to segfault my brain? |
| 00:09:29 | FromDiscord | <Elegantbeef> That's a hard quit, it's just wrong to allow recovery from that 😄 |
| 00:17:36 | FromDiscord | <Invisible> Any way I can prevent 2 of the same application from running at the same time? |
| 00:18:28 | FromDiscord | <graveflo> you mean recognize if the app you are making is already running at start up? |
| 00:18:37 | FromDiscord | <Invisible> In reply to @graveflo "you mean recognize if": No |
| 00:18:42 | FromDiscord | <Invisible> I mean like if the user spam opens it |
| 00:18:45 | FromDiscord | <Invisible> it wont open 60 windows |
| 00:19:09 | FromDiscord | <graveflo> yea so the program must recognize that it is already open and exit itself in that case, right? |
| 00:19:15 | FromDiscord | <Invisible> yup |
| 00:19:23 | FromDiscord | <Invisible> thought u meant pc startup |
| 00:19:36 | FromDiscord | <graveflo> okay there are several ways to do this but there are trade offs |
| 00:19:42 | FromDiscord | <Invisible> such as? |
| 00:20:31 | FromDiscord | <graveflo> the dumbest way to do it is to use a signal file. its stupid easy but fragile. If you just have the program check if a file exists and if it does then exit but if it doesnt then it starts and creates the file you will kinda have this behavior |
| 00:20:47 | FromDiscord | <graveflo> Im only mentioning this bc it seems you are trying to do things the easiest way possible |
| 00:21:00 | FromDiscord | <Invisible> but if it exits, it won't delete the file |
| 00:21:07 | FromDiscord | <Invisible> such as if the pc abruptly loses power or something |
| 00:21:14 | FromDiscord | <graveflo> the real way to do it is to use IPC or you can query the OS for information about the processes currently running |
| 00:21:16 | FromDiscord | <Elegantbeef> Right that's the issue with lock files |
| 00:21:24 | FromDiscord | <Elegantbeef> for reference the Unity engine uses a lock file |
| 00:21:46 | FromDiscord | <Elegantbeef> you can just search by process name for if your process exists |
| 00:21:48 | FromDiscord | <Invisible> I have na idea |
| 00:22:00 | FromDiscord | <Invisible> What if I try to edit the file and add 1 byte to it, and if it doesn't let me then it's running |
| 00:22:03 | FromDiscord | <Invisible> but if it does then it's not |
| 00:22:07 | FromDiscord | <Invisible> nvm |
| 00:22:11 | FromDiscord | <Invisible> it cant edit itself |
| 00:22:12 | FromDiscord | <Invisible> im dumb |
| 00:22:14 | FromDiscord | <graveflo> no thats the lock file method |
| 00:22:25 | FromDiscord | <graveflo> you have to keep it "open" at all times |
| 00:22:37 | FromDiscord | <graveflo> but that will work and the OS usually will free the lock on its own |
| 00:22:43 | FromDiscord | <Elegantbeef> It's just a file with a read/write lock |
| 00:22:46 | FromDiscord | <graveflo> in some bad events that is |
| 00:23:02 | FromDiscord | <Invisible> As long as it works and I can edit it easily and not shit myself trying to understand it, I don't care |
| 00:23:12 | FromDiscord | <Invisible> any method that isn't overly complicated will work for me |
| 00:23:24 | FromDiscord | <graveflo> looks like you will use a lock file then |
| 00:23:53 | FromDiscord | <Invisible> sent a code paste, see https://play.nim-lang.org/#ix=4yb3 |
| 00:24:22 | FromDiscord | <graveflo> yea p much |
| 00:24:24 | FromDiscord | <Invisible> But if the program is constantly running, and then crashes / the pc shuts off, it will still think it's running when it restarts |
| 00:24:29 | * | Jjp137 quit (Ping timeout: 240 seconds) |
| 00:24:52 | FromDiscord | <Elegantbeef> Right |
| 00:24:54 | FromDiscord | <graveflo> well thats why you should try and open it for writing and see how it reacts instead... idk if nim has a file access query |
| 00:25:06 | FromDiscord | <Elegantbeef> The unity engine has(had?) this problem |
| 00:25:13 | FromDiscord | <Elegantbeef> So many times I've had to manually delete a lock file |
| 00:25:34 | FromDiscord | <graveflo> right but you can get around this if you use the OS locking mech.. again there are choices here with trade offs |
| 00:25:44 | FromDiscord | <graveflo> using the OS write lock can get around this issue |
| 00:25:50 | FromDiscord | <Invisible> In reply to @graveflo "well thats why you": can you provide me with a basic exmaple to get me started? |
| 00:25:54 | FromDiscord | <Elegantbeef> \nix generally doesnt use file locks iirc, might be crazy |
| 00:26:02 | FromDiscord | <Invisible> Ill figure the rest out on my own, I just want something to base myself off of |
| 00:26:15 | FromDiscord | <graveflo> Im sure you can but this is a windows user i think |
| 00:28:18 | FromDiscord | <huantian> In reply to @Elegantbeef "So many times I've": one way to counter this is to write your pid to the file |
| 00:28:29 | FromDiscord | <huantian> and other instance not only check if the file exists, but if a process with that pid exists I think |
| 00:28:33 | * | Onionhammer quit (Server closed connection) |
| 00:28:36 | FromDiscord | <Invisible> No easy way to get the PID from nimlang |
| 00:28:39 | FromDiscord | <Elegantbeef> Yea that makes sense |
| 00:28:43 | FromDiscord | <Invisible> or check it |
| 00:28:49 | * | Onionhammer joined #nim |
| 00:29:00 | FromDiscord | <Elegantbeef> https://nim-lang.org/docs/os.html#getCurrentProcessId |
| 00:29:18 | FromDiscord | <Elegantbeef> Damn I hate how hard that is |
| 00:29:22 | FromDiscord | <Elegantbeef> I'm going to write a PR to make it easier |
| 00:29:55 | FromDiscord | <graveflo> In reply to @huantian "one way to counter": hey thats pretty smart 😄 |
| 00:30:42 | FromDiscord | <huantian> the internet does occasionally have good ideas 😛 |
| 00:30:45 | FromDiscord | <Elegantbeef> There is a chance that you have a program with that PID open |
| 00:30:55 | FromDiscord | <Elegantbeef> It's not like PIDs are GUIDs |
| 00:31:02 | FromDiscord | <Invisible> In reply to @Elegantbeef "Damn I hate how": How bout checking if the PID is active? |
| 00:31:07 | FromDiscord | <huantian> yeah I'm not too sure what the chance of that is for a userland process on unix |
| 00:31:18 | FromDiscord | <Elegantbeef> Using class names or program names makes more sense imo |
| 00:31:24 | FromDiscord | <graveflo> In reply to @Elegantbeef "There is a chance": true ig you would have to follow that up with another check but its still p cool.. You ruined it a little |
| 00:31:58 | FromDiscord | <Invisible> In reply to @Elegantbeef "Using class names or": they could js rename it between executions, which would defeat this entire thing |
| 00:32:00 | FromDiscord | <Elegantbeef> I just iterate `/proc` 😄 |
| 00:32:10 | FromDiscord | <Elegantbeef> rename it between executions? |
| 00:32:14 | FromDiscord | <graveflo> In reply to @Invisible "they could js rename": who is "they" |
| 00:32:17 | FromDiscord | <Elegantbeef> Are you writing malware or a proper program |
| 00:32:26 | FromDiscord | <Invisible> In reply to @Elegantbeef "Are you writing malware": A proper program, what malware has a GUI |
| 00:32:33 | FromDiscord | <Elegantbeef> Ransomware |
| 00:32:34 | FromDiscord | <graveflo> youd be suprised |
| 00:32:38 | FromDiscord | <graveflo> yea exactly XD |
| 00:32:42 | FromDiscord | <Invisible> In reply to @Elegantbeef "Ransomware": 🤷♂️ ig |
| 00:32:48 | FromDiscord | <Invisible> don't most ransomwaes use text notes? |
| 00:32:49 | FromDiscord | <graveflo> but anyway this is a proper example of user error |
| 00:32:55 | FromDiscord | <graveflo> if they mess with the lock file thats on them |
| 00:33:03 | FromDiscord | <Invisible> In reply to @graveflo "if they mess with": ig |
| 00:33:09 | FromDiscord | <Invisible> 🤷♂️ |
| 00:33:17 | FromDiscord | <Invisible> ill use cmd to query all of the pids |
| 00:33:30 | FromDiscord | <Elegantbeef> Or use winlean or winim |
| 00:33:30 | FromDiscord | <Invisible> and check if the one in the file matches one in there |
| 00:33:37 | FromDiscord | <Elegantbeef> There is no reason to go out to a shell |
| 00:33:48 | FromDiscord | <Invisible> Dont know how to use eiteher |
| 00:35:42 | FromDiscord | <graveflo> I already posted code that queries all PIDs and uses them to get process handles |
| 00:37:10 | * | oldpcuser_ joined #nim |
| 00:39:59 | * | oldpcuser quit (Ping timeout: 240 seconds) |
| 00:40:06 | * | oldpcuser_ quit (Max SendQ exceeded) |
| 00:40:34 | * | oldpcuser_ joined #nim |
| 00:42:38 | * | oldpcuser_ quit (Read error: Connection reset by peer) |
| 00:42:41 | FromDiscord | <Elegantbeef> The best way to learn how to use any system API is to look how they're used in C/C++ |
| 00:42:51 | * | oldpcuser joined #nim |
| 00:42:56 | * | oldpcuser quit (Remote host closed the connection) |
| 00:44:14 | * | oldpcuser joined #nim |
| 00:45:13 | FromDiscord | <graveflo> does not really not have a divmod? |
| 00:46:07 | FromDiscord | <JeysonFlores> can I specify the javascript backend in nimble? |
| 00:49:34 | FromDiscord | <Elegantbeef> make a `config.nims` with `--backend:js` |
| 00:50:22 | FromDiscord | <Elegantbeef> sent a code paste, see https://play.nim-lang.org/#ix=4yb6 |
| 00:51:01 | FromDiscord | <juan_carlos> `backend = "js"` https://nim-lang.org/docs/nimscript.html#backend |
| 00:51:13 | * | jkl quit (Server closed connection) |
| 00:51:21 | FromDiscord | <JeysonFlores> thanks |
| 00:51:34 | * | jkl joined #nim |
| 01:05:56 | * | jmdaemon quit (Ping timeout: 240 seconds) |
| 01:12:19 | * | m5zs7k quit (Ping timeout: 255 seconds) |
| 01:12:47 | * | blackbeard420 quit (Ping timeout: 255 seconds) |
| 01:12:47 | * | cedb quit (Ping timeout: 255 seconds) |
| 01:12:49 | * | mahlon quit (Ping timeout: 255 seconds) |
| 01:13:03 | * | krydos5 joined #nim |
| 01:13:13 | * | systemdsucks quit (Ping timeout: 255 seconds) |
| 01:14:41 | * | mronetwo quit (Ping timeout: 260 seconds) |
| 01:15:50 | * | casaca quit (Ping timeout: 260 seconds) |
| 01:15:51 | * | henrytill quit (Ping timeout: 260 seconds) |
| 01:15:55 | * | hexeme quit (Ping timeout: 260 seconds) |
| 01:16:25 | * | oldpcuser_ joined #nim |
| 01:16:26 | * | noeontheend quit (Ping timeout: 260 seconds) |
| 01:16:26 | * | oddish quit (Ping timeout: 260 seconds) |
| 01:16:27 | * | anddam quit (Ping timeout: 260 seconds) |
| 01:16:28 | * | oisota quit (Ping timeout: 260 seconds) |
| 01:16:29 | * | mal`` quit (Ping timeout: 260 seconds) |
| 01:17:01 | * | fallback quit (Ping timeout: 260 seconds) |
| 01:18:04 | * | calebtj quit (Ping timeout: 255 seconds) |
| 01:18:04 | * | mahlon_ joined #nim |
| 01:18:04 | * | Lord_Nightmare quit (Ping timeout: 255 seconds) |
| 01:18:05 | * | tiorock joined #nim |
| 01:18:15 | * | tiorock quit (Changing host) |
| 01:18:15 | * | tiorock joined #nim |
| 01:18:48 | * | mal``` joined #nim |
| 01:19:03 | * | fallback_ joined #nim |
| 01:19:52 | * | rockcavera quit (Killed (tungsten.libera.chat (Nickname regained by services))) |
| 01:19:52 | * | tiorock is now known as rockcavera |
| 01:20:15 | * | blackbeard420_ joined #nim |
| 01:20:15 | * | Onionhammer2 joined #nim |
| 01:20:20 | * | tk quit (Ping timeout: 240 seconds) |
| 01:20:20 | * | derpydoo quit (Ping timeout: 240 seconds) |
| 01:20:20 | * | Amun-Ra quit (Ping timeout: 240 seconds) |
| 01:20:21 | * | marcus quit (Ping timeout: 240 seconds) |
| 01:20:21 | * | Onionhammer quit (Ping timeout: 240 seconds) |
| 01:20:22 | * | dza quit (Ping timeout: 240 seconds) |
| 01:20:22 | * | krux02 quit (Ping timeout: 240 seconds) |
| 01:20:22 | * | antranigv quit (Ping timeout: 240 seconds) |
| 01:20:22 | * | xet7 quit (Ping timeout: 240 seconds) |
| 01:20:22 | * | krydos quit (Ping timeout: 240 seconds) |
| 01:20:22 | * | Onionhammer2 is now known as Onionhammer |
| 01:20:27 | * | anddam_ joined #nim |
| 01:20:29 | * | Onionhammer quit (Read error: Connection reset by peer) |
| 01:20:35 | * | blackbeard420 joined #nim |
| 01:20:36 | * | mronetwo_ joined #nim |
| 01:20:38 | * | deadmarshal_ quit (Ping timeout: 240 seconds) |
| 01:20:39 | * | oddish_ joined #nim |
| 01:20:40 | * | krydos5 quit (Ping timeout: 240 seconds) |
| 01:20:41 | * | jkl quit (Ping timeout: 240 seconds) |
| 01:20:41 | * | oldpcuser quit (Ping timeout: 240 seconds) |
| 01:20:43 | * | 042AAA61F joined #nim |
| 01:20:44 | * | tiorock joined #nim |
| 01:20:45 | * | anddam_ quit (Changing host) |
| 01:20:45 | * | anddam_ joined #nim |
| 01:21:00 | * | Onionhammer joined #nim |
| 01:21:16 | * | oddish_ quit (Read error: Connection reset by peer) |
| 01:21:19 | * | mronetwo_ quit (Read error: Connection reset by peer) |
| 01:21:22 | * | oddish joined #nim |
| 01:21:24 | * | tiorock quit (Changing host) |
| 01:21:24 | * | tiorock joined #nim |
| 01:21:24 | * | rockcavera is now known as Guest3205 |
| 01:21:24 | * | Guest3205 quit (Killed (molybdenum.libera.chat (Nickname regained by services))) |
| 01:21:24 | * | tiorock is now known as rockcavera |
| 01:21:25 | * | Amun-Ra joined #nim |
| 01:21:27 | * | mronetwo joined #nim |
| 01:22:59 | * | 042AAA61F quit (Ping timeout: 240 seconds) |
| 01:23:08 | * | blackbeard420_ quit (Ping timeout: 240 seconds) |
| 01:23:27 | * | mahlon_ quit (Ping timeout: 255 seconds) |
| 01:24:20 | * | anddam_ quit (Ping timeout: 240 seconds) |
| 01:25:44 | * | azimut quit (Ping timeout: 240 seconds) |
| 01:25:49 | * | mahlon_ joined #nim |
| 01:30:50 | * | tk joined #nim |
| 01:31:00 | * | xet7 joined #nim |
| 01:31:43 | * | noeontheend joined #nim |
| 01:31:44 | * | Lord_Nightmare joined #nim |
| 01:31:48 | * | casaca joined #nim |
| 01:32:05 | * | henrytill joined #nim |
| 01:32:07 | * | deadmarshal_ joined #nim |
| 01:32:36 | * | hexeme joined #nim |
| 01:32:45 | * | m5zs7k joined #nim |
| 01:32:57 | * | jkl joined #nim |
| 01:33:58 | * | systemdsucks joined #nim |
| 01:34:01 | * | marcus joined #nim |
| 01:35:15 | * | oldpcuser_ is now known as oldpcuser |
| 01:35:54 | * | anddam_ joined #nim |
| 01:36:59 | * | ced1 joined #nim |
| 01:37:46 | * | calebjohn joined #nim |
| 01:39:51 | * | dza joined #nim |
| 01:42:16 | FromDiscord | <!&luke> Is there a way to get vars/functions from nimscript with nimscripter without them being exported |
| 01:47:32 | FromDiscord | <Elegantbeef> Nope |
| 01:48:55 | FromDiscord | <Invisible> How do I get the last item in a list |
| 01:49:03 | FromDiscord | <Invisible> list[-1] doesn't work |
| 01:49:11 | FromDiscord | <Elegantbeef> \`^1 |
| 01:49:16 | FromDiscord | <Elegantbeef> `list[^1]`\ |
| 01:49:26 | FromDiscord | <Invisible> ty |
| 01:58:42 | FromDiscord | <!&luke> In reply to @Elegantbeef "Nope": Oh |
| 02:00:08 | * | deadmarshal_ quit (Ping timeout: 255 seconds) |
| 02:00:35 | * | m5zs7k_ joined #nim |
| 02:01:12 | * | Lord_Nightmare2 joined #nim |
| 02:01:21 | * | m5zs7k quit (Quit: m5zs7k) |
| 02:01:22 | * | calebjohn quit (Ping timeout: 255 seconds) |
| 02:01:22 | * | ced1 quit (Ping timeout: 255 seconds) |
| 02:01:28 | FromDiscord | <Elegantbeef> If you care to know, the VM only allows you to access global variables which must be exported |
| 02:01:48 | * | ced1 joined #nim |
| 02:01:48 | * | calebjohn joined #nim |
| 02:01:49 | * | Lord_Nightmare quit (Ping timeout: 255 seconds) |
| 02:04:10 | * | Lord_Nightmare2 is now known as Lord_Nightmare |
| 02:04:12 | FromDiscord | <!&luke> In reply to @Elegantbeef "If you care to": Yea it makes sense |
| 02:05:14 | FromDiscord | <!&luke> Do you know how I can call a function only if it exists? |
| 02:06:29 | FromDiscord | <Elegantbeef> Invoke should raise an exception, so handle that i guess |
| 02:08:40 | FromDiscord | <!&luke> Ok |
| 02:09:11 | FromDiscord | <!&luke> Would the same work for variables |
| 02:09:15 | * | m5zs7k_ is now known as m5zs7k |
| 02:10:45 | FromDiscord | <Elegantbeef> Seems so |
| 02:11:17 | FromDiscord | <!&luke> Oh wait for variables I can just use defaults |
| 02:16:48 | * | deadmarshal_ joined #nim |
| 02:25:22 | * | Jjp137 joined #nim |
| 02:57:17 | FromDiscord | <!&luke> Is there something like strformat that works with runtime strings |
| 02:58:15 | FromDiscord | <Elegantbeef> `%` in strutils |
| 02:58:47 | FromDiscord | <!&luke> Ok thanks |
| 03:27:49 | * | jtbx joined #nim |
| 03:32:16 | * | calebjohn quit (Ping timeout: 255 seconds) |
| 03:34:06 | * | calebjohn joined #nim |
| 03:59:29 | * | Jjp137 quit (Ping timeout: 240 seconds) |
| 03:59:53 | * | Jjp137 joined #nim |
| 04:45:22 | * | jtbx quit (Quit: Lost terminal) |
| 04:45:56 | * | ntat joined #nim |
| 04:51:55 | * | rockcavera quit (Remote host closed the connection) |
| 05:54:10 | FromDiscord | <Invisible> Is there a thing in nim that's similar to python's f strings? |
| 05:54:22 | FromDiscord | <Invisible> like var string = "{string1}, hello, {string2" |
| 05:54:28 | FromDiscord | <Invisible> (edit) "{string2"" => "{string2}"" |
| 05:56:14 | FromDiscord | <Elegantbeef> `std/strformat` |
| 05:57:30 | FromDiscord | <Invisible> k |
| 06:05:12 | FromDiscord | <dissolved.girl> The syntax you're probably looking for is `&"{string1}, hello, {string2}"` |
| 06:05:27 | FromDiscord | <dissolved.girl> From `std/strformat` |
| 06:09:30 | FromDiscord | <Invisible> In reply to @dissolved.girl "The syntax you're probably": It doesn't work if I have curly brackets in my string tho |
| 06:09:35 | FromDiscord | <Invisible> And I can't escape them |
| 06:09:51 | FromDiscord | <Elegantbeef> `{{` |
| 06:10:04 | FromDiscord | <Invisible> Doesn't work |
| 06:10:16 | FromDiscord | <Invisible> already tried |
| 06:10:19 | FromDiscord | <Elegantbeef> > Parts of the string that are enclosed in the curly braces are interpreted as Nim code. To escape a { or }, double it. |
| 06:10:25 | FromDiscord | <Elegantbeef> It certainly does |
| 06:10:27 | FromDiscord | <Invisible> nvm, worded that wrong |
| 06:10:37 | FromDiscord | <Invisible> It works when I run it but it fucks up my syntax coloring |
| 06:11:10 | FromDiscord | <dissolved.girl> That doesn't seem like a Nim issue 😅 |
| 06:11:26 | FromDiscord | <Invisible> 🤷♂️ |
| 06:47:21 | termer | clicked on #nim |
| 06:47:24 | termer | saw Beef |
| 06:47:31 | termer | just another typical day |
| 06:51:48 | * | PMunch joined #nim |
| 07:27:11 | FromDiscord | <Chronos [She/Her]> Would it be easy to treat Redis as a native Nim variable using a proc or something? Why shouldn't I do that? |
| 08:54:39 | FromDiscord | <kaddkaka> NIM spells WIN if u spin it 180 degrees 🤯 |
| 08:56:37 | FromDiscord | <heysokam> you need to spin it recursively to get the real message |
| 08:59:58 | * | beholders_eye joined #nim |
| 09:02:07 | * | anddam_ is now known as anddam |
| 10:21:43 | FromDiscord | <chmod222> You need an australian mirror |
| 11:03:15 | FromDiscord | <Andreas> hi, i have a minor problem passing a proc into a constructor-proc. Heres a simplified example https://play.nim-lang.org/#ix=4ycr↵It involves two concepts, which test ok. Calling `makeChunk` from inside the constructor works. I'm too dump to pass it in. ↵I'm happy for any insight.. |
| 11:11:37 | NimEventer | New thread by ploxotnuj1: Library for tracking pressed keys, see https://forum.nim-lang.org/t/10273 |
| 11:14:18 | * | azimut joined #nim |
| 12:21:13 | FromDiscord | <deech> Is there a `seq` iterator that yields the items in reverse order? |
| 12:21:41 | FromDiscord | <deech> (edit) "Is there a `seq` iterator that yields the items in reverse order? ... " added "I get that `deque` exists but sometimes it's convenient." |
| 12:22:59 | PMunch | `iterator backwards[T](x: seq[T]): T = for i in countdown(x.high, x,low): yield x[i]` |
| 12:23:33 | FromDiscord | <deech> Does this exist in the standard library? |
| 12:25:00 | FromDiscord | <Andreas> @pmunch hi, pls take a look here https://play.nim-lang.org/#ix=4ycr i worked around it, but that solution is silly.. |
| 12:25:10 | FromDiscord | <Andreas> (edit) "that" => "my" |
| 12:25:21 | PMunch | deech, no only in your imagination :P |
| 12:25:47 | PMunch | I don't think there's anything like that in there, if there was I'd hope it would take an openarray instead |
| 12:26:03 | PMunch | Andreas, take a look at what? |
| 12:26:39 | FromDiscord | <intellj_gamer> There's `reversed` (https://nim-lang.org/docs/algorithm.html#reversed%2CopenArray%5BT%5D) but it sadly doesn't have an iterator overload |
| 12:26:49 | FromDiscord | <deech> Thanks! I try to use the stdlib when possible but this is neat as well. |
| 12:28:02 | FromDiscord | <deech> In reply to @intellj_gamer "There's `reversed` (https://nim-lang.org/docs/algor": I saw that but it actually does the work of in place reversing the seq which is I don't need, I just need to read it backwards so an iterator or for loop is ideal. |
| 12:28:18 | FromDiscord | <deech> (edit) removed "is" |
| 12:30:41 | FromDiscord | <intellj_gamer> That's `reverse`, `reversed` returns a new seq. |
| 12:30:45 | FromDiscord | <intellj_gamer> (edit) "seq." => "seq" |
| 12:31:06 | FromDiscord | <Andreas> In reply to @PMunch "<@754284517316886598>, take a look": https://play.nim-lang.org/#ix=4ycL this on. i am obviously too dumb too pass a proc into a constructor |
| 12:31:25 | FromDiscord | <Andreas> (edit) "on." => "one here." |
| 12:31:25 | FromDiscord | <deech> Oh I missed that! I'll keep it in mind for the future. Thanks! |
| 12:31:38 | * | ntat_ joined #nim |
| 12:31:55 | * | ntat quit (Ping timeout: 260 seconds) |
| 12:33:19 | * | ced1 is now known as cedb |
| 12:34:46 | FromDiscord | <intellj_gamer> think the convention is past tense means it returns a new value while present does in place 🤔 like `sort` is in place but `sorted` returns a new seq |
| 12:35:47 | * | oisota joined #nim |
| 12:37:31 | PMunch | Andreas, should just be a matter of passing the name |
| 12:37:56 | PMunch | But I'm not reading 83 lines of code to figure out where you actually try to do it |
| 12:40:10 | FromDiscord | <Andreas> In reply to @PMunch "But I'm not reading": i want to pass `proc chunkMaker[T]` into the `initPool`-constructor, and thats not working. Neither when i defina a type for it or describe it as `proc initPool[T]( cb :proc( pc :PoolConcept[T]) :ChunkConcept[T] ) :DumbArena[T] ` |
| 12:41:51 | PMunch | Try to create a minimal reproduction |
| 12:46:09 | FromDiscord | <Andreas> In reply to @PMunch "Try to create a": its < 50-loc https://play.nim-lang.org/#ix=4ycO problem is line-48 |
| 13:03:29 | * | krux02 joined #nim |
| 13:22:15 | * | xet7 quit (Quit: Leaving) |
| 13:25:30 | * | ntat_ is now known as ntat |
| 13:27:29 | FromDiscord | <voidwalker> Do I remember well, was there some way to tell nim not to pre-allocate/initialize the result var in procs ? |
| 13:27:40 | PMunch | {.noinit.} |
| 13:29:45 | PMunch | Andreas, what is up with this spacing?! |
| 13:30:17 | FromDiscord | <Andreas> In reply to @PMunch "<@754284517316886598>, what is up": wdym ? its great and VSCode wants it that way |
| 13:30:37 | PMunch | You should fix your VSCode setup, nothing about this is great.. |
| 13:31:38 | FromDiscord | <Andreas> In reply to @PMunch "You should fix your": well, its your playground, isn't it - you could pipe all user-code thru nimpretty ? |
| 13:31:43 | FromDiscord | <voidwalker> hm ok that skips the initialization, but it still allocates memory for the result, right ? can this be skipped as well if we return another variable ? |
| 13:31:58 | PMunch | Hmm, I thought [:int] would work, but then it complains about braces.. Not sure if there's a way to do this, might not like the concepts mixed with generics in that way.. |
| 13:32:35 | PMunch | voidwalker, what do you mean? The result needs space to live in.. |
| 13:33:46 | PMunch | When you make a call the arguments are pushed onto the stack, then it skips forward the amount of space of the return value, and then it jumps to the code. When the code returns the caller can expect the return value to be the last thing on the stack, that's just how function calling works |
| 13:34:01 | FromDiscord | <voidwalker> yeah I guess I don't really understand how this works internally. let's say you return the string and choose to build it in another variable, not result (cause result is too long, maybe if it was ret like in Go..). That means you have two initializations of memory space for the same goal. How can this be avoided ? |
| 13:34:21 | PMunch | By using result.. |
| 13:34:36 | PMunch | That's the whole point of result |
| 13:34:40 | FromDiscord | <voidwalker> is there no way to alias result to another name ? : ) |
| 13:34:47 | FromDiscord | <voidwalker> like r or ret or res |
| 13:34:48 | PMunch | Well.. |
| 13:34:55 | PMunch | You could use a template |
| 13:35:45 | FromDiscord | <odexine> In reply to @PMunch "<@754284517316886598>, what is up": You haven’t met them yet huh |
| 13:35:45 | FromDiscord | <Andreas> In reply to @PMunch "Hmm, I thought [:int]": hmm, that was my assumption, too. Is that smth. i should report ? Cos' passing around concepts is important and the only way to make use of them - other than that it a pretty "java-ish"-kata for the docs.. |
| 13:36:09 | PMunch | voidwalker, like so: https://play.nim-lang.org/#ix=4yd0 |
| 13:37:23 | FromDiscord | <voidwalker> hm i guess.. this is quite an okay solution |
| 13:37:53 | PMunch | Of course if you want it slightly more ergonomic and plan to use this a lot you could do something like this: https://play.nim-lang.org/#ix=4yd1 |
| 13:38:19 | PMunch | Feel free to swap x and y if you prefer that |
| 13:38:37 | FromDiscord | <voidwalker> yeah that is quite perfect now : ) |
| 13:39:05 | PMunch | odexine, I think I've complained about it before :P |
| 13:39:09 | FromDiscord | <voidwalker> Good for when you have a big ass proc with a huge object to build, and have to type result 100 times |
| 13:39:44 | FromDiscord | <voidwalker> if it's a ref object, you can just do `var r = result` |
| 13:39:50 | FromDiscord | <odexine> Personally I don’t think “ret” explains enough of what it actually is |
| 13:39:59 | FromDiscord | <odexine> Nor does “r” |
| 13:40:00 | PMunch | Andreas, sure reporting it seems like a good idea. But I guess since the concept is still open I'm not sure that could work.. |
| 13:40:33 | PMunch | Definitely makes your code harder to scan |
| 13:40:50 | PMunch | Especially if your editor highlights `result` in a different colour like mine does |
| 13:40:55 | FromDiscord | <voidwalker> well, there's a middle ground between readability (by non nim user), readability by nim user, and writing efficiently, succint code as a proficient nim user |
| 13:41:37 | FromDiscord | <voidwalker> I normally don't mind `result` except when it's all over the place, and makes lines too long so you have to split them etc |
| 13:43:41 | FromDiscord | <voidwalker> Another topic I'd like some insight into. Nim strings allow 0/null value, cstrings don't. Should strings be used as storage for raw binary data then? |
| 13:46:26 | FromDiscord | <odexine> In reply to @voidwalker "Another topic I'd like": ? |
| 13:46:28 | FromDiscord | <odexine> They do? |
| 13:46:37 | FromDiscord | <odexine> Please check again |
| 13:46:52 | FromDiscord | <voidwalker> they do ? I thought strings in C (not cstrings) don't |
| 13:47:01 | FromDiscord | <odexine> Wait you mean in the fact that they can contain zeroed bytes |
| 13:47:06 | FromDiscord | <odexine> Okay that’s a strange sentence to parse |
| 13:47:23 | FromDiscord | <Andreas> In reply to @PMunch "<@754284517316886598>, sure reporting it": wdym "concept is still open" ? it's experimental ? |
| 13:48:18 | FromDiscord | <arnetheduck> In reply to @voidwalker "Good for when you": you can avoid both problems by using the object construction syntax (`Obj(field: a, field2: b)`: https://status-im.github.io/nim-style-guide/language.objconstr.html) |
| 13:50:07 | * | PMunch_ joined #nim |
| 13:50:36 | FromDiscord | <odexine> In reply to @voidwalker "Another topic I'd like": Strings can indeed be used to store raw binary data but i would point you at seq[byte] instead |
| 13:51:16 | * | PMunch quit (Ping timeout: 260 seconds) |
| 13:51:28 | * | psydruid quit (Write error: Connection reset by peer) |
| 13:51:28 | * | frenchboy[m] quit (Write error: Connection reset by peer) |
| 13:51:33 | FromDiscord | <arnetheduck> In reply to @PMunch "Of course if you": fwiw, solutions like this go out of their way to preserve the poor style that results from `result` usage, ie that of mutating an object that initially starts out in an invalid state - it's a very common source of bugs in Nim, that code either forgets to write or overwrites a partially initialized obect because of this |
| 13:51:43 | FromDiscord | <voidwalker> @arnetheduck that's not efficient, I don't want double initialization |
| 13:53:06 | FromDiscord | <voidwalker> why would I init with 0/empty values only to init again with the proper computed values |
| 13:53:26 | FromDiscord | <odexine> You wouldn’t using object constructor syntax? |
| 13:53:58 | FromDiscord | <odexine> Not sure where you think the double initialisation comes from |
| 13:54:13 | FromDiscord | <arnetheduck> In reply to @voidwalker "why would I init": you're in for a bad surprise: nim will zero your object twice _before_ you can start using `result....` due to bugs |
| 13:54:32 | FromDiscord | <odexine> Why twice? That sounds strange |
| 13:54:54 | FromDiscord | <voidwalker> I know it inits it once, and with {.noInit.} we're supposed to get 0 inits |
| 13:54:59 | FromDiscord | <arnetheduck> once in the caller and once in the function itself |
| 13:55:16 | * | PMunch joined #nim |
| 13:55:18 | FromDiscord | <voidwalker> that sounds like a bad bug, but if it's a bug, then it will be fixed some day ? |
| 13:55:30 | FromDiscord | <voidwalker> We should code for the future, not the buggy present : P |
| 13:55:32 | * | PMunch_ quit (Ping timeout: 240 seconds) |
| 13:55:59 | FromDiscord | <voidwalker> if the function returns a ref object, does it still do inits ? |
| 13:56:05 | FromDiscord | <arnetheduck> in that future, you want to use object construction syntax as much as possible - it's the one that gives the most relevant intent to the compiler so the compiler can know how to optimize it |
| 13:56:16 | FromDiscord | <voidwalker> https://github.com/sgmihai/torrentim/blob/8e9e0491a3eae432b28b9d0e6a11ffd80658c802/torrentim.nim#L112 |
| 13:56:22 | FromDiscord | <arnetheduck> `noinit` is another bug workaround, essentially |
| 13:56:40 | FromDiscord | <voidwalker> this is the constructor I had in mind when asking about this |
| 13:56:52 | FromDiscord | <arnetheduck> compilers can _generally_ understand that when writing 0 then writing a value to a memory location, the first write can be removed - that only works for specific cases though |
| 13:56:53 | FromDiscord | <voidwalker> What could I do better ? |
| 13:57:33 | FromDiscord | <arnetheduck> nim here does a few things which prevents this optimization from happening, but that's a different story (related to said bugs) |
| 13:58:15 | FromDiscord | <demotomohiro> sent a long message, see http://ix.io/4yda |
| 13:58:46 | FromDiscord | <arnetheduck> In reply to @voidwalker "https://github.com/sgmihai/torrentim/blob/8e9e0491a": this is exactly the kind of code that would be better off using init syntax - it would save performance and be more elegant / less boiler-plated |
| 13:59:00 | * | frenchboy[m] joined #nim |
| 13:59:19 | FromDiscord | <arnetheduck> you'd also reduce the risk of overwriting fields several times and a few other ugly things.. |
| 13:59:20 | * | psydruid joined #nim |
| 14:00:05 | * | PMunch_ joined #nim |
| 14:00:35 | * | PMunch quit (Ping timeout: 260 seconds) |
| 14:00:39 | FromDiscord | <voidwalker> still not sure how this would look like.. |
| 14:01:45 | FromDiscord | <voidwalker> put each field of the object in a separate var, then call init again with all that huge chunk of values ? |
| 14:01:56 | FromDiscord | <voidwalker> that sounds very messy |
| 14:01:56 | FromDiscord | <arnetheduck> sent a code paste, see https://paste.rs/lQThu |
| 14:02:05 | FromDiscord | <voidwalker> oh |
| 14:03:06 | FromDiscord | <arnetheduck> when written like that, nim usually makes use of something called RVO where if you call `field: someFunction()`, the function will write to the field directly instead of makign a copy first |
| 14:03:16 | FromDiscord | <arnetheduck> (edit) "makign" => "making" |
| 14:04:12 | FromDiscord | <voidwalker> so I should have a function to parse every field :\ |
| 14:05:06 | FromDiscord | <arnetheduck> well, you don't have to - you can mix in expressions and other things as well - they will _typically_ write directly to the memory location of the field except for a number of cases which are bugs |
| 14:06:29 | * | azimut_ joined #nim |
| 14:06:43 | * | azimut quit (Remote host closed the connection) |
| 14:09:35 | FromDiscord | <arnetheduck> sent a code paste, see https://play.nim-lang.org/#ix=4ydg |
| 14:11:08 | FromDiscord | <odexine> I was going to mention whether these double initialisations actually mattered in the application |
| 14:11:19 | FromDiscord | <odexine> Sounded like premature optimisation |
| 14:12:23 | FromDiscord | <voidwalker> this all talk is about premature optimisation : P |
| 14:13:19 | FromDiscord | <voidwalker> Probably doesn't matter too much, it's about adding/parsing a torrent. There's many more optimisations to be done at parsing the bencoded data efficiently, vs using a generic parser |
| 14:13:50 | FromDiscord | <arnetheduck> I look at it from the point of view of clarity - if you're doing lots of redundant repetitive work (such as repeating `result` all the time in your code), so does the reader _and so does the compiler preventing it from optimising for you_ - it's a lose-lose-lose situation |
| 14:14:01 | FromDiscord | <demotomohiro> Nim and backend C compiler do many good optimizations. You don't need to do manual optimization that compiler do. |
| 14:14:29 | FromDiscord | <arnetheduck> there are a few notable exceptions to that rule, but unless you can notice them when running the app, well .. |
| 14:16:10 | * | PMunch_ quit (Quit: Leaving) |
| 14:17:41 | * | xet7 joined #nim |
| 14:21:45 | * | rockcavera joined #nim |
| 14:21:58 | FromDiscord | <spotlightkid> Why do I see many threads with activity since my last visit on the Nim forum, but if I open them, the last post is from months ago? Spam? New posts from moderated users? |
| 14:25:20 | FromDiscord | <yandall> Hi, someone has already faced this problem? I have a memory leak when using prologe, I think my problem is related to this one but I haven't been able to find a solution https://github.com/nim-lang/Nim/issues/21631 |
| 14:28:28 | FromDiscord | <mratsim> In reply to @deech "I saw that but": you can use `dup`to transform in-place in out-place, reusing buffers if needed. |
| 14:30:28 | FromDiscord | <mratsim> In reply to @voidwalker "yeah I guess I": stack allocation is just `sub rsp, 0x10` assuming you need 16 bytes on the stack |
| 14:31:02 | FromDiscord | <mratsim> if the compiler finds that 200 bytes are needed for 10 objects it won't make 10 allocations, it will do it all at once. |
| 14:32:42 | FromDiscord | <odexine> In reply to @spotlightkid "Why do I see": both yes |
| 14:32:51 | FromDiscord | <voidwalker> Alright, about using strings to store null values, should that be done ? What are the downsides ? compatibility problems if using the nim->c code in a c lib ? |
| 14:33:34 | FromDiscord | <mratsim> In reply to @voidwalker "Alright, about using strings": null values?↵↵a null pointer? or a 0x00 byte? |
| 14:33:35 | FromDiscord | <voidwalker> I am asking cause, I asked for a patch for a great nim library, binarylang, to support 0 values in strings, and author told me strings should be used for text only :\ |
| 14:33:40 | FromDiscord | <voidwalker> 0 value byte |
| 14:34:08 | FromDiscord | <mratsim> he is right. And cstring ends with 0x00 byte. |
| 14:34:22 | FromDiscord | <mratsim> so you can't use them |
| 14:34:27 | FromDiscord | <voidwalker> I have fixed/known-size string to parse in binarylang, and it breaks before the size because the default end termination for strings is 00 value |
| 14:34:42 | FromDiscord | <voidwalker> (edit) "I have fixed/known-size string to parse in binarylang, and it breaks before the size because the default end termination ... for" added "check in binaryalng" |
| 14:35:02 | FromDiscord | <voidwalker> but there is no other data structure to hold chunks of binary data in binarylang, so we have to use string |
| 14:35:13 | FromDiscord | <odexine> strings can have \0 bytes in them |
| 14:35:17 | FromDiscord | <mratsim> if you want to deal with binary data, use seq[byte] |
| 14:35:39 | FromDiscord | <graveflo> In reply to @demotomohiro "Anyone know how to": Are you looking to do something specific? It seems you have most of all you need to get started |
| 14:35:41 | FromDiscord | <mratsim> In reply to @odexine "strings can have \0": not if you need to pass them to a C function after |
| 14:35:57 | FromDiscord | <odexine> that is correct only if the C function expects \0 termination |
| 14:36:08 | FromDiscord | <voidwalker> ok so the whole nim stdlib is badly design by using mostly strings for everything binary |
| 14:36:11 | FromDiscord | <odexine> if it uses buf+len style, then it could work |
| 14:36:32 | FromDiscord | <odexine> In reply to @voidwalker "ok so the whole": thats what backward compatibility gets you id say |
| 14:36:42 | FromDiscord | <mratsim> In reply to @voidwalker "ok so the whole": I agree: https://github.com/nim-lang/RFCs/issues/32 |
| 14:36:47 | FromDiscord | <odexine> but yes |
| 14:37:39 | FromDiscord | <spotlightkid> @voidwalker\: then binarylang apparently isn't the right tool for the job. If the author won't change it, you have two options\: fork it or look for another tool. |
| 14:38:30 | FromDiscord | <voidwalker> he might change it, I was just curious if the reasoning is right, about strings having/not having 0x00 values in them |
| 14:39:19 | FromDiscord | <voidwalker> this whole null terminated string bs is 50 year old legacy now :\ |
| 14:41:56 | FromDiscord | <demotomohiro> In reply to @graveflo "Are you looking to": It seems `jiro4989/setup-nim-action` installs devel Nim using choosenim and I think installed Nim contains compiler modules.↵But `import compiler/platform` fails.↵Setting `--path` options might fix the problem but there are 3 OS that tests runs and they installs Nim in different places. |
| 14:41:57 | FromDiscord | <spotlightkid> maybe point him at https://docs.python.org/3/library/struct.html ?↵`struct.unpack("4s", b"foo\0")` will happily include the null-byte in the returned string. |
| 14:45:42 | FromDiscord | <voidwalker> https://sealmove.github.io/binarylang/#features-strings |
| 14:45:53 | FromDiscord | <voidwalker> `Strings are special because they don't have a fixed size. Therefore, you must provide enough information regarding their termination. ` |
| 14:46:02 | FromDiscord | <voidwalker> `s: a # null/eos-terminated (because next field doesn't use assertion)` |
| 14:46:47 | FromDiscord | <voidwalker> But I also think it's kind of wrong to arbitrarily assume a string would be terminated by a null byte, in a binary data structure ? |
| 14:46:56 | FromDiscord | <odexine> i think thats a different string from what we thought what you meant? |
| 14:47:11 | * | azimut_ quit (Remote host closed the connection) |
| 14:47:41 | * | azimut joined #nim |
| 14:48:06 | FromDiscord | <odexine> `Strings are null/eos-terminated unless assertion is used on the same field or on the next field`\ |
| 14:48:08 | FromDiscord | <odexine> (edit) "field`\" => "field`" |
| 14:48:20 | FromDiscord | <odexine> so use an assertion on the next field, i dont understand the problem |
| 14:48:25 | FromDiscord | <voidwalker> I was asking in a general sense, about the theoretical (philosophical even) aspects of using strings vs other data structures. In the context of trying to figure out the correct way to fix binarylang |
| 14:48:55 | FromDiscord | <odexine> strings are not meant to hold binary data, the meaning of string is usually attributed to text |
| 14:49:44 | FromDiscord | <voidwalker> well I cannot force my data structure to use assertion, if there's no assertion to be had. I just know the length of the string to be read. And this works, unless the string has 0x00 inside it, in which case the parsing breaks (it could be argued that no parsing should be done in this case, but a simple blind copy) |
| 14:52:46 | FromDiscord | <odexine> then use a seq[byte] (or integer)↵`8: thefield[length]` |
| 14:53:51 | FromDiscord | <odexine> i dont understand why you wanted to use a string for this |
| 14:54:41 | FromDiscord | <voidwalker> well I guess I could do that, and then I would have to copymem into a string again.. would have been better if it was string to begin with |
| 14:55:39 | FromDiscord | <odexine> why do you want it to be a string |
| 14:56:16 | FromDiscord | <voidwalker> Cause it's data that gets sent/received from socket, at a high speed. It would be best to avoid it being copied uselessly many times |
| 14:56:27 | FromDiscord | <odexine> that does not answer my question |
| 14:56:48 | FromDiscord | <odexine> what does the string type net you that the seq[int] data type does not |
| 14:56:50 | FromDiscord | <voidwalker> well, socket works with pointers as well.. |
| 14:57:12 | FromDiscord | <voidwalker> convenience I guess, and not having to rewrite any code : D |
| 14:57:53 | FromDiscord | <voidwalker> I guess I have no choice but to rewrite to `8: field[length]` style, and use pointers for socket instead of string var |
| 14:58:17 | FromDiscord | <odexine> so basically nim stdlib strikes again i guess |
| 14:58:17 | FromDiscord | <odexine> xd |
| 15:00:08 | FromDiscord | <voidwalker> why couldn't they make string type in nim with no backwards compatibility, so it can be interchanged with arrays, seq of byte etc, without needing copies and such. This is one of the things that bugs me the most about nim. In other languages, data arrays are more homogenous |
| 15:00:15 | FromDiscord | <odexine> okay so uh |
| 15:00:33 | FromDiscord | <odexine> technically you can cast strings to seqs as long as the backing type is 8 bits |
| 15:00:49 | FromDiscord | <odexine> strings and seqs are internally represented the same as of now |
| 15:01:37 | FromDiscord | <voidwalker> https://github.com/nim-lang/Nim/issues/14810 |
| 15:02:05 | FromDiscord | <odexine> yes |
| 15:02:54 | FromDiscord | <voidwalker> it's really breaking ergonomics if I can't have a byte array be treatead as a string and so on |
| 15:27:15 | NimEventer | New Nimble package! nimmicrograd - Nim implementation of micrograd autograd engine., see https://github.com/soheil555/nimmicrograd |
| 15:30:01 | FromDiscord | <odexine> huh |
| 15:43:06 | FromDiscord | <michaelb.eth> In reply to @voidwalker "why couldn't they make": string is a gc'd thing in Nim, which leads to it needing to be treated separately from other datastructures that have the same shape internally, that's what I understood from the manual, if I'm remembering correctly |
| 15:44:02 | FromDiscord | <michaelb.eth> maybe things are different/better with orc/arc, but with refc I'm aware you have to be a bit careful going back and forth betweeng string and seq |
| 15:44:28 | FromDiscord | <michaelb.eth> stew/byteutils has facilities for that purpose which avoid problems you can run into |
| 15:46:41 | FromDiscord | <odexine> In reply to @michaelb.eth "string is a gc'd": but sequences are as well, so |
| 15:47:18 | FromDiscord | <michaelb.eth> true, I'm re-reading now |
| 15:48:27 | FromDiscord | <voidwalker> not closing httpclient in a proc.. the http objects gets garbage collected, but it will leak file descriptors ? am I right ? |
| 15:49:54 | FromDiscord | <voidwalker> I thought it was okay not to close it, but running the program in console got me an exception because of too many file descriptors, while running it in vscode didn't. Probably vscode sets some env vars so the value is higher ? |
| 15:49:59 | FromDiscord | <Chronos [She/Her]> Does `chronos` just slot in to any async project when changing the async backend or does that screw with anything? |
| 15:50:17 | FromDiscord | <Chronos [She/Her]> Like do I need to do anything specifically besides toggling the flag? |
| 15:50:34 | FromDiscord | <voidwalker> I'd very much like to know that myself, I found no info of how this swapping actually works, what to look out for etc |
| 15:50:46 | FromDiscord | <voidwalker> what happens if you import libs that import asyncdispatch |
| 15:51:23 | FromDiscord | <Chronos [She/Her]> Yeah, I've heard orc can cause memory leaks with async so I'd like to avoid it |
| 15:51:44 | FromDiscord | <Chronos [She/Her]> In reply to @voidwalker "what happens if you": I believe in the docs it says it'll use the other backend? |
| 15:51:51 | FromDiscord | <Chronos [She/Her]> But not sure tbh |
| 15:51:52 | FromDiscord | <michaelb.eth> asyncdispatch event loop and chronos event loop cannot run on the same thread, so I believe that per-thread you have to be consistent |
| 15:52:11 | FromDiscord | <Chronos [She/Her]> In reply to @michaelb.eth "asyncdispatch event loop and": If my entire application uses chronos then it's fine? |
| 15:52:46 | FromDiscord | <michaelb.eth> as long as you're not making use of some async thing that will implicitly start the asyncdispatch event loop |
| 15:52:54 | FromDiscord | <michaelb.eth> you should be ok |
| 15:52:55 | FromDiscord | <voidwalker> how does this compile param switch even work ? you can import either chronos or asyndispatch and it won't matter ? |
| 15:54:55 | FromDiscord | <michaelb.eth> In reply to @voidwalker "how does this compile": is it `-d:useChronos` or something like that? |
| 15:54:56 | FromDiscord | <voidwalker> if so, what stops the nim stdlib asyncdispatch stuff to be compiled to work with chronos ? |
| 15:55:16 | FromDiscord | <voidwalker> (edit) "if so, what stops the nim stdlib asyncdispatch stuff to be compiled to work with chronos ? ... " added "(if otherwise compatible syntax/features are used)" |
| 15:55:32 | FromDiscord | <Chronos [She/Her]> In reply to @michaelb.eth "is it `-d:useChronos` or": Iirc `--asyncBackend:chronos`...? |
| 15:56:09 | FromDiscord | <michaelb.eth> ah right: https://github.com/status-im/nim-chronos#multiple-async-backend-support |
| 15:56:28 | FromDiscord | <michaelb.eth> that's not something baked into Nim, it's a suggested convention that library/module authors can use |
| 15:58:48 | FromDiscord | <voidwalker> oh so it comes from importing chronos, it cannot work without import chronos |
| 16:01:25 | FromDiscord | <michaelb.eth> per that readme, a library author could choose which async backend libary to import by checking `when defined(asyncBackend): if asyncBackend == chronos: ... else: ...` |
| 16:01:46 | FromDiscord | <michaelb.eth> (edit) "per that readme, a library author ... could" added "(or you in your code)" |
| 16:02:07 | FromDiscord | <voidwalker> ohhh so it's just a convention to be used |
| 16:02:11 | FromDiscord | <michaelb.eth> alternatively, a library may provide different entrypoints, one for chronos compat, another for asyncdispatch compat |
| 16:02:16 | FromDiscord | <michaelb.eth> In reply to @voidwalker "ohhh so it's just": yes, correct |
| 16:02:26 | FromDiscord | <voidwalker> WHY didn't they say so 😄 |
| 16:02:55 | * | junaid_ joined #nim |
| 16:02:56 | FromDiscord | <michaelb.eth> I think they do, just may not have been clear on first reading |
| 16:05:38 | * | beholders_eye quit (Ping timeout: 265 seconds) |
| 16:08:56 | FromDiscord | <michaelb.eth> In reply to @voidwalker "if so, what stops": the dispatchers (event loops) are different; I'd need to check, but I think the `{.async.}` macro of asyncdispatch ties you to asyncdispatch's dispatcher, while chronos' `{.async.}` ties you to chronos' dispatcher |
| 16:09:09 | FromDiscord | <michaelb.eth> and as I mentioned before, you can't have both dispatchers running in the same thread |
| 16:09:51 | FromDiscord | <voidwalker> yes I am aware of that. I read that doc entry a few times and it is still not clear to me that asyncBackend is a convention and not something built in chronos (or asyncdispatch) |
| 16:10:01 | FromDiscord | <voidwalker> (edit) "yes I am aware of that. I read that doc entry a few times and it is still ... not" added "was" |
| 16:12:51 | FromDiscord | <voidwalker> I wonder then how much of stdlib works with chronos out of the box, maybe they could patch the sources to allow this chronos/asyncdispatch switch, would make using it in such projects easier |
| 16:13:23 | FromDiscord | <michaelb.eth> good point, maybe it does need to be stated more clearly:↵> Libraries built on top of async/await may wish to support multiple async backends. The are suggested conventions for doing that, but keep in mind they are only conventions and it is up to individual library and package authors to implement them↵> ↵> Tthe best way to do so is to create separate modules for each backend that may be imported side-by-side.... |
| 16:13:33 | FromDiscord | <michaelb.eth> (edit) "The" => "There" |
| 16:13:53 | FromDiscord | <michaelb.eth> (edit) "that," => "so," | "them↵>" => "them.↵>" |
| 16:21:57 | * | junaid_ quit (Quit: leaving) |
| 17:05:16 | FromDiscord | <!&luke> Is there a way to use fakeroot in nim |
| 17:10:55 | FromDiscord | <!&luke> Can setCurrentDir be used in conjunction with exec |
| 17:15:37 | FromDiscord | <demotomohiro> `proc execProcess` has `workingDir` parameter. |
| 17:16:48 | * | oldpcuser quit (Ping timeout: 248 seconds) |
| 17:33:31 | * | oldpcuser joined #nim |
| 17:55:57 | * | beholders_eye joined #nim |
| 18:05:27 | FromDiscord | <!&luke> In reply to @demotomohiro "`proc execProcess` has `workingDir`": I don't think I can use that in nimscript can I? |
| 18:16:52 | * | frenchboy[m] quit (Remote host closed the connection) |
| 18:18:10 | * | frenchboy[m] joined #nim |
| 18:19:28 | * | cedb quit (Quit: WeeChat 3.8) |
| 18:29:00 | FromDiscord | <Prestige> Anyone else using the latest release of pixie? It seems broken |
| 18:30:14 | FromDiscord | <demotomohiro> In reply to @ripluke "I don't think I": Probably you cannot use them in nimscript. |
| 18:31:25 | FromDiscord | <Prestige> Ah it was an issue with nimble I guess, I had to delete chroma and reinstall dependencies |
| 19:06:31 | * | beholders_eye quit (Ping timeout: 240 seconds) |
| 19:26:36 | * | ntat quit (Quit: Leaving) |
| 19:52:40 | NimEventer | New Nimble package! nimegenerator - Random name/word generator., see https://github.com/nirokay/nimegenerator |
| 21:25:43 | * | ehmry quit (Quit: https://quassel-irc.org - Chat comfortably. Anywhere.) |
| 21:26:33 | * | adium quit (Server closed connection) |
| 21:26:49 | * | adium joined #nim |
| 21:27:10 | * | ehmry joined #nim |
| 22:00:08 | * | candlestick joined #nim |
| 22:00:13 | * | candlestick left #nim (#nim) |
| 22:21:41 | * | ced1 joined #nim |
| 22:21:51 | * | ced1 is now known as cedb |
| 23:51:04 | termer | Doing a livestream trying out the CPS concurrency library. I'm aiming to learn it for some experiments outside of async/await since asyncdispatch is so bad |
| 23:51:07 | termer | https://live.termer.net/ |