| 31 Jul 2021 |
 timotheecour#6096 | * hm... how come it's so hard to find this functionality in other languages; ya i'm having a hard time trying to break it | 19:12:17 |
 saem | timotheecour#6096: was it intentional to not have semLowerLetVarCustomPragma work for a let/var section with multiple identdefs? | 20:09:16 |
| timotheecour#6096 | the implementation was intentionally restrictive to tackle this case 1st; it can always be extended in future work to support multiple identdefs; but note that https://github.com/nim-lang/Nim/issues/15920 should be fixed first before extending the syntax | 20:15:43 |
| saem | personally, i don't see how restricting this helps, but ok
let
s = @["foo"]
foo {.whatever.} = s[0]
| 20:18:53 |
| saem | * personally, i don't see how restricting this helps, but ok
let
s = @["foo"]
foo {.whatever.} = s[0]
| 20:19:01 |
| saem | * personally, i don't see how restricting this helps, but ok
let
s = @["foo"]
foo {.whatever.} = s[0]
| 20:19:10 |
| saem | * personally, i don't see how restricting this helps, but ok
let
s = @["foo"]
foo {.whatever.} = s[0]
| 20:19:19 |
| timotheecour#6096 | PR welcome to do that, I can review it if you write it | 20:28:41 |
| saem | I think the feature needs a rethink, because the lowering should be var foo(a, MyType, expr) and not foo(a, MyType, expr), as it's on the identdef and not the var or let section. | 20:42:12 |
| timotheecour#6096 | tell me how you'd implement byaddr with your lowering | 20:51:31 |
| saem | sure would be nice if those sections were desugared. | 21:00:33 |
|  diego_garro joined the room. | 23:30:00 |
| 1 Aug 2021 |
|  yusepgl joined the room. | 07:04:50 |
| timotheecour#6096 | I don't understand your suggestion and how it would help writing things like byaddr | 16:36:45 |
| saem | Make let and var sections single identdefs as a desugaring step would mean byaddr and such templates would remain trivial and written as is. | 16:54:42 |
| saem | This for of lowering is used in CPS plenty and it dramatically simplifies things. | 16:55:24 |
|  Daniel Developer joined the room. | 17:45:56 |
|  mnimk joined the room. | 17:48:38 |
| timotheecour#6096 | oh, that's what you mean, yes, this would be a welcome simplification for var/let/const sections (not type sections because of mutual recursion) | 17:53:32 |
| saem | Yup, type sections would need some prototype/placeholder concept in order to break cycles and honestly I don't know if it's worth the juggling. | 17:55:19 |
| timotheecour#6096 | it's not, but what's worth fixing is this: https://github.com/nim-lang/Nim/issues/13830 which is very related. | 18:35:16 |
| timotheecour#6096 | * it's not, but what's worth fixing is this: https://github.com/nim-lang/Nim/issues/13830 which is very related. | 18:35:22 |
 Clyybber#8533 | I think moduleOverride is the mechanism that could allow package managers to supply different versions of a package to different modules | 19:35:37 |
| Clyybber#8533 | it's equivalent to the .path file idea that would be like a config file but read in subdirectories too | 19:36:09 |
 leorize | wouldn't package-local path works better? | 19:50:27 |
 dom96#7486 | ideally the package manager should be able to pass a dependency tree to Nim | 20:20:29 |
| Clyybber#8533 | yeah, basically the same thing | 21:12:31 |
| saem | Is there a list of issues that are a must prior to releasing 1.6? | 22:21:54 |
| dom96#7486 | likely those tagged "showstopper" | 22:22:26 |
|  dave the knave joined the room. | 23:48:34 |
