🦋 Welcome to the IRC channel of the core developers of the Raku Programming Language (raku.org #rakulang). This channel is logged for the purpose of history keeping about its development | evalbot usage: 'm: say 3;' or /msg camelia m: ... | log inspection situation still under development | For MoarVM see #moarvm
Set by lizmat on 22 May 2021.
00:00 frost joined 00:05 reportable6 left 00:07 reportable6 joined
Xliff Is there a way to get zef to unmute the output of rakudo subprocesses? 00:13
vrurg Xliff: tests? --verbose --debug 00:36
Just --debug, actually. 00:37
01:17 linkable6 left, evalable6 left 01:19 evalable6 joined 01:20 linkable6 joined 03:20 evalable6 left, linkable6 left, linkable6 joined 03:21 frost left 03:22 evalable6 joined 04:22 evalable6 left, linkable6 left 04:23 evalable6 joined, linkable6 joined 05:18 Geth left 06:05 reportable6 left 06:06 reportable6 joined
lizmat Files=1352, Tests=117183, 291 wallclock secs (35.34 usr 9.90 sys + 4058.11 cusr 331.18 csys = 4434.53 CPU) 07:39
07:54 frost joined 09:04 sena_kun left 09:07 sena_kun joined
lizmat I wish Azure could tell me WTF is wrong with PRs: github.com/rakudo/rakudo/pull/4855...5882849026 09:37
"PowerShell exited with code '1'." somehow does not give me enough information :-(
moon-child don't some cis let you ssh in and poke around when a test fails? 09:39
(though I guess you wouldn't get that on windows anyway...)
nine Geth down? 09:43
lizmat checks 09:52
09:52 Geth joined
lizmat should really spend some TLC on the Geth code 09:52
nine: any specific repo I should look at re-issuing push messages ? 09:53
nine rakudo/rakuast 09:54
lizmat what time approx?
nine Today starting 10:28:17 2022 +0200 09:55
Geth rakudo/rakuast: befb3026e8 | (Stefan Seifert)++ | 3 files
Fix EXPORT_SYMBOL not finding @*PACKAGES

Instead of having the package stack in a dynvar, in RakuAST, the resolver carries the package stack.
09:56
lizmat that one?
nine also fe78b5c689789d51dbcc660c55e5c69f0013ca5f 09:57
Geth rakudo/rakuast: fe78b5c689 | (Stefan Seifert)++ | src/Raku/ast/code.rakumod
Fix getcodename requires a concrete object when applying traits

The export trait wants to ask the Routine for its name, but this requires $!do to be initialized. Since traits are run at BEGIN time, we don't have the proper routine body in place yet, so instead add a stub carrying the name.
10:00 Geth left, Geth joined 10:02 RakuIRCLogger joined 10:50 Altai-man joined
lizmat so what is this mythical NULL.c setting ? /me has been unable to find a way to have an attempt at loading 11:19
[Tux] Rakudo v2022.03-161-g86927cba5 (v6.d) on MoarVM 2022.03-15-g43e1b00a6
csv-ip5xs0.787 - 0.810
csv-ip5xs-205.114 - 5.221
csv-parser3.526 - 4.880
csv-test-xs-200.394 - 0.430
test7.084 - 7.190
test-t1.456 - 1.478
test-t --race0.829 - 0.896
test-t-2020.376 - 21.300
test-t-20 --race6.866 - 10.190
11:34
12:05 reportable6 left 12:07 reportable6 joined 12:17 frost left
Geth DBIish/rbt.mysql-options: 13f0db1820 | (Humberto Massa)++ (committed by Rod Taylor) | 3 files
Add options to MySQL/MariaDB connections

Author: Humberto Massa [email@hidden.address]
  :$connection-timeout
  :$read-timeout
  :$write-timeout
Author: Humberto Massa <humbertomassa@gmail.com>
12:25
DBIish/rbt.mysql-options: 123936f57d | (Humberto Massa)++ (committed by Rod Taylor) | 3 files
Add options to MySQL/MariaDB connections

  :$connection-timeout
  :$read-timeout
  :$write-timeout
Author: Humberto Massa Minor adjustments: Rod Taylor
12:27
DBIish/rbt.mysql-options: b2a6e37652 | (Humberto Massa)++ (committed by Rod Taylor) | 3 files
Add options to MySQL/MariaDB connections

  :$connection-timeout
  :$read-timeout
  :$write-timeout
Author: Humberto Massa Minor adjustments: Rod Taylor
12:30
DBIish/rbt.doc-fix: 51016a03f4 | (Rod Taylor)++ | README.md
Mention mariadb library lookup in docs
12:41
DBIish/main: b2a6e37652 | (Humberto Massa)++ (committed by Rod Taylor) | 3 files
Add options to MySQL/MariaDB connections

  :$connection-timeout
  :$read-timeout
  :$write-timeout
Author: Humberto Massa Minor adjustments: Rod Taylor
12:42
DBIish/rbt.mysql-connection-settings: 37 commits pushed by (Ancient Wizard)++, (Rod Taylor)++, (Vadim Belman)++, rbt++, (Humberto Massa)++
review: github.com/raku-community-modules/...a6e3765211
12:43
DBIish/rbt.doc-fix: 7e9f981d8d | (Rod Taylor)++ | README.md
Mention mariadb library lookup in docs

This was an oversight in ddfa8b5.
12:45
rakudo: 2cadc4d869 | (Elizabeth Mattijsen)++ | src/Perl6/ModuleLoader.nqp
Streamlining some module / setting loading logic

  - give bootstrap loading its own method. This was done for code
   clarity and the hope that the alternate path would be inlineable.
   The alternate path is apparently only used for loading modules
   from nqp, like "use QAST:from<NQP>".
  - if setting "d" is loaded already, then we don't need to check for
   previous settings, as these have then already been loaded as well
   when the "d" setting was loaded.
12:47
lizmat afk for a few hours&
MasterDuke lizmat: if you click through to see the build logs, you get dev.azure.com/Rakudo/rakudo/_build...a&l=83 `src\moar.h(44): fatal error C1083: Cannot open include file: 'stdatomic.h': No such file or directory` 12:52
Geth DBIish: massa++ created pull request #226:
Solve problems with NULL columns
13:07
13:14 frost joined 13:40 carlmasak joined 13:46 frost left
Geth DBIish/main: 7e9f981d8d | (Rod Taylor)++ | README.md
Mention mariadb library lookup in docs

This was an oversight in ddfa8b5.
13:58
rakudo/rakuast: b59ce45da7 | (Jonathan Worthington)++ | 3 files
RakuAST node for 2⁴⁵ postfixes

And support it in the RakuAST-based compiler.
14:10
DBIish/rbt.mysql-blob: 4 commits pushed by (Humberto Massa)++, (Rod Taylor)++ 14:58
DBIish/rbt.mysql-blob: 61f9aff816 | (Rod Taylor)++ | t/26-mysql-blob.t
Minor tidying of the file
14:59
DBIish/rbt.mysql-blob: a15ff9d011 | (Rod Taylor)++ | t/26-mysql-blob.t
WIP Blob fix

  .prepare.execute() and .execute() work differently. They shouldn't.
DBIish/rbt.mysql-blob: fef8b69033 | (Rod Taylor)++ | 2 files
Use newer mysql_exec_statement API whenever possible

There are a very small number of statements which fail under this API, such as LOCK TABLE. In nearly every other case it provides an improved interface for correctness. For consistency in things like NULL handling, which are tricky to fully test otherwise, always try this API and fallback if it fails with the specific 'This command is not supported in the prepared statement protocol yet' message.
DBIish/rbt.mysql-blob: 4d11ed0fcf | (Rod Taylor)++ | t/26-mysql-blob.t
Add additional NULL testing for blob column
15:00
DBIish/rbt.mysql-blob: 781fcd363d | (Rod Taylor)++ | 2 files
Use newer mysql_exec_statement API whenever possible

There are a very small number of statements which fail under this API, such as LOCK TABLE. In nearly every other case it provides an improved interface for correctness. For consistency in things like NULL handling, which are tricky to fully test otherwise, always try this API and fallback if it fails with the specific 'This command is not supported in the prepared statement protocol yet' message.
DBIish/rbt.mysql-blob: 4 commits pushed by (Rod Taylor)++ 15:07
DBIish/rbt.mysql-blob: 4 commits pushed by (Rod Taylor)++ 15:09
lizmat MasterDuke: from github.com/rakudo/rakudo/pull/4856 I don't see a link that gets me to the page you showed 15:11
Geth rakudo/rakuast: df9b0a2571 | (Jonathan Worthington)++ | 2 files
Fix topic var and placeholder signature semantics

For { :$^foo } we would end up generating an optional `$_` parameter instead of just binding the outer value into the block's `$_`. This was observable because the `$_` parameter that was mistakenly generated was optional but code gen'd before the required `$^foo`, which in turn blew up bytecode compilation. The previous approach failed to work due to timing: the implicit delcarations are needed earlier. Even if it wasn't for that timing issue, however, blocks should still get a `$_` bound to the outer scope's one.
15:16
DBIish/rbt.mysql-blob: dc42a21725 | (Rod Taylor)++ | 2 files
Use mysql_exec_statement API whenever possible

The list of statements supported by the prepared statement API includes most things:
  dev.mysql.com/doc/refman/8.0/en/sq...-permitted
Consistently using this API when possible provides consistency with NULL handling and potentially other divergences which we've not yet discovered.
A few, such as LOCK TABLE or LOAD TABLE or SHOW WARNINGS, are yet supported so fallback to the older query mechanism when required. If performance with this type of query is an issue, peaking at the first word might be possible rather than relying strictly on the try/catch.
15:21
DBIish/rbt.mysql-blob: 9baab1eee0 | (Rod Taylor)++ | 5 files
Improve MySQL error handling

The current addition is only the sqlstate, though other pieces such as the statement are available.
Model this off the PostgreSQL error handling.
16:05
DBIish/rbt.mysql-blob: 8d030cab79 | (Rod Taylor)++ | 2 files
Use mysql_exec_statement API whenever possible

The list of statements supported by the prepared statement API includes most things:
  dev.mysql.com/doc/refman/8.0/en/sq...-permitted
Consistently using this API when possible provides consistency with NULL handling and potentially other divergences which we've not yet discovered.
A few, such as LOCK TABLE or LOAD TABLE or SHOW WARNINGS, are yet supported so fallback to the older query mechanism when required. If performance with this type of query is an issue, peaking at the first word might be possible rather than relying strictly on the try/catch.
DBIish/rbt.mysql-blob: 174eefdac3 | (Rod Taylor)++ | 5 files
Improve MySQL error handling

The current addition is only the sqlstate, though other pieces such as the statement are available. Additional diagnostics information will be added in the future.
Model this off the PostgreSQL error handling.
16:07
DBIish/rbt.mysql-blob: 8d49f22fde | (Rod Taylor)++ | 2 files
Use mysql_exec_statement API whenever possible

The list of statements supported by the prepared statement API includes most things:
  dev.mysql.com/doc/refman/8.0/en/sq...-permitted
Consistently using this API when possible provides consistency with NULL handling and potentially other divergences which we've not yet discovered.
A few, such as LOCK TABLE or LOAD TABLE or SHOW WARNINGS, are yet supported so fallback to the older query mechanism when required. If performance with this type of query is an issue, peaking at the first word might be possible rather than relying strictly on the try/catch.
DBIish/rbt.mysql-blob: 4 commits pushed by (Rod Taylor)++ 16:33
rakudo: 4b8f674171 | (Elizabeth Mattijsen)++ | src/Perl6/ModuleLoader.nqp
Simplify stash_hash logic

No extra variables needed
16:43
DBIish/rbt.mysql-blob: e7435fba97 | (Rod Taylor)++ | 6 files
Improve MySQL error handling

The current addition is only the sqlstate, though other pieces such as the statement are available. Additional diagnostics information will be added in the future.
Model this off the PostgreSQL error handling.
16:55
DBIish/rbt.mysql-blob: a6557343ef | (Rod Taylor)++ | 3 files
Use mysql_exec_statement API whenever possible

The list of statements supported by the prepared statement API includes most things:
  dev.mysql.com/doc/refman/8.0/en/sq...-permitted
Consistently using this API when possible provides consistency with NULL handling and potentially other divergences which we've not yet discovered.
A few, such as LOCK TABLE or LOAD TABLE or SHOW WARNINGS, are yet supported so fallback to the older query mechanism when required. If performance with this type of query is an issue, peaking at the first word might be possible rather than relying strictly on the try/catch.
DBIish/rbt.mysql-blob: 002140e285 | (Rod Taylor)++ | Changes
Changes message for b2a6e37
rakudo: 93402f1813 | (Elizabeth Mattijsen)++ | src/Perl6/ModuleLoader.nqp
Remove some superstitious parentheses
16:58
rakudo/rakuast: 6a63ae3268 | (Jonathan Worthington)++ | src/Raku/ast/operator-properties.rakumod
Add some missing operator properties
17:11
rakudo/rakuast: 888ad5f305 | (Jonathan Worthington)++ | 4 files
Start to improve undeclared symbol reporting

  * Gather nodes during CHECK time that are marked as unresolved
  * Some basic mechanism for constructing exceptions (although will
   need more work to handle suggestions and to aggregate by line, so
   this API likely changes)
DBIish/rbt.mysql-blob: 0e2008806e | (Rod Taylor)++ | 6 files
Improve MySQL error handling

The current addition is only the sqlstate, though other pieces such as the statement are available. Additional diagnostics information will be added in the future.
Model this off the PostgreSQL error handling.
17:16
DBIish/rbt.mysql-blob: 502e1265df | (Rod Taylor)++ | 3 files
Use mysql_exec_statement API whenever possible

The list of statements supported by the prepared statement API includes most things:
  dev.mysql.com/doc/refman/8.0/en/sq...-permitted
Consistently using this API when possible provides consistency with NULL handling and potentially other divergences which we've not yet discovered.
A few, such as LOCK TABLE or LOAD TABLE or SHOW WARNINGS, are yet supported so fallback to the older query mechanism when required. If performance with this type of query is an issue, peaking at the first word might be possible rather than relying strictly on the try/catch.
DBIish/rbt.mysql-blob: ff5eea415e | (Rod Taylor)++ | Changes
Changes message for b2a6e37
17:17 Altai-man left 17:18 carlmasak left
Geth DBIish/main: 6 commits pushed by (Rod Taylor)++ 17:23
17:28 Xliff left 18:05 reportable6 left 18:07 reportable6 joined
Geth nqp/faster_sorted_keys: 9be5f71c07 | (Elizabeth Mattijsen)++ | src/core/Hash.nqp
Speed up "sorted_keys" sub for < 3 elements

  - check number of elements in hash before doing anything
  - if zero, don't do anything
  - if more than zero, fetch the *keys* instead of pairs
  - if 1 element, we're done
  - if 2 elements, swap if necessary
... (5 more lines)
18:33
nqp: lizmat++ created pull request #769:
Speed up "sorted_keys" sub for < 3 elements
18:59 sena_kun left 19:08 sena_kun joined 22:38 nebuchadnezzar left
Geth ¦ rakudo: lizmat self-assigned Zef stopped installing packages github.com/rakudo/rakudo/issues/4857 22:59
23:10 Xliff joined