Recent activity

Re: Windows support a month ago

From Никола Радојевић to ~sircmpwn/hare-users

This is not the place to talk about political regimes, if the project maintainers say they don't want to support non-free systems, that's how it's going to be 

Дана 24. август 2022. 18:32:43 CEST, Dmitry B <bubnenkoff@gmail.com> написа:
>There is no problem with free/non-free operating systems.
>There is a problem with free and non-free political regimes.
>Operating Systems don't kill people.
>If you want freedom, you'd better state your position on the war in Ukraine.
>The Russian military uses free Linux to kill people.
>
>ср, 17 авг. 2022 г. в 12:58, Drew DeVault <sir@cmpwn.com>:
>>
>> We will not reconsider support for non-free operating systems. Do not
>> ask again.

Re: Slicing silently discards const 3 months ago

From Никола Радојевић to ~sircmpwn/hare-users

Compiler doesn't properly check for const when generating code yet, it's a known issue. Not sure what would be the intended behavior here though 

Дана 17. јун 2022. 06:19:43 CEST, Julian Maingot <julianmaingot@gmail.com> написа:
>Hello,
>
>I was working through some toy examples and noticed that the following
>program compiles and runs without warnings or errors.
>
>export fn main() void = {
>	const arr: [_]int = [1,2,3];
>	let slc = arr[..len(arr)];
>	slc[0] = 5;
>	assert(arr[0] == 5);
>};

[PATCH hare] rt: Fix stack alignment in _start 3 months ago

From Nikola to ~sircmpwn/hare-dev

Until now, we assumed that the stack was 16 byte aligned by the OS when
the process was started. This is not guaranteed, so we should align the
stack ourselves.

This is a companion to the similar patch for harec.

Signed-off-by: Nikola <nikola@radojevic.rs>
---
 rt/+freebsd/start+aarch64-libc.s | 1 +
 rt/+freebsd/start+x86_64-libc.s  | 1 +
 rt/+linux/start+aarch64-libc.s   | 1 +
 rt/+linux/start+x86_64-libc.s    | 1 +
 4 files changed, 4 insertions(+)
[message trimmed]

[PATCH harec] rt: Fix stack alignment in _start 3 months ago

From Nikola to ~sircmpwn/hare-dev

Until now, we assumed that the stack was 16 byte aligned by the OS when
the process was started. This is not guaranteed, so we should align the
stack ourselves.

Fixes: https://todo.sr.ht/~sircmpwn/hare/656

Signed-off-by: Nikola <nikola@radojevic.rs>
---
 rt/+freebsd/start+aarch64.s | 1 +
 rt/+freebsd/start+x86_64.s  | 1 +
 rt/+linux/start+aarch64.s   | 1 +
 rt/+linux/start+x86_64.s    | 1 +
 rt/+openbsd/start+aarch64.s | 1 +
 rt/+openbsd/start+x86_64.s  | 1 +
[message trimmed]

[PATCH hare] math::complex: Fix failing test due to signed zeros 4 months ago

From Nikola to ~sircmpwn/hare-dev

This fixes the failing tests, but I am not sure if checking for
different signes of zeros is important somewhere else, comments
apprecieted.

The test failed at complex/+test.ha:1088.

Signed-off-by: Nikola <nikola@radojevic.rs>
---
math/complex/+test.ha | 3 +++
1 file changed, 3 insertions(+)

diff --git a/math/complex/+test.ha b/math/complex/+test.ha
index a1cc0916..51e6af71 100644
--- a/math/complex/+test.ha
[message trimmed]

[PATCH hare v2] driver: make progress bar width fixed 4 months ago

From Nikola to ~sircmpwn/hare-dev

Signed-off-by: Nikola <nikola@radojevic.rs>
---
 cmd/hare/plan.ha     | 1 +
 cmd/hare/progress.ha | 4 ++--
 cmd/hare/schedule.ha | 2 ++
 3 files changed, 5 insertions(+), 2 deletions(-)

diff --git a/cmd/hare/plan.ha b/cmd/hare/plan.ha
index abab6695..ae20a2f1 100644
--- a/cmd/hare/plan.ha
+++ b/cmd/hare/plan.ha
@@ -65,6 +65,7 @@ type plan_progress = struct {
	complete: size,
	total: size,
[message trimmed]

[PATCH hare] Build driver: The size of the progress bar is constant now. 4 months ago

From Nikola to ~sircmpwn/hare-dev

Fixes: https://todo.sr.ht/~sircmpwn/hare/685

---
 cmd/hare/plan.ha     | 1 +
 cmd/hare/progress.ha | 4 ++--
 cmd/hare/schedule.ha | 2 ++
 3 files changed, 5 insertions(+), 2 deletions(-)

diff --git a/cmd/hare/plan.ha b/cmd/hare/plan.ha
index abab6695..ae20a2f1 100644
--- a/cmd/hare/plan.ha
+++ b/cmd/hare/plan.ha
@@ -65,6 +65,7 @@ type plan_progress = struct {
	complete: size,
[message trimmed]

[PATCH hare] sort::search now returns an index to the element instead of a pointer 4 months ago

From Nikola to ~sircmpwn/hare-dev

Fixes: https://todo.sr.ht/~sircmpwn/hare/678

---
 fnmatch/fnmatch.ha |  6 +++---
 sort/+test.ha      |  6 +++---
 sort/search.ha     | 19 +++++++------------
 3 files changed, 13 insertions(+), 18 deletions(-)

diff --git a/fnmatch/fnmatch.ha b/fnmatch/fnmatch.ha
index 85d948f9..d0578c4f 100644
--- a/fnmatch/fnmatch.ha
+++ b/fnmatch/fnmatch.ha
@@ -347,10 +347,10 @@ fn ctype_name_to_func(name: str) nullable *fn(c: rune) bool = {
		("upper", &ascii::isupper), ("xdigit",&ascii::isxdigit),
[message trimmed]

[PATCH hare v2] Handling hash collisions by comparing both hashes and idents 4 months ago

From Nikola to ~sircmpwn/hare-dev

Fixes: https://todo.sr.ht/~sircmpwn/hare/679

---
 cmd/hare/schedule.ha | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/cmd/hare/schedule.ha b/cmd/hare/schedule.ha
index 514ebc4b..a4f4d119 100644
--- a/cmd/hare/schedule.ha
+++ b/cmd/hare/schedule.ha
@@ -29,7 +29,8 @@ fn sched_module(plan: *plan, ident: ast::ident, link: *[]*task) *task = {
	let hash = ident_hash(ident);
	let bucket = &plan.modmap[hash % len(plan.modmap)];
	for (let i = 0z; i < len(bucket); i += 1) {
[message trimmed]

[PATCH hare] Handling hash collisions by comparing both hashes and idents 4 months ago

From Nikola to ~sircmpwn/hare-dev

This is a fix for #679.

---
 cmd/hare/schedule.ha | 14 +++++++++++++-
 1 file changed, 13 insertions(+), 1 deletion(-)

diff --git a/cmd/hare/schedule.ha b/cmd/hare/schedule.ha
index 514ebc4b..eeefb10c 100644
--- a/cmd/hare/schedule.ha
+++ b/cmd/hare/schedule.ha
@@ -16,6 +16,17 @@ use path;
use strings;
use strio;

[message trimmed]