From Enno Tensing to ~sircmpwn/hare-dev
On 24/06/16 06:13:13, Lorenz (xha) wrote: > apart from the fact that this patch does not apply (because the > subjectprefix should've been "PATCH harec"), isn't /usr/ > the right PREFIX for linux and not /usr/local? so wouldn't > patching config.mk the right move? Both /usr and /usr/local can be the 'right' PREFIX for linux. For BINDIR hier(7) says for /usr/bin that it is the primary directory for executable programs and /usr/local/bin is for binaries local to the site. For most distros that usually means that the package manager installs to /usr and manual-installations (e.g. make install) use /usr/local. However not every system will have all required directories --- though
From Enno Tensing to ~sircmpwn/hare-dev
Currently the blog links to index.xml on harelang.org, not on harelang.org/blog. This results in all pages on harelang.org being included, which is not the desired behaviour of a blog's rss-feed. To remedy this, change index.xml to {{.RelPermalink}}/index.xml. Signed-off-by: Enno Tensing <tenno@suij.in> --- layouts/blog/section.html | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/layouts/blog/section.html b/layouts/blog/section.html index ce76930..d373a33 100644 --- a/layouts/blog/section.html [message trimmed]
From Enno Tensing to ~sircmpwn/hare-users
On 24/02/09 17:46:27, Ember Sawady wrote: > On Fri Feb 9, 2024 at 5:44 PM UTC, Enno Tensing wrote: > > Can confirm that this happens with the link given on harelang.org/blog > > (which is harelang.org/index.xml). That seems to be an error? It should > > point to harelang.org/blog/index.xml (which behaves as it should). > > mind sending a patch? Will do.
From Enno Tensing to ~sircmpwn/hare-users
Hi, On 24/02/09 17:13:08, nirogu@vivaldi.net wrote: > Hi everyone. > > I'm trying to read the harelang blog using my RSS reader and I noticed > this weird bug. > > It seems that integrating the RSS blog downloads every subpage in the > harelang.org site, instead of only the blog entries. Can confirm that this happens with the link given on harelang.org/blog (which is harelang.org/index.xml). That seems to be an error? It should point to harelang.org/blog/index.xml (which behaves as it should).
From Enno Tensing to ~sircmpwn/hare-dev
On 24/02/05 17:32:25, Ember Sawady wrote: > we can decide whether to make modules without hare sources illegal to > import later, but the simplest thing to do is to just make them work as > you'd expect I'd either vote for a warning ('imported module xyz is empty') or just exposing all submodules with the parent as prefix? Anyway... Builds and fixes the issue, without having to do shady stuff with ctx.tests :) Thanks for investigating!
From Enno Tensing to ~sircmpwn/hare-users
Hi Dmitry, On 24/02/02 17:08:51, Dmitry B wrote: > So I have two question: > 1. Is Valgrind works with Hare? I do not know how it works but it did > not showed unfreed memory in my hare app. Valgrind can profile hare-programs, if they've been linked with libc. > 2. Is it enough to log size of every memory allocated in heap to get > an understanding how much app use RAM? There are tools, like memusage, that can give you that information, though I don't know if that is enough to get the full picture.
From Enno Tensing to ~sircmpwn/hare-dev
If mod.srcs.only_readme is set, the current module is empty and hare should not continue with the compilation. However with only that check is present 'make check' fails, with the message 'module hare is empty'. To prevent this, don't check for empty modules if ctx.test is true. Signed-off-by: Enno Tensing <tenno@suij.in> --- cmd/hare/build/queue.ha | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/cmd/hare/build/queue.ha b/cmd/hare/build/queue.ha index 56696672..c2a6ec0c 100644 --- a/cmd/hare/build/queue.ha [message trimmed]
From Enno Tensing to ~sircmpwn/hare-dev
Add new bool-field 'only_readme' to srcset, which is set to true if a module only contains a readme. Signed-off-by: Enno Tensing <tenno@suij.in> --- hare/module/srcs.ha | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/hare/module/srcs.ha b/hare/module/srcs.ha index 6ac3f855..7e5e3dff 100644 --- a/hare/module/srcs.ha +++ b/hare/module/srcs.ha @@ -38,6 +38,8 @@ export type srcset = struct { o: []str, [message trimmed]
From Enno Tensing to ~sircmpwn/hare-dev
Add an error that indicates that a module provides no sources, either because it is faulty or only provides submodules (like encoding does). Signed-off-by: Enno Tensing <tenno@suij.in> --- hare/module/types.ha | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/hare/module/types.ha b/hare/module/types.ha index 8a146f58..96469b72 100644 --- a/hare/module/types.ha +++ b/hare/module/types.ha @@ -31,6 +31,9 @@ export type file_conflict = ![]str; // Context for another error. [message trimmed]
From Enno Tensing to ~sircmpwn/hare-dev
This is an attempt to fix the abort() when trying to compile an empty module (e.g. #909[1]). For this a new bool-field is added to module::srcset that is true, if a module only contains a readme. If such a module is then encountered during run_task() a module::empty_module error is returned. This solves the issue quite nicely, but I am somewhat unhappy with the fact that it needs to be disabled for 'hare test' (at least for the selftest with make check), otherwise it raises an error before even getting to the first test.