tangled.org
Commits
Signed-off-by: Anirudh Oppiliappan <anirudh@tangled.org>
Signed-off-by: Anirudh Oppiliappan <anirudh@tangled.org>
Signed-off-by: Anirudh Oppiliappan <anirudh@tangled.org>
it fails on file:// clone
Signed-off-by: Seongmin Lee <git@boltless.me>
Signed-off-by: dawn <dawn@tangled.org>
Lewis: May this revision serve well! <did:plc:3fwecdnvtcscjnrx2p4n7alz>
Signed-off-by: Trey Bastian <hello@treybastian.com>
fixed reversing the env var names
Signed-off-by: Trey Bastian <hello@treybastian.com>
Signed-off-by: dawn <dawn@tangled.org>
Lewis: May this revision serve well! <did:plc:3fwecdnvtcscjnrx2p4n7alz>
Signed-off-by: Seongmin Lee <git@boltless.me>
Signed-off-by: Seongmin Lee <git@boltless.me>
Signed-off-by: dawn <dawn@tangled.org>
Signed-off-by: dawn <dawn@tangled.org>
Signed-off-by: Seongmin Lee <git@boltless.me>
Signed-off-by: dawn <dawn@tangled.org>
Signed-off-by: Anirudh Oppiliappan <anirudh@tangled.org>
Signed-off-by: dawn <dawn@tangled.org>
initCgroupParent skips the parent-process move at group "/", assuming
the real root cgroup, which is exempt from the no-internal-process
constraint. in a private cgroup namespace (podman and docker's default
on cgroup v2) "/" is a populated delegated NON-root cgroup instead, so
the first workflow fails: 'failed to write subtree controllers
[memory pids] to "/sys/fs/cgroup/cgroup.subtree_control"': EBUSY.
userspace cannot tell the two "/"s apart — cgroup namespaces
virtualize /proc/self/cgroup and mountinfo for the reader — but the
constraint can: it exempts only the real root. probe it by writing
+memory +pids to cgroup.subtree_control: succeeds at the real root
(the same write prepareCgroup does per microVM anyway), EBUSY at a
populated namespace root, where we then vacate exactly as on the
group != "/" path. a read-only cgroupfs now fails loudly at init
instead of mid-workflow.
verified in scratch podman containers: unpatched reproduces the EBUSY,
patched passes the new regression test and enforces memory.max.
Signed-off-by: Winter <winter@madoka.systems>
Signed-off-by: noriaki watanabe <nabeyang@gmail.com>
Signed-off-by: Seongmin Lee <git@boltless.me>
This still marks the unshown ones as "emailed" so we don't send them in
the next digest tick.
Signed-off-by: Anirudh Oppiliappan <anirudh@tangled.org>
It was still using HTMX based comment toggle which has been removed from
`64c94491`.
Signed-off-by: Seongmin Lee <git@boltless.me>
Signed-off-by: dawn <dawn@tangled.org>
Signed-off-by: dawn <dawn@tangled.org>
Signed-off-by: oppiliappan <me@oppi.li>
Signed-off-by: oppiliappan <me@oppi.li>
Signed-off-by: dawn <dawn@tangled.org>
Signed-off-by: dawn <dawn@tangled.org>
This was mixing multiple forms, breaking them with `required`
constraints from completely different form. Both resubmit button and run
CI button were broken due to required comment body textarea included via
hx-include.
Signed-off-by: Seongmin Lee <git@boltless.me>
This reverts commit f3543173498febdf964251e1c994b96073e3f06c.
Fixes #364
Allows users to manually change theme, useful for cases like
LibreWolf where it defaults to light-mode irrespective of system
settings (with resistFingerprinting enabled).
Stores theme as 'light', 'dark', or 'auto' in the database, and
dynamically adds/removes the 'dark' class from the HTML.
Signed-off-by: TheBananaPancake <did:plc:3ywraszv5eqhnlvzumwju5fa>
Signed-off-by: oppiliappan <me@oppi.li>
Signed-off-by: oppiliappan <me@oppi.li>
Signed-off-by: oppiliappan <me@oppi.li>
Signed-off-by: oppiliappan <me@oppi.li>
Signed-off-by: Seongmin Lee <git@boltless.me>
Signed-off-by: Anirudh Oppiliappan <anirudh@tangled.org>
Signed-off-by: Anirudh Oppiliappan <anirudh@tangled.org>
Signed-off-by: Anirudh Oppiliappan <anirudh@tangled.org>
Similar to follow records from previous commit, deduplicate stars.
Define a `deduped_stars` view to query deduplicated stars. Oldest star
wins so refreshing stars won't put repository to trending list.
Signed-off-by: Seongmin Lee <git@boltless.me>
From model layer, we expect follows to be unique per (did, subject). But
from PDS, there is no such constraint. So hide the rkey from model while
keeping duplicated records in DB.
We keep duplicated records in DB to delete them all at once on unfollow
Signed-off-by: Seongmin Lee <git@boltless.me>
passing untyped dict so `.HxSwapOob` will resolve to nil, not panic.
Signed-off-by: Seongmin Lee <git@boltless.me>
add `BaseParams` to `LoginParams`. We won't set it, but it will make
template to render without failing.
Signed-off-by: Seongmin Lee <git@boltless.me>
It's just style change which can be done by CSS.
Signed-off-by: Seongmin Lee <git@boltless.me>
Signed-off-by: Seongmin Lee <git@boltless.me>
Signed-off-by: Seongmin Lee <git@boltless.me>
Signed-off-by: Seongmin Lee <git@boltless.me>
Signed-off-by: Seongmin Lee <git@boltless.me>
Signed-off-by: oppiliappan <me@oppi.li>
was a minor regression from manual-runs
Signed-off-by: oppiliappan <me@oppi.li>
if a user skips onboarding at step 0, empty profile records were not
created.
Signed-off-by: oppiliappan <me@oppi.li>
both pages now have an "advanced" section for users to configure
spindles, the default selection is "No Spindle"
Signed-off-by: oppiliappan <me@oppi.li>
initCgroupParent skips the parent-process move at group "/", assuming
the real root cgroup, which is exempt from the no-internal-process
constraint. in a private cgroup namespace (podman and docker's default
on cgroup v2) "/" is a populated delegated NON-root cgroup instead, so
the first workflow fails: 'failed to write subtree controllers
[memory pids] to "/sys/fs/cgroup/cgroup.subtree_control"': EBUSY.
userspace cannot tell the two "/"s apart — cgroup namespaces
virtualize /proc/self/cgroup and mountinfo for the reader — but the
constraint can: it exempts only the real root. probe it by writing
+memory +pids to cgroup.subtree_control: succeeds at the real root
(the same write prepareCgroup does per microVM anyway), EBUSY at a
populated namespace root, where we then vacate exactly as on the
group != "/" path. a read-only cgroupfs now fails loudly at init
instead of mid-workflow.
verified in scratch podman containers: unpatched reproduces the EBUSY,
patched passes the new regression test and enforces memory.max.
Signed-off-by: Winter <winter@madoka.systems>
Fixes #364
Allows users to manually change theme, useful for cases like
LibreWolf where it defaults to light-mode irrespective of system
settings (with resistFingerprinting enabled).
Stores theme as 'light', 'dark', or 'auto' in the database, and
dynamically adds/removes the 'dark' class from the HTML.
Signed-off-by: TheBananaPancake <did:plc:3ywraszv5eqhnlvzumwju5fa>