summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
-rw-r--r--doc/forum/Fstab.mounted_could_call_swapon_when_activating_swap.mdwn3
-rw-r--r--doc/forum/Fstab.mounted_could_call_swapon_when_activating_swap/comment_1_8ab6b313c80486f8f87a5e13e830bfa9._comment20
-rw-r--r--doc/forum/Fstab.mounted_could_call_swapon_when_activating_swap/comment_2_773fc1441dd06e9dd41508bd800298eb._comment13
-rw-r--r--doc/forum/Fstab.mounted_could_call_swapon_when_activating_swap/comment_3_f48a6191c56bed41eda55436f0aa3e9c._comment15
-rw-r--r--doc/forum/Git.cloned_deletes_harmless_empty_directory.mdwn3
-rw-r--r--doc/forum/Git.cloned_deletes_harmless_empty_directory/comment_1_7cd0521c6d071b25852f8355f4f61f94._comment20
-rw-r--r--doc/forum/Git.cloned_deletes_harmless_empty_directory/comment_2_289f157f129511242d93beae76fd03a3._comment11
-rw-r--r--doc/todo/initial_spin_compile_failure_recovery.mdwn5
-rw-r--r--src/Propellor/Property/SiteSpecific/JoeySites.hs4
9 files changed, 92 insertions, 2 deletions
diff --git a/doc/forum/Fstab.mounted_could_call_swapon_when_activating_swap.mdwn b/doc/forum/Fstab.mounted_could_call_swapon_when_activating_swap.mdwn
new file mode 100644
index 00000000..61cd10cc
--- /dev/null
+++ b/doc/forum/Fstab.mounted_could_call_swapon_when_activating_swap.mdwn
@@ -0,0 +1,3 @@
+The mount command won't work when activating a swap partition/file, so we should call swapon instead.
+
+https://github.com/ArchiveTeam/glowing-computing-machine/tree/fstab-swap
diff --git a/doc/forum/Fstab.mounted_could_call_swapon_when_activating_swap/comment_1_8ab6b313c80486f8f87a5e13e830bfa9._comment b/doc/forum/Fstab.mounted_could_call_swapon_when_activating_swap/comment_1_8ab6b313c80486f8f87a5e13e830bfa9._comment
new file mode 100644
index 00000000..4a144df5
--- /dev/null
+++ b/doc/forum/Fstab.mounted_could_call_swapon_when_activating_swap/comment_1_8ab6b313c80486f8f87a5e13e830bfa9._comment
@@ -0,0 +1,20 @@
+[[!comment format=mdwn
+ username="joey"
+ subject="""comment 1"""
+ date="2017-04-05T02:17:00Z"
+ content="""
+This idea kind of makes sense, because swap partitions in /etc/fstab
+get swaponed at boot.
+
+But, the implementation doesn't take the types into account. The `mounted`
+property takes a FilePath for the mountpoint, but for swap that
+needs to be "none", which is not really a file-path. Also, the `fstabbed`
+property has a separate `SwapPartition` type, so making `mount` support
+swap partitions without using that type feels wrong.
+
+It might be simpler all round to treat swap partitions being able to
+be specified in /etc/fstab as a historical accident, which it kind of
+is (increasingly so, since eg systemd has other ways to accomplish
+that), and instead of shoehorning this into the `mounted` property,
+add a new `swaponed` property.
+"""]]
diff --git a/doc/forum/Fstab.mounted_could_call_swapon_when_activating_swap/comment_2_773fc1441dd06e9dd41508bd800298eb._comment b/doc/forum/Fstab.mounted_could_call_swapon_when_activating_swap/comment_2_773fc1441dd06e9dd41508bd800298eb._comment
new file mode 100644
index 00000000..62cabc0a
--- /dev/null
+++ b/doc/forum/Fstab.mounted_could_call_swapon_when_activating_swap/comment_2_773fc1441dd06e9dd41508bd800298eb._comment
@@ -0,0 +1,13 @@
+[[!comment format=mdwn
+ username="db48x@80bd751a72d5a80737e2f875342cf845629c7202"
+ nickname="db48x"
+ avatar="http://cdn.libravatar.org/avatar/ad2688127feb555a92154b16d8eeb5d3"
+ subject="comment 2"
+ date="2017-04-05T02:48:08Z"
+ content="""
+Yes, perhaps if it took an Option FilePath (am I saying this correctly in Haskellese?) it would be nicer.
+
+I don't mind much how it's structured; this was just the smallest obvious change, since it was failing to mount it. Perhaps breaking it up into smaller, more primitive, pieces would help. Fstab.mounted could = Fstab.fstabbed `onChange` Fstab.mounted, for instance, and then I could write Fstab.fstabbed `onChange` Swap.swapEnabled (oh, but Fstab.fstabbed already exists; I'm not using it because it replaces the whole file, which seems like an odd thing to do. Maybe call it Fstab.listed instead?).
+
+Also, for maximum irony I was just perusing your most recent dozen commits or so, and saw you enable Apt.serviceInstalledRunning \"swapspace\" on one of your machines. That's amazing; I had no idea it existed! I am re-evaluating all of my life choices now.
+"""]]
diff --git a/doc/forum/Fstab.mounted_could_call_swapon_when_activating_swap/comment_3_f48a6191c56bed41eda55436f0aa3e9c._comment b/doc/forum/Fstab.mounted_could_call_swapon_when_activating_swap/comment_3_f48a6191c56bed41eda55436f0aa3e9c._comment
new file mode 100644
index 00000000..95c69551
--- /dev/null
+++ b/doc/forum/Fstab.mounted_could_call_swapon_when_activating_swap/comment_3_f48a6191c56bed41eda55436f0aa3e9c._comment
@@ -0,0 +1,15 @@
+[[!comment format=mdwn
+ username="joey"
+ subject="""comment 3"""
+ date="2017-04-05T03:08:30Z"
+ content="""
+I like the idea of composing smaller properties to build the current
+property, and add flexability.
+
+Renaming the existing `fstabbed` would probably be too much bother.
+(Also, I think I picked that name because it kind of hints that the
+existing fstab does not come out alive.)
+
+(The swapspace package is great if you can eat the now tiny overhead of a
+swap file compared to a swap partition.)
+"""]]
diff --git a/doc/forum/Git.cloned_deletes_harmless_empty_directory.mdwn b/doc/forum/Git.cloned_deletes_harmless_empty_directory.mdwn
new file mode 100644
index 00000000..ce3c192c
--- /dev/null
+++ b/doc/forum/Git.cloned_deletes_harmless_empty_directory.mdwn
@@ -0,0 +1,3 @@
+In my case I have carefully set up the directory that I'm going to clone into with the correct group ownership and setgid permission, so that the cloned files will also have the correct ownership. This change just checks to see if the directory actually has anything in it before it deletes it.
+
+https://github.com/ArchiveTeam/glowing-computing-machine/tree/git-in-emtpy-directory
diff --git a/doc/forum/Git.cloned_deletes_harmless_empty_directory/comment_1_7cd0521c6d071b25852f8355f4f61f94._comment b/doc/forum/Git.cloned_deletes_harmless_empty_directory/comment_1_7cd0521c6d071b25852f8355f4f61f94._comment
new file mode 100644
index 00000000..91b403b0
--- /dev/null
+++ b/doc/forum/Git.cloned_deletes_harmless_empty_directory/comment_1_7cd0521c6d071b25852f8355f4f61f94._comment
@@ -0,0 +1,20 @@
+[[!comment format=mdwn
+ username="joey"
+ subject="""comment 1"""
+ date="2017-04-05T02:22:54Z"
+ content="""
+I am not entirely happy with this patch, because it seems that if
+Git.cloned took care to preserve permissions in this case, it could be
+argued that it should also preserve permissions when the directory already
+exists but has the wrong content. Or equally well argued that it should not
+preserve permissions, which might be a leftover from some past unwanted
+state.
+
+Is that really the best way to do it? You could instead say:
+
+ Git.cloned user repo dir Nothing
+ `onChange` recursiveSetGID user dir
+
+And then you just have to write a recursiveSetGID which would be a
+generally useful property.
+"""]]
diff --git a/doc/forum/Git.cloned_deletes_harmless_empty_directory/comment_2_289f157f129511242d93beae76fd03a3._comment b/doc/forum/Git.cloned_deletes_harmless_empty_directory/comment_2_289f157f129511242d93beae76fd03a3._comment
new file mode 100644
index 00000000..1a8c1447
--- /dev/null
+++ b/doc/forum/Git.cloned_deletes_harmless_empty_directory/comment_2_289f157f129511242d93beae76fd03a3._comment
@@ -0,0 +1,11 @@
+[[!comment format=mdwn
+ username="db48x@80bd751a72d5a80737e2f875342cf845629c7202"
+ nickname="db48x"
+ avatar="http://cdn.libravatar.org/avatar/ad2688127feb555a92154b16d8eeb5d3"
+ subject="comment 2"
+ date="2017-04-05T02:37:44Z"
+ content="""
+Yea, I guess that's a fair point about the other cases.
+
+It just seems inelegant to go back over all the files and fix up their permissions, when it could just have been set right to begin with.
+"""]]
diff --git a/doc/todo/initial_spin_compile_failure_recovery.mdwn b/doc/todo/initial_spin_compile_failure_recovery.mdwn
new file mode 100644
index 00000000..423b279c
--- /dev/null
+++ b/doc/todo/initial_spin_compile_failure_recovery.mdwn
@@ -0,0 +1,5 @@
+When initial propellor --spin host fails to compile propellor
+perhaps due to a ghc compatability bug, spinning again doesn't fix the
+problem. IIRC /usr/local/propellor has a git repo set up, but no remote
+set, and so the subsequent spin doesn't update it, since propellor is not
+running there to receive a git push into the repo. --[[Joey]]
diff --git a/src/Propellor/Property/SiteSpecific/JoeySites.hs b/src/Propellor/Property/SiteSpecific/JoeySites.hs
index 28246dfe..063a2eda 100644
--- a/src/Propellor/Property/SiteSpecific/JoeySites.hs
+++ b/src/Propellor/Property/SiteSpecific/JoeySites.hs
@@ -573,8 +573,8 @@ kiteMailServer = propertyList "kitenet.net mail server" $ props
, "# Enable postgrey."
, "smtpd_recipient_restrictions = permit_tls_clientcerts,permit_sasl_authenticated,,permit_mynetworks,reject_unauth_destination,check_policy_service inet:127.0.0.1:10023"
- , "# Enable spamass-milter, amavis-milter, opendkim"
- , "smtpd_milters = unix:/spamass/spamass.sock unix:amavis/amavis.sock inet:localhost:8891"
+ , "# Enable spamass-milter, amavis-milter (opendkim is not enabled because it causes mails forwarded from eg gmail to be rejected)"
+ , "smtpd_milters = unix:/spamass/spamass.sock unix:amavis/amavis.sock"
, "# opendkim is used for outgoing mail"
, "non_smtpd_milters = inet:localhost:8891"
, "milter_connect_macros = j {daemon_name} v {if_name} _"