From 34604a4a0e4cd251e8d27e4db5e15334748df74d Mon Sep 17 00:00:00 2001
From: "http://joeyh.name/" <http://joeyh.name/@web>
Date: Wed, 1 Jan 2014 19:56:32 +0000
Subject: Added a comment

---
 .../comment_1_3128b025d15d3af6e9c8725d4659cdfa._comment      | 12 ++++++++++++
 1 file changed, 12 insertions(+)
 create mode 100644 doc/bugs/assistant_unable_to_auth___40__windows__41__/comment_1_3128b025d15d3af6e9c8725d4659cdfa._comment

diff --git a/doc/bugs/assistant_unable_to_auth___40__windows__41__/comment_1_3128b025d15d3af6e9c8725d4659cdfa._comment b/doc/bugs/assistant_unable_to_auth___40__windows__41__/comment_1_3128b025d15d3af6e9c8725d4659cdfa._comment
new file mode 100644
index 000000000..66827bf8d
--- /dev/null
+++ b/doc/bugs/assistant_unable_to_auth___40__windows__41__/comment_1_3128b025d15d3af6e9c8725d4659cdfa._comment
@@ -0,0 +1,12 @@
+[[!comment format=mdwn
+ username="http://joeyh.name/"
+ ip="209.250.56.227"
+ subject="comment 1"
+ date="2014-01-01T19:56:32Z"
+ content="""
+I can reproduce this bug. Seems the problem is that sshTranscript is called with a string that is fed in on stdin. But I guess that on Windows, ssh password prompting is also done using stdin (on Linux it would use /dev/tty), and so this prevents it from prompting.
+
+I've fixed this for the remote ssh server setup in the webapp, which did not actually need to feed anything in on stdin. Confirmed ssh adding works!
+
+However, for rsync.net, the ssh pubkey is fed into dd using stdin, and so I was unable to fix it there yet. Since rsync.net does not support redirection to files in its shell, the only approach that would seem to work is to first download the authorized_keys file, add the line, and rsync in the new file. Until I get a chance to do that, I am leaving this bug report open.
+"""]]
-- 
cgit v1.2.3


From 051aded96c6df8375867bf54f01f7c36730d58b9 Mon Sep 17 00:00:00 2001
From: "http://joeyh.name/" <http://joeyh.name/@web>
Date: Wed, 1 Jan 2014 19:59:12 +0000
Subject: Added a comment

---
 .../comment_2_e0b64ab2d19f0b0afbd466aad22c4df9._comment           | 8 ++++++++
 1 file changed, 8 insertions(+)
 create mode 100644 doc/bugs/assistant_unable_to_auth___40__windows__41__/comment_2_e0b64ab2d19f0b0afbd466aad22c4df9._comment

diff --git a/doc/bugs/assistant_unable_to_auth___40__windows__41__/comment_2_e0b64ab2d19f0b0afbd466aad22c4df9._comment b/doc/bugs/assistant_unable_to_auth___40__windows__41__/comment_2_e0b64ab2d19f0b0afbd466aad22c4df9._comment
new file mode 100644
index 000000000..6998c2b2d
--- /dev/null
+++ b/doc/bugs/assistant_unable_to_auth___40__windows__41__/comment_2_e0b64ab2d19f0b0afbd466aad22c4df9._comment
@@ -0,0 +1,8 @@
+[[!comment format=mdwn
+ username="http://joeyh.name/"
+ ip="209.250.56.227"
+ subject="comment 2"
+ date="2014-01-01T19:59:11Z"
+ content="""
+Also, fixing rsync.net that way would mean two ssh connections to set up authorized_keys, so 2 password prompts. suboptimal..
+"""]]
-- 
cgit v1.2.3


From b4ee53fe1197dcef2868f74f016e659ae93af443 Mon Sep 17 00:00:00 2001
From: "http://joeyh.name/" <http://joeyh.name/@web>
Date: Wed, 1 Jan 2014 20:13:04 +0000
Subject: Added a comment

---
 .../comment_1_847ff393fe3a0227d61440f6be899907._comment        | 10 ++++++++++
 1 file changed, 10 insertions(+)
 create mode 100644 doc/bugs/assistant_eats_all_CPU/comment_1_847ff393fe3a0227d61440f6be899907._comment

diff --git a/doc/bugs/assistant_eats_all_CPU/comment_1_847ff393fe3a0227d61440f6be899907._comment b/doc/bugs/assistant_eats_all_CPU/comment_1_847ff393fe3a0227d61440f6be899907._comment
new file mode 100644
index 000000000..cd14d9084
--- /dev/null
+++ b/doc/bugs/assistant_eats_all_CPU/comment_1_847ff393fe3a0227d61440f6be899907._comment
@@ -0,0 +1,10 @@
+[[!comment format=mdwn
+ username="http://joeyh.name/"
+ ip="209.250.56.227"
+ subject="comment 1"
+ date="2014-01-01T20:13:04Z"
+ content="""
+This strongly resembles ghc RTS bugs such as <http://bugs.debian.org/677096> (except that one got fixed, and affected the non-threaded RTS, while this is using the threaded RTS).
+
+The assistant and the webapp are the same process. The only difference is whether a web browser is talking to it. Unless the other bug is regarding the CPU used by the web browser, that other bug is a duplicate of this one.
+"""]]
-- 
cgit v1.2.3