Instead, make sure we do the check as part of CheckFolderHealth before
pulling, and individually per file to try to not run out of space at
that stage.
(The latter is far from fool proof as we may pull lots of stuff in
parallell, but it's worth a try.)
This makes it possible to run multiple instances on the same box, all
receiving local discovery packets. Tested on Mac, Windows, supposed to
work on at least Linux too. For Windows, there may be issues with XP and
earlier, but meh...
Either Angular or the browser sometimes returns cached repsonse header,
causing a flap between requests that return the new version and requests
that return the old one. Here, instead, we trust the actual data
returned by the uncached /rest/system/version call.
Resetting the timeout doesn't fully cut it, as it may timeout after we
got an event and be delivered later. This should fix it well enough for
the moment. https://github.com/golang/go/issues/11513
I figured we're missing out on being cool and awesome by not having an
alphabetically based release code name like the big guys. This commit
fixes that. I've unilaterally decided on a theme of "$metal $bug"
because metals are kind of cool, and bugs, well, ...
This fixes a corner case I discovered in the symlink branch, where we
unexpectedly succeed in "replacing" an entire non-empty directory tree
with a file or symlink. This happens when archiving is in use, as we
then just move the entire tree away into the archive. This is wrong as
we should just archive files and fail on non-empty dirs in all cases.
New handling first checks what the (old) thing is, and if it's a
directory or symlink just does the delete, otherwise does conflict
handling or archiving as appropriate.
This will decrease the risk of running out of file descriptors for the
database and other bad things, which could otherwise potentially happen
if we're serving lots of requests and scanning in parallel, etc.
Windows doesn't have a per process open file limit like Unix so we don't
need to worry about it there.
The number of copiers and pullers is set to default at config loading
time, but the new folder configuration doesn't pass through config
loading so we start up with 0 copiers and 0 pullers and hence get stuck.
I moved the default handling to the puller itself instead. I think this
way is also cleaner as we get to keep the 0 in the config and the puller
gets to decide the defaults on it's own.
- Move the Go files into script/ instead of random places
- Rewrite check-contrib.sh into check-authors.go and check-copyright.go
- Clean up build.sh a little bit
* v0.11:
Translations and docs update
Enable browser caching of static resources
Handle multiple case insensitivity prefixes in ignores (fixes#2134)
Make rescan available for unshared folders
Add timeout for peek (fixes#1035)
Fix TestReset when Syncthing shuts down too fast
Clarify password in integration tests
Properly rename config files during integration tests (fixes#1769)
This sends the Cache-Control header to allow caching of static resources,
and checks the If-Modified-Since header to allow browser to use the
cached resource on refresh.
* pr-2066:
Configurable home disk percentage, translations
Add minimum disk free percentage to GUI
Stop folder when running out of disk space (fixes#2057)
This sends the Cache-Control header to allow caching of static resources,
and checks the If-Modified-Since header to allow browser to use the
cached resource on refresh. Also fixes some paths that caused redirects
(core//foo -> core/foo)
The very deep tree structure didn't really aggree with me, sorry. This
makes the core module rather large, but on the other hand that just
highlights that it is rather large.
Without this, as soon as we'd touched 1000 files in the last minute
(which can happen), we got stuck doing cache cleaning all the time,
burning a lot of CPU time.
This captures the common pattern of writing to a temp file and moving it
to it's real name only if everything went well. It reduces the amount of
code in some places where we do this, but maybe not as much as I
expected because the upgrade thing is still a special snowflake...
commit d60fbce311
Author: Jacek Szafarkiewicz <szafar@linux.pl>
Date: Mon Jun 1 11:16:36 2015 +0200
Correct order of deb files
commit 3b2ecfcc45
Merge: f4daebbc23a601
Author: Jacek Szafarkiewicz <szafar@linux.pl>
Date: Mon Jun 1 11:15:06 2015 +0200
Merge github.com:syncthing/syncthing
Conflicts:
build.go
commit f4daebb851
Author: Jacek Szafarkiewicz <szafar@linux.pl>
Date: Tue May 26 12:58:25 2015 +0200
Add me to AUTHORS
commit 9e77f4bea0
Author: Jacek Szafarkiewicz <szafar@linux.pl>
Date: Tue May 26 12:57:40 2015 +0200
Add systemd files to deb packate
This captures the common pattern of writing to a temp file and moving it
to it's real name only if everything went well. It reduces the amount of
code in some places where we do this, but maybe not as much as I
expected because the upgrade thing is still a special snowflake...
Jakob Borg