This fixes both a race condition where we could assign s.stop from one
goroutine and then read it from another without locking, and handles the
fact that listener may be nil at shutdown if we've had a bad
CommitConfiguration call in the meantime.
Every time a JSON object is returned in an HTTP response, the
appropriate header needs to be set and the object itself needs to be
encoded. Doing this in every function is repetitive and error prone
(getDBFile and postDBScan, for instance, never set any headers).
This adds a helper function to centralize the appropriate JSON response
handling.
This replaces the current 3072 bit RSA certificates with 384 bit ECDSA
certificates. The advantage is these certificates are smaller and
essentially instantaneous to generate. According to RFC4492 (ECC Cipher
Suites for TLS), Table 1: Comparable Key Sizes, ECC has comparable
strength to 3072 bit RSA at 283 bits - so we exceed that.
There is no compatibility issue with existing Syncthing code - this is
verified by the integration test ("h2" instance has the new
certificate).
There are browsers out there that don't understand ECC certificates yet,
although I think they're dying out. In the meantime, I've retained the
RSA code for the HTTPS certificate, but pulled it down to 2048 bits. I
don't think a higher security level there is motivated, is this matches
current industry standard for HTTPS certificates.
This implements a new debug/trace infrastructure based on a slightly
hacked up logger. Instead of the traditional "if debug { ... }" I've
rewritten the logger to have no-op Debugln and Debugf, unless debugging
has been enabled for a given "facility". The "facility" is just a
string, typically a package name.
This will be slightly slower than before; but not that much as it's
mostly a function call that returns immediately. For the cases where it
matters (the Debugln takes a hex.Dump() of something for example, and
it's not in a very occasional "if err != nil" branch) there is an
l.ShouldDebug(facility) that is fast enough to be used like the old "if
debug".
The point of all this is that we can now toggle debugging for the
various packages on and off at runtime. There's a new method
/rest/system/debug that can be POSTed a set of facilities to enable and
disable debug for, or GET from to get a list of facilities with
descriptions and their current debug status.
Similarly a /rest/system/log?since=... can grab the latest log entries,
up to 250 of them (hardcoded constant in main.go) plus the initial few.
Not implemented in this commit (but planned) is a simple debug GUI
available on /debug that shows the current log in an easily pasteable
format and has checkboxes to enable the various debug facilities.
The debug instructions to a user then becomes "visit this URL, check
these boxes, reproduce your problem, copy and paste the log". The actual
log viewer on the hypothetical /debug URL can poll regularly for new log
entries and this bypass the 250 line limit.
The existing STTRACE=foo variable is still obeyed and just sets the
start state of the system.
Not necessarily the easiest way to fix just this bug, but the root cause
was using the (at that point uninitialized) cfg variable, so it seemed
sensible to just get rid of it to avoid that kind of crap.
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 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 reset of all folders failed when there was no data for a given
folder, as it was not returned by db.ListFolders then. But we don't
really care about that, we can "reset" it anyway...
The API never got the first few events ("Starting" etc) as it subscribed
too late. Instead, set up a subscription for it early on. If the API is
configured not to run this is unnecessary but doesn't hurt very much.
By doing it init(), the monitor process also gets a subscription thing
running, which is unnecessary (and really confused me when seeing it in
the debug output).
The "Invalid" config attribute is retained for errors discovered during
config loading (empty path, duplicate ID). This can only be set or
cleared at config loading time.
Errors discovered during runtime (I/O problems, etc) are now in the
folder state instead. Changes to these are sent as any other folder
state change.
Replace the current mix of UpperCamelCase und lowerCamelCase with
consistent lowerCamelCase keys for the JSON API. Also adapt the frontend
so it works with the changed API.
Attention: this will break existing consumers of the API.
Integers are for numbers, enabling arithmetic like subtractions and for
loops without getting shot in the foot. Unsigneds are for bitfields.
- "int" for numbers that will always be laughably smaller than four
billion, and where we don't care about the serialization format.
- "int32" for numbers that will always be laughably smaller than four
billion, and will be serialized to four bytes.
- "int64" for numbers that may approach four billion or will be
serialized to eight bytes.
- "uint32" and "uint64" for bitfields, depending on required number of
bits and serialization format. Likewise "uint8" and "uint16", although
rare in this project since they don't exist in XDR.
- "int8", "int16" and plain "uint" are almost never useful.
Request to terminate currently ongoing downloads and jump to the bumped file
incoming in 3, 2, 1.
Also, has a slightly strange effect where we pop a job off the queue, but
the copyChannel is still busy and blocks, though it gets moved to the
progress slice in the jobqueue, and looks like it's in progress which it isn't
as it's waiting to be picked up from the copyChan.
As a result, the progress emitter doesn't register on the task, and hence the file
doesn't have a progress bar, but cannot be replaced by a bump.
I guess I can fix progress bar issue by moving the progressEmiter.Register just
before passing the file to the copyChan, but then we are back to the initial
problem of a file with a progress bar, but no progress happening as it's stuck
on write to copyChan
I checked if there is a way to check for channel writeability (before popping)
but got struck by lightning just for bringing the idea up in #go-nuts.
My ideal scenario would be to check if copyChan is writeable, pop job from the
queue and shove it down handleFile. This way jobs would stay in the queue while
they cannot be handled, meaning that the `Bump` could bring your file up higher.
Make sure we have a good random seed on the default RNG, that the
predictable RNG is clearly marked as such, that random strings are
actually the length requested, and that they contain a restricted set of
characters only.