NEWS

v0.29.2
=======

Adds the ability to ban fields based on regular expression. Also improves
the replace transformer that only supported metadata fields, now it also
supports the data field.

Adds new data type support to cast transformer. Now you will be able to
transform between byte-arrays (blobs) and strings. 

v0.28.0
=======

Release date: 2024-09-27

Adds the ability to ban content with a regular expression. Also bumps some,
but not all modules.

Fixes an issue where transformers were not visible when running with -show.

PS: There were some github action issues that caused some minor history
issues which can be seen in the git log as we re-did the tags of v0.28.x

v0.27.2
=======

Release date: 2024-03-18

Very minor release to improve debugging by adding timestamps to
metric-related messages. Also a few library bumps for security advisories.

v0.27.1
=======

Release date: 2023-12-01

Minor, second attempt at fixing the ban transformer. Should no longer
cause out of bounds errors. Hopefully.

v0.27.0
=======

Release date: 2023-11-22

Two minor, but important bug fixes:

- Fixes an issue with out-of-bounds checking on the ban transformer.
- Make the SNMP trap sender iterate over all metrics in a container, not
  just the first.

Other than that, some more relevant documentation and examples.

v0.26.0
=======

Release date: 2023-10-12

At a glance: Improvements to the SNMP sender and security fixes.

This addresses:

- CVE-2023-37475 (avro issues)
- CVE-2023-39325 (HTTP/2 rapid reset)

The SNMP trap sender fixes included make the sender more usable. It allows
manually configuring snmpTrapOID among other things, needed to make the
sender actually send valid snmpTraps.

v0.25.1
=======

Release date: 2023-09-25

Allow Dictsplit to split on the top-object.

Previously Skogul could split::

	{
		"metrics": [
			{
				"metadata": { },
				"timestamp": { ... },
				"data": {
					"foo": {
						"eth0": { "key": 123 },
						"eth1": { "key": 567 }
					}
				}
			}
		]
	}

Into::

	{
		"metrics": [
			{
				"timestamp": { ... },
				"metadata": { "name": "eth0" },
				"data": { "key": 123 }
			},
			{
				"timestamp": { ... },
				"metadata": { "name": "eth1" },
				"data": { "key": 567 }
			}
		]
	}

but always needed a top key bellow "data" to refer to. Now you can also
split the following original::

	{
		"metrics": [
			{
				"metadata": { },
				"timestamp": { ... },
				"data": {
					"eth0": { "key": 123 },
					"eth1": { "key": 567 }
				}
			}
		]
	}

To do this, simply omit the Field option (or leave it as an empty array).

E.g.::

	"transformers": {
		"type": "dictsplit",
		"metadataname": "name"
	}

As before, "Metadataname" will be the metadata key where the dictionary
index is stored.

Also, some printf-debugging was accidentally included in v0.25.0, this is
removed.

v0.25.0
=======

Release date: 2023-08-21

Skogul, now with more blobs.

The blob encoder and "parser" introduces a way of shuffeling arbitrary data
through Skogul without actually parsing it. By using the blob "parser",
Skogul will simply add the input data to a generic "data" key as a byte
array. This can then be forwarded and handled with any regular Skogul
mechanism. The regular JSON encoding will actually base64-encode byte
arrays, so you can forward it just fine like that.

To unpack the data, you can use the blob encoder, which should return the
data to the original unparsed state. Since you may have introduced a batch
sender and put multiple metrics into a container, the blob encoder has an
optional delimiter which will be injected in between metrics.

The TCP sender expands upon the net-sender. Technically, the net-sender can
do TCP just fine, but it is extremely inefficient at scale, because it will
open a new TCP connection for each container sent, instead of re-using it.
The TCP sender tries to address this and has been tested. It is still not
perfect in error-handling, so it should not be used if you need a guarantee
that messages are delivered, but it is actually reasonably performant.

Putting this together you can receive syslog messages on a UDP port and
forward it with the TCP sender without either side knowing there was
something like Skogul in between, and thus making sure rsyslogd doesn't get
backed up and blocks due to a slow receiving end and full buffers. NOT THAT
THAT EVER HAPPENS.

docs/examples/misc/tcp.json includes an example of two receivers using the
same handler, the blob encoder and tcp sender.

v0.24.0
=======

Release date: 2023-08-08

Adds two new, somewhat experimental modules: an SNMP trap sender and
rabbitmq sender and receiver. Both are expected to get some changes after
we get them tested, but of course any feedback is welcome.

v0.23.0
=======

Release date: 2023-06-06

Bump Juniper telemetry protocolbuffers from Junos 21.3R1 to 23.2R1.

Adds a small fix to the influx sender, preventing it from sending
empty tag strings to Influx. Fixes #294.

Adds support for renaming metadata and data fields using the "rename"
key to the metadata/data transformer. Example::

	"transformers": {
		"ok": {
			"type": "data",
			"rename": [
				{ "source": "old_key_name", "destination": "new_key_name" }
			]
	}



v0.22.0
=======

Release date: 2023-05-09

Add JSON unmarshalling to SQL receiver, add unflatten to metadata
transformer as well, and allow banning with ban transformer by path.

v0.21.1
=======

Release date: 2023-02-27

Add SIGHUP-handling to file-sender.

Add JSON-casting to cast transformer.

Upgrade to Go 1.20, and bump various libs past security lol.

v0.21.0
=======
Release date: 2023-02-10

Add NATS sender and receiver (thanks Niklas Holmstedt!), add support for
"match any"/presence in switch-transformer and fix a minor issue where
blank string could never match.

v0.20.11
========
Release date: 2023-01-05

Releases v0.20.0 to v0.20.10 are all build-system/CI/CD-related tweaks and
basically irrelevant to end users.

The following is the release notes for v0.20.0:

This is a small, but important release that adds support for loadable
plugins, and also the LineFileAdvanced file reader that allows log
rotation-based file reading (e.g.: Move a file, send sigup, then parse it).

The plugin support is very much experimental, see docs/plugins.rst for
discussion. There's also an example-plugin at
https://github.com/telenornms/skogul-plugin-example which demonstrates how
to build an out-of-tree plugin.

v0.20.10
========

Release date: 2023-01-05

Releases v0.20.0 to v0.20.10 are all build-system/CI/CD-related tweaks and
basically irrelevant to end users. See v0.20.0 for actual notes.


v0.20.0
=======

Release date: 2023-01-05

This is a small, but important release that adds support for loadable
plugins, and also the LineFileAdvanced file reader that allows log
rotation-based file reading (e.g.: Move a file, send sigup, then parse it).

The plugin support is very much experimental, see docs/plugins.rst for
discussion. There's also an example-plugin at
https://github.com/telenornms/skogul-plugin-example which demonstrates how
to build an out-of-tree plugin.

v0.19.1
=======

Release date: 2022-11-15

This is an "emergency release" of sorts, that REMOVES SQLITE SUPPORT.

We need to nuke this because it introduces what turned out to be
unacceptable dependencies which can't be (easily) met on, e.g., CentOS 7.

This means Skogul 0.19.0 RPMS wont work on CentOS/RHEL 7, while Skogul
0.19.1 should.

Sorry for the inconvenience. We will re-visit SQLite when a more
appropriate library can be used.

v0.19.0
=======

Release date: 2022-11-14

Somewhat large release, but:

- Drone is on the fringe, so manual release. GH actions incoming.
- New SQLite sender
- New *Experimental* prometheus parser
- New unflatten transformer
- Cast transformer: Add casting of IPs to integers

Minor:

- Several minor chores from Håkon Solbjørg that increases consistency and
  correctness, with minor negative user-impact.
- Splunk: Send metadata as fields to spluk HEC
- Deperecate "Path" and use "File" instead for consistency (deprecation,
  old configs will keep working)
- Add KeepOriginal flag to the Data transformer to 
- Many log cleanups

All in all, this is a larger release than we'd like, but also largely
isolated features.

Since this will be a manual release, it might deviate SLIGHTLY from the
other releases, but should be mostly the same.

v0.18.0
=======

Release date: 2022-09-20

Adds USP protocol buffer parser and AVRO encoder/parser time support, adds
worksarounds for Juniper -Inf transceivers, adds dummy parser.

USP
---

The USP parser will parse a protocol buffer-encoded USP message contained
in a record, and use the payload as JSON. This might change if hardware
surfaces that use USP differently.


AVRO
----

The AVRO encoder/parser now handles time fields, using a minor workaround.
Shouldn't require any changes for users.


Juniper
-------

We now apply various hacks to fix #218, where juniper telemetry reports
-Infinity, which is impossible to encode as JSON. Intead, we silently
convert it to -40, horrifying any mathematicians in the room.


Dummy parser
------------

The dummy parser is meant as a developer-tool where we can save un-parsed
bytes to disk, thereby allowing us to acquire test data for development.
Not useful in any production environment.

v0.17.0
=======

Release date: 2022-08-24

In short:

- AVRO and GOB encoding and parsing
- SQL receiver
- Encoder-support in file sender
- Docker images!

Docker images - first release, so let's see how well this works :D

SQL receiver will periodically poll an SQL database for data, it is a bit
immature, but it should be solid. You write a query, the query needs a
field named 'time' (select foo as TIME ...) for the timestamp, and you also
select which columns are metadata - the rest will be data.

While GOB could theoretically improve performance, it's not possible in
practice because GOB isn't schemaless, but embeds a schema in the initial
packet. This means that GOB data depends on the receiver having the initial
packet, and there is no convenient way to achieve that except
re-initializing the encoder every time. The end result is that gob is no
better or worse then JSON in regards to performance, but, unlike JSON, it
can support every data type that Go supports, fully, without losing
accuracy.

The AVRO format is still missing support for timestamps, see #247, but
that's being worked on.

To use AVRO, you need to specify a schema. One is provided in
docs/examples/avro, which is a very generic schema. You can write a more
specific one as well, for your specific data, but so far there is no
support for time stamps. To be able to actually USE Avro today, you need to
add a transformer on the handler to add a timestamp.

v0.16.0
=======

Release date: 2022-05-28

TL;DR:

- Kafka sender and receiver
- Default configuration file moved to /etc/skogul/conf.d in systemd units
- Enricher transformer MUST be used together with enricher updater sender
- "Wholefile" receiver will read a file and treat it as a single container
- Much needed docs/examples cleanup
- Dummy timestamp transformer

If you are using an unaltered systemd unit file and our RPM packages, you
will have to move your configuration from /etc/skogul/default.json into
/etc/skogul/conf.d/ - the file name can be whatever, as long as it ends in
.json - Skogul will load every configuration file present.

This better allows dynamic configuration through, e.g., Ansible and is
what we've been using for a while.

The Kafka sender has one little quirk: Since Kafka allows batching messages
natively, it was decided that it is better for overall performance to
encode each metric in a container as an individual message, instead of
encoding the entire container as one message. This will better exploit the
nature of Kafka and make it easier to use the data, hopefully.

The enricher also went through a major simplification: It no longer has a
native loader function, and instead you have to configure an enricher
updater sender. The configuration becomes slightly more cumbersome, but the
end result is a much more elegant solution.

v0.15.2
=======

Release date: 2022-05-09

Adds plumbing for modular encoding, but more importantly, an enrichment
updater sender.

The sender "enrichmentupdater" and the corresponding changes to the
enrichment transformer will allow on-line updates to the enrichment
database using regular Skogul pipelines. You can load a JSON file at
start-up with an initial database of enrichment, then update that database
on the fly over an HTTP interface for example, if you set up a
receiver-pipeline that ends in an enrichment updater.

Both the enrichment updater and the enrichment transformer is STILL
experimental, but I'm feeling good about them, and they will be a major
addition to medium to large-scale Skogul deployments.

PS: I botched the NEWS in v0.15.1 so let's go lucky v0.15.2.

v0.15.0
=======

Release date: 2022-04-10

Add an experimental enrichment transformer, and a slightly less
experimental switch sender.

The enrichment transformer WILL change, but needs testing, so it is
included. It allows an operator to provide a JSON file to enrich metrics
based on metadata fields, e.g.: Add customer name to interface stats, or
add if description. To use it, see docs/examples/tester_to_stdout_enrich.json ,
you will need a JSON file that has a list of metadata fields to match, and
the data field will be the fields added *to metadata* (yes, this is
confusing, this is one of the reasons it's considered experimental).

The switch sender is simpler, but also probably a bit immature, but testing
is required. It takes a list of conditions and checks if a metric(!)
matches this, and if it matches a complete set, data will be sent to the
associated sender. Use case: Dupe sender sends data to your own influxdb,
and to a switch sender, the switch sender then selectively streams
customer-specific data to a customer.

v0.14.3
=======

Release date: 2022-04-01

Introduce float flattening to the cast transformer, which allows flattening
floating point numbers into "non-exponential" string representations. E.g.:
For serial numbers, instead of having them encoded as 3.141519e+X when sent
to influx as tags, they'll be encoded as 3141519.

Introduces much better request logging as well for the HTTP receiver.


v0.14.2
=======

Release date: 2022-03-22

Introduce IgnorePartialErrors in handlers to let a handler modify a
container by removing invalid metrics instead of dropping the entire
container.

Also introduces several small but important log-quality fixes.

v0.14.1
=======

Release date: 2022-03-22

Add Cast transformer, which allows specifying a shallow schema for input.
It can cast data and metadatafields to strings, integers and 64-bit floats.

It currently ignores missing fields, and for fields already of the correct
value, it "just works" with no added costs. For fields needing conversion
it will somewhat cumbersomely go through fmt.Sscanf(fmt.Sprintf(...),...),
which is somewhat costly, but also ensures that it can convert anything
that fmt knows how to handle reasonably well.

It is meant to use as a "we can't fix the source data", not as a "we don't
want to fix the source provider".

v0.14.0
=======

Release date: 2022-03-22

Updates juniper telemetry interface, including a change of protobuf
compiler. Tested on various routers.

Greatly improves error messages, specially for protobuf, this is the
beginning of a major revamp to use wrapped errors, which will hopefully
both reduce log spam, and make the log messages more useful for real life
trouble shooting. For protobuf, this means actually useful errors when
marshaling fails, including what router and the underlying errors. Next up:
Workaround for -Infinity.

Also adds http stats, courtesy of eiriktaa <eirikamdal@yahoo.com>.

Also switches rpm builds from CentOS 8 which no longer works at all, to
Rocky Linux 8. This should have no impact on end users, regardless of
platform.

PS: This is a verbatim copy of the entire v0.13.x debacle, since it failed
to build the release. There are 0 code changes between 0.13.0 and 0.14.0,
and this is just to get the actual NEWS section fixed and a clean start.

v0.13.2
=======

Release date: 2022-03-22

Only build fixes since 0.13.0 failed to check on our CI pipeline. See
v0.13.0 for actual news.

v0.13.1
=======

Release date: 2022-03-22

Only build fixes since 0.13.0 failed to check on our CI pipeline. See
v0.13.0 for actual news.

v0.13.0
=======

Release date: 2022-03-22

Updates juniper telemetry interface, including a change of protobuf
compiler. Tested on various routers.

Greatly improves error messages, specially for protobuf, this is the
beginning of a major revamp to use wrapped errors, which will hopefully
both reduce log spam, and make the log messages more useful for real life
trouble shooting. For protobuf, this means actually useful errors when
marshaling fails, including what router and the underlying errors. Next up:
Workaround for -Infinity.

Also adds http stats, courtesy of eiriktaa <eirikamdal@yahoo.com>.

v0.12.1
=======

Release date: 2021-12-10

Fixes a bug in the stats receiver where it would not run transformers before
sending data.

v0.12.0
=======

Release date: 2021-12-10

Adds a Stats receiver for shipping internal skogul stats. This is a bit more
detailed than the Stats _sender_, and will be further extended in the future.
For now there are stats for the UDP receiver, protobuf parser and HTTP sender.
Furthermore, in the future we will try to get rid of more log messages about
errors and rather use stats for that.

Fixes the Structured Data parser to support more formats, and tackle all
formats with the same efficiency and priority.

Adds a concept of skogul.Identity to identify a skogul module by name. The
name is fetched from the configuration, and is typically the key used to
define a module.

This release also adds CopyFromData, a transformer to copy a field from data
to metadata without removing it. The ExtractFromData transformer is marked as
obsolete and will be removed in the future. To achieve the same effect, use
CopyFromData togheter with a Remove transformer.

v0.11.0
=======

Release date: 2021-12-02

The Batch sender now has a concept of a burning metrics if the sender is
too slow. If a "Burner" sender is configured, the batch sender will
redirect metrics to the burner if the regular sender would otherwise block,
e.g.: If the backlog is too big. This is recommended for setups where data
is duplicated to multiple backends, such as to a staging and test
environment, where you don't want the staging and test environment to slow
down data replication to the production environment.

Additionally, the UDP Read buffer size can now be configured for the UDP
receiver, which can be very useful for bursty telemetry data.

On a side note, shout out to Lasse Haugen for inspiring the DictSplitter
included in v0.10.9.

v0.10.9
=======

Release date: 2021-12-01

Skogul now figures out if the configuration path is a file or a directory,
and will work with either using the same options (-f or -d). We will
probably deprecate one of the flags at some point in the future.

Add a new DictSplit transformer.

Improved language in documentation.

Make the null-sender configure itself automatically.

The Net sender now checks for required parameters.

When running tests, try to find a new port if the selected one was already in use.

v0.10.8
=======

Release date: 2021-11-15

Fixes a bug that would crash skogul if it failed to convert
a protocol buffer packet into the internal data format of
skogul. The underlying problem was an interesting value
for a numerical metric, which there are now tests added for.

v0.10.7
=======

Release date: 2021-05-07

Fixes the systemd(1) unit file so that it supports being enabled
and automagically started on boot. We all WantedBy this.

v0.10.6
=======

Release date: 2021-04-27

Adds even better support for escaped characters in the influx parser,
specifically if the values had previously been escaped.

v0.10.5
=======

Release date: 2021-04-27

Supports parsing escaped hex, decimal and unicode character sequences
in the influxdb parser. Previously this would not have the expected behaviour.


v0.10.4
=======

Release date: 2020-11-12

Adds support for specifying a metadata field for Source in the Splunk sender
or setting it as a constant.
Adds experimental support for specifying what loglevel to log a failure as
in the UDP receiver.

v0.10.3
=======

Release date: 2020-11-06

Adds support for renewing the Client ID in the MQTT client upon reconnects,
as reconnecting (too quickly) with the same client ID might make the broker
forcibly close the connection after having reconnected. Enable this option with
RenewClientID: true.

v0.10.2
=======

Release date: 2020-10-26

Makes UDP receiver packet size configurable, instead of being hard-coded to 9000.

This was the first contribution from someone outside our organisation, thanks a lot
to @moogacs for this feature!

v0.10.1
=======

Release date: 2020-10-16

Fixes a memory leak in the Structured Data parser. No other parts of
skogul were affected by this bug.

Other, small fixes:
- Remove a silly debug log line
- Inform the user if they supplied an invalid log level

v0.10.0
=======

Release date: 2020-10-14

This is a big one! The changes are mainly new and opt-in features, so things
should work "just as before".

Features:
- New sender: splunk. Can be used to send to Splunk HEC.
- New parser: Structured Data (RFC5424). Only the data, not a syslog.
- New transformer: parse. Can be used to apply a parser to a
                   specific field in the data.
- new feature: TLS Client Certificate support in HTTP sender and receivers.
- new feature: TLS Client Certificate SAN DNSName verification.

Fixes:
- UDP receiver has been buffed with more robust handling of requests as well
  as some more tests.

v0.9.0
======

Release date: 2020-09-25

Promotes parsers to modules, making the configurable. To keep simplicity,
parsers can be "auto-instantiated" without having to create a "parsers"
key in the configuration file. To configure a parser, a "parsers" section
has to be added and the appropriate configurations made.

The auto-instantiation feature has been added to the debug sender and the
template transformer as well, to reduce the amount of boilerplate required.

Breaking change: Previously, not supplying a parser in a handler was allowed,
and would default to use the json parser for skogul containers. This is no
longer the case, and a parser always has to be defined.

v0.8.1
======

Release date: 2020-09-22

Adds a parser for MNR data and adds some more logging to the file sender.

v0.8.0
======

Release date: 2020-09-04

Adds a file sender which supports writing to file.
Fixes the RPM build spec so that rpm will not overwrite the default config
file upon upgrading skogul.

v0.7.2
======

Release date: 2020-07-30

Fixes a bug in the influx parser so that it now supports '=' (equals sign)
in tag values without dropping the rest of the tags.

v0.7.1
======

Release date: 2020-07-03

Fixes a bug in the influx receiver so that it now supports '=' (equals sign)
in tag values. This is useful for command lines where arguments might be
-foo=bar.

v0.7.0
======

Release date: 2020-06-18

Configuration can now be read from multiple files by using the new `-d`
option instead of `-f` to refer to a configuration directory instead of a
configuration file.

All configuration files need to end with `.json`, and all are read. When
all configuration files are read, any reference is reviewed. This means
that you can have a set of standard senders in a configuration file and
refer to them from an other.

v0.6.2
======

Release date: 2020-04-03

Fixes a bug introduced in the influx-sender in v0.6.1 which would send any integers
as quoted string values, as in `key="123i"` instead of the expected `key=123i`.
Fixes a bug in the influx-receiver which would skip the final byte in a float
when parsing it.

v0.6.1
======

Release date: 2020-04-03

Fixes a bug in the influx-sender which didn't send integers to influx as <int>i,
which is what InfluxDB expects. Since this is a breaking change and requires the
measurement fields to be dropped/updated/altered, there's a feature flag added:
"ConvertIntToFloat". Enabling this will convert ALL numbers sent through that
sender to FLOATS. Don't do this unless you know you need this. The default value
is OFF.

v0.6.0
======

Release date: 2020-04-02

Introduces a parses for the InfluxDB line protocol. The parser is to be considered
in beta/testing. Activate it by configuring a handler with the "influx" parser.

v0.5.5
======

Release date: 2020-03-03

Fixes a bug with the build pipeline which would reuse a previously generated
skogul binary when archiving releases. The binary in question did not contain
the build flag for setting version number, so now we clean the directory before
building a release to make sure to rebuild it with the proper flags.

v0.5.4
======

Release date: 2020-03-03

Fixes a bug where a previous bugfix for generating builds with version numbers
wasn't in master and therefore wasn't present for future builds.

v0.5.3
======

Release date: 2020-03-03

Fixes a bug with -version which wouldn't work for prebuilt binaries/packages.

v0.5.2
======

Release date: 2020-03-02

Fixes a bug with the install RPMs which would error during install
and uninstall. The RPMs are now built on a system running yum, which
will make sure build dependencies are correctly satisfied. Also reworks the
build system using Make, and introduces a -version flag.

Refactors the internal Handler API to expose Parse, Transform and Send
(also TransformAndSend which transforms a container and then sends it,
and Handle which parses, transforms and sends a container). This should
make the API easier to understand and makes sure relevant procedures (
like Validate()) are run before e.g. Send()-ing things off.

Specific changes:

- Fixes a bug with the RPM building which would make it error
  during install and uninstall (reinstall). The package gets
  installed correctly, but not uninstalled. This bug has to be
  manually worked a round if any previous skogul RPM is
  installed, by manually uninstalling the version using rpm.
  (rpm -e --noscripts skogul*)
- Inverts the influxdb-sender field validation to white-list instead of
  black-list data types, which should fix #119.

v0.5.1
======

Release date: 2020-01-30

Add custom flattening separator, and a magic 'drop' separator to
remove any intermediate keys. Also adds some benchmarks for custom
parsers as well as cleaning up some package names for test files.

Specific changes:

- Support for custom flattening separator as well as dropping intermediate keys.

v0.5.0
======

Release date: 2020-01-22

Skogul 0.5.0 is a small release, but should make life much easier for
InfluxDB-users.

Specific changes:

- InfluxDB sender now detects invalid metrics instead of sending "garbage"
  to InfluxDB that fails - it will send what it can.
- InfluxDB sender has much better logging
- Split transformer now correctly transforms remainder of metrics in a
  container after encountering one it can't handle.
- RPM packages and systemd unit files now agree upon default config file
  path.

v0.4.1
======

Release date: 2020-01-06

Skogul 0.4.1 brings support for exposing MQTT topics in the metadata. Yup, that's it.

v0.4.0
======

Release date: 2019-12-04

Skogul 0.4.0 brings MQTT configuration up to speed with other receivers and senders.
A new transformer to set the timestamp is introduced, as well as a custom JSON parser
for use where the incoming JSON format does not conform to the one of Skogul.
Furthermore, new versions now ship with prebuilt RPMs as well as prebuilt archives.
We also run automated tests on all example configuration in ./docs/examples!

Specific changes:

- JSON syntax errors offsets correctly with tabs
- Support building RPMs and build RPMs for new releases
- Bring MQTT configuration up to speed (configuration updates required!)
- MQTT wildcard support (+ and #)
- A custom JSON parser, for use when the JSON is not already on Skogul format
- Timestamp transformer, used to set the timestamp of a metric from data

v0.3.0
======

Release date: 2019-11-18

Skogul 0.3.0 contains several changes aimed at project management, some to
improve core functionality, and some module improvements. It's not a huge
change since 0.2.0, and that's how we like it. As per usual, we aim for
high quality releases.

Specific changes:

- Use of CI and automated building of release tar-balls
- Cleanup and refresh of documentation
- There's now a single unified way of handling modules that dramatically
  reduces code duplication in core.
- Modules can now provide documentation for nested data structures, which
  is correctly exposed in manual pages.
- A conditional transformer (#71) was added that provides a mechanism to
  only execute a transformer if certain conditions are present.
- The "extract" feature of the metadata transformer no longer creates empty
  metadata keys, and will remove the original to avoid duplicating
  information.
- The logrus receiver replaces the log-receiver, allowing better control of
  log data. More work in this area is to be expected.
- The HTTP receiver can now apply different authorization rules for
  different paths.

The latter constitutes a bit of a change in configuration.