Refactoring to prepare the addition of CSV parsing rules. The parsing
rules will contain information about which columns to ingest or ignore.
This will be used to add the ability to upload files via HTTP post in
addition to the TcpIngestor.
The icons were made with Inkscape. This tool does not care
for file size. By rewriting the icons by hand I could reduce
the file size siginificantly. E.g. 6.1kb to 441 bytes for
cumulative-distribution-chart.svg.
Classpathes in Eclipse are different from classpathes in Gradle due to
Gradle's 'implementation' configuration which does not provide
dependency transitively in Gradle.
Remove the controller method that returned the VueJS index page.
Add resource handlers that redirect to the Angular application.
I added two implementations, but activated only one. At the moment I am
not sure which solution is the better. We keep both so that we can
easily switch if need arises.
Generating all images for a gallery view can take
long. It might not be obvious when the computation
is done, because there is no progress indicator.
Fixed by adding a progress bar.
You might want to abort the computation. E.g.
because you already found what you were looking
for or you want to change the parameters.
This can now be done by clicking 'Abort'.
Note, it does not abort the current computation.
In the previous changeset the code that determined
which axis the plots used was implemented as a
side effect of getting the Gnuplot definition of
an axis.
Changed that to an explit update call with simpler
logic.
We do not know which fields exist at compile time.
But it is a great help to have some pre-selected
fields in groupBy.
Solved by adding a configuration option.
Gnuplot does not handle long x-axis ticks very good.
It should know how wide the labels are and could adapt
the increment size accordingly, but it doesn't.
Fixed by explicitly defining the increment for x-axis
labels.
Making the filter panel thinner by 18%. To do this
we moved the date range next to the query box.
The thinner filter box gives us more width for the plot.
We generate CSV files with comma as separator.
When we write times with milli seconds, then
we use floating point numbers. Depending
on the locale those floating point numbers
may be written with comma instead of point.
If that happens, then the plots are messed up.
Fixed by enforcing the locale when formatting floats.
This makes it easier to use the mouse wheel
to zoom in. Without it you could zoom into
a region that had not data and then had to
use the date picker to change the date.
