Samuel GAIST authored Oct 10, 2019 and Samuel GAIST committed Nov 15, 2019

The selection can be done for both remote and local
development.

Samuel GAIST authored Oct 10, 2019 and Samuel GAIST committed Nov 15, 2019

Now the user has the choice to configure a remote
or docker environment. This allows for local testing
and then select a suitable environment for the platform.

This content is not valid in the sense that it's mostly empty.

The names are now less confusing with regard to their usage.

Empty as in minimal "{}"

Algorithm are a special case since without api version
an additional question is asked therefore the test
provides that in the JSON

Empty as in "{}"

Empty as in "{}"

Empty as in "{}"

Empty as in "{}"

Empty as in "{}"

Empty as in "{}"

This allows to write tests more easily.

This will allow loading of unfinished experiments.

The editor tab will show an icon when dealing with
invalid asset content.

This also start the implementation for support "wrong"
experiments as in not yet fully configured.

It also now does updating of the IO mapping when selecting
a new algorithm.

This will allow to have a more accurate list of assets to
choose from.

This makes the tests better separated.

The prototype while conceptually valid doesn't contain valid
plotter. This commit fixes that by selecting a valid Plotter
automatically.

Related to #244

The prototype while conceptually valid doesn't contain valid
data. This commit fixes that by selecting a valid Dataformat
automatically.

Fixes #244

This implementation auto-fiels the experiment content whith
whatever it can find matches input/outputs in number.

It currently doesn't check for compatibility.

Part of #185

This allows to support edition of "not yet created" asset
like for example the experiment which uses a two steps
setup.

This new implementation also opens the newly created asset
for edition.

Fixes #226