Closed Hywan closed 8 years ago
I think it's a bit soon to start working on something like this without first speaking with the others and seeing some of the issues that are actually open ;)
Also, architecture wise, I disagree with some of the assumptions made in the code. Will elaborate later.
What's the public for this kind of dashboard ?
At the time it is really hard to deploy it for "clients" since it will require a new install of a11ym
for each set of rule that need to be audited. A developer does not really care about history to the point of needing a dashboard like this. This does not really advance us on the SaaS either mostly from the same reason than the "clients" ones.
@krtek4 Please, elaborate, because this is a very small program with no specific architecture or assumptions, so it's hard to figure out what's wrong ;-).
The public for this dashboard is everyone who would like to track the evolution of their audits/tests. However, I admit the README.md
could be enhanced in order to specifiy that a11ym
should run as a cron job.
I don't understand why it needs a new install per set of rule. When you've fixed all your issues, you just would like to track the progression or regression, and this dashboard helps by providing an overview. For the SaaS, it provides an easy to use and to read listing, which is much better than the actual solution with a tricky Apache configuration file.
This PR introduces the dashboard. Let's start with a screenshot:
The listing shows all the reports. The red link points to the report with a URL of the form:
/report/…
. The graph only shows the 8th last statistics.To generate the report, nothing changed:
To start the dashboard server:
And then browse to
127.0.0.1:8080
.We could add tooltips over the graph dots/circles but maybe in another PR.
Thoughts @krtek4 and @jeanmonod?