otto-de / jlineup Goto Github PK

The webserver part of JLineup should have a nice overview of running and recent Jobs. At the moment, the edison status detail indicator is (mis)used for this.

First, thanks a lot for this tool!

I switched from the window-widths to device config recently and tried the device-config key like described in the docs (https://github.com/otto-de/jlineup/blob/master/docs/CONFIGURATION.md#device-config), but got an error for that. The key device worked.

Since two or three versions, the CLI jar increased in size by ~30 MB, please investigate.

Hallo,

momentan ist es nicht möglich den User-Agent in der jlineup.json zu setzen.
Wäre ggf. ein Feature Request?

Gruß
David

In der report.json fehlt momentan die Information ob ein Test fehlgeschlagen ist. Es wird zwar das difference ausgewiesen, dieses muss dann aber manuell wieder gegen das max-dif aus der lineup.json verglichen werden. Wäre es nicht Sinnvoll ein failure Status oder das max-diff auch in das report.json mit aufzunehmen?

Gruß
David

In the pom of 3.0.0-rc1 jlineup-core is not properly referenced.
Is:

<dependency>
    <groupId>de.otto</groupId>
    <artifactId>core</artifactId>
    <version>3.0.0-rc1</version>
    <scope>compile</scope>
</dependency>

Should:

<dependency>
    <groupId>de.otto</groupId>
    <artifactId>jlineup-core</artifactId>
    <version>3.0.0-rc1</version>
    <scope>compile</scope>
</dependency>

This leads to problems if you want to use jlineup in gradle:
Execution failed for task ':jlineupBefore'.

Could not resolve all files for configuration ':runtime'.
> Could not find de.otto:core:3.0.0-rc1.
Searched in the following locations:
https://repo1.maven.org/maven2/de/otto/core/3.0.0-rc1/core-3.0.0-rc1.pom
https://repo1.maven.org/maven2/de/otto/core/3.0.0-rc1/core-3.0.0-rc1.jar
Required by:
project : > de.otto:jlineup:3.0.0-rc1

It would be great to retrieve the 'before' run id after requesting the /runs endpoint either via body response or redirect in order to get the url to request for the 'after' phase

JLineup Web doesn't have any persistence besides the screenshots and the generated reports on disk. To make it possible to restart the server without loosing currently running or pending jobs, we need a persistence.

At the moment, errors and exceptions, that are thrown during a JLineup web run, go to the configured log on the web server. It would be nice to have at least the root error cause in the job status.

It would be great to see the browser version in order to see if the bug shown in a report depends on a certain version.

Hi
Could you create a readme.md file to help with the start and using the jlineup tool and some tips as to how would you use it and also test scenarios where some page interaction is required to go to the desired page like login or click or something else ??

thanks ....

It would be nice to call custom URLs before and / or after JLineup run steps. This would enable you to create and remove test data through custom controllers in your 'JLineupped' service.

Tracking issue for:

• https://github.com/otto-de/jlineup/security/code-scanning/1

Hi,

I just noticed a strange behavior that seems to be existing since 4.4.0. 4.3.4 works fine:

I build a service that runs two comparisions: one where the after step runs right after the before step (no issues), and one where the before step runs, then some software updates are triggered, and after the updates are finished (might take a few minutes), the after step is exectued. This works with 4.3.4, but since 4.4.0, the after step seems to not notice the before step. It just creates additional folders for the screenshots, instead of adding the after screenshots in the folders created by the before step, and the report.json only contains screenshot-after-file-name entries, screenshot-before-file-name is null.

Currently chrome profiles are not deleted after a jlineup run. We noticed that a profile may take up ~300MB of disk space, so that the disk runs full quite rapidly when you have many jlineup runs.

I have a comparision with a very small difference value (0.00036101398601399). When I use exactly that value for the max-diff config value, the check fails. Increasing it to something like 0.00037101398601399 works, so I wondered if there is some rounding done internally?