Some time ago I wrote an article about vRO 8.0 and first steps. Now with version 8.4 I decided to write a second article, checking if anything changed what I pointed out in my last article, or if there are some cool new features or new glimpses.
One of the really missing parts have been released already with 8.1 or 8.2: the tree view in the HTML5client is back. Nothing new, won’t talk about this topic.
First of all, and that’s based on the new mechanism, the most important troubleshooting script is/opt/scripts/deploy.sh. This redeploys all containers. Guys, don’t know what you feel about containerizing installed applications, but my personal opinion is, it’s not improving stability. I started using vRO with version 8.2 in our production environment, and I use the redeploy script at least five times a week because something broke in vRO. I hope, stability will be better in 8.4.
Another big issue which bit mea couple of times is the logout behavior. If the UI is idle for a certain period of time, you’ll be logged out. If you’re currently editing a workflow, all changes are lost. This is especially annoying if you wrote many lines of code in a scriptable task and the phone rings. If you get logged out all the lines of code are lost. I suggest to not pick up the phone. Hopefully this will change, either having a “no logout timer” by default or maybe an auto-save feature. Being honest, I don’t know what the best solution would be. 8.4 still has an idle time logout, but the time interval seems to have increased.
Found one major improvement: variables in a workflow can now be sorted ascending or descending by name. Same for Inputs and Outputs. If you wonder why this is important: in the past, variables were sorted by the length of their names (descending) – at least my guess. If you had a workflow with many variables, best of luck finding the one needed.
Versions of workflows still need to be increased manually. If you are using packages to transfer content between independent vRO installations, that’s not ideal, because you need to think and not forget increasing the versions. If there would be an option to turn an automated version increment on, that would be fantastic.
Also objects in the API explorer are now sorted alphabetically instead of a random mess. Nevertheless, writing code in a scriptable task and using the API explorer is still a pain.The API explorer pane still can’t be resized and instead of selecting methods and properties of objects (was possible in the past but the description was empty), this selection is no longer possible, because in the scriptable task, an object in the API explorer is not expandable anymore.
API explorer inside a scriptable task looks different, but is still not very useful. That’s not thatbad, because the “real” API explorer is better.
Speaking off, let’s take a closer look:
API explorer can be selected from the main menu and opens in a new tab in the browser. Via the “Type” of the object you can navigate to the scripting object by clicking the link, makes sense.
But still, if you look at the attributes, it doesn’t make a lot of sense. All “attributes” are shown at the top, but at least in my world, those are called properties of an object. Furthermore, the type is shown, but not if it’s a single object or an array. That information was available in the previous version of the API explorer in the Java based client. Without this information, programming ends up being a huge guessing game.
If you scroll down you’ll the constructor and methods section. The constructor should be at the top of the page and for the methods it’s the same: the return type is shown but not if it’s a single object or an array.
Summary: the API explorer is still useless.
Arranging icons the schema pane is still weird, because adding additional elements sill moves them around not adding standardized spacings. I know, not everyone cares about that, but I love having my workflows look nice. If everything else would work, I could oversee this strange behavior.
One last interesting finding:now there is a “deleted items” entry in the main menu.
Sounds interesting but haven’t tested its functionality.
Latest next week I am going to test the most important functionality for me: I’ll verify if the policy runs are now stable or not. In 8.2 there is an issue that at one point in time, vRO still collects messages from RMQ but is not starting the workflow anymore, for whatever reason. I’ve heard that this has been fixed, but I will test myself.
Summarizing my first look into vRO 8.4: if it’s somehow possible, stay with 7.6 because the UI is still not really working. If you can’t, try using third party tools instead of the API explorer, because it’s just not working.
Thanks for reading and take care!!