FlexDash, quo vadis?

[tve]

Someone asked in another thread about the choice of dashboards and FlexDash came up, so I thought I might as well create a separate thread to talk about the future of FD.

My assessment is that given the release of a new official dashboard it's completely futile to be developing FlexDash in its current state. Even if FD were the best dashboard in the universe it doesn't really matter: the number of users that are even interesting in looking at alternatives to the official dashboard is infinitesimal and the potential of having 3rd party widgets contributed is nil.

I think this is a good moment to thank all of you who have contributed to FD with testing, ideas, or simple encouragement! I would not have gotten as far without you! It has been a lot of fun and a lot of good learning! Plus a lot of banging my head against the limitations and idiosyncrasies of Node-RED...

Looking at FD as is stands, I believe it innovated in a number of ways:

• using modern JS technologies on the browser side, incl. Vue and Vuetify
• dynamic reloading of widgets (not having to restart Node-RED just to test a change in a widget)
• proper editor for writing custom widgets within the Node-RED editor (the "FD custom" widget)
• dynamic arrays of widgets allowing a dynamic number of sensors or equivalent to be displayed
• support for subflows allowing the bundling of UI elements with simple subflows such that instantiating a subflow automatically adds the UI
• automatic caching of values such that a new browser that connects gets up-to-date values automatically
• pop-up overlays with widgets that allow drill-down type of data to be shown on a dashboard tab
• visual editing of the dashboard layout within the Node-RED editor

(and probably some other stuff I forget right now )

The area that I still find quite unsatisfactory is data drill down, meaning, how to present data in the dashboard and let the user drill down into details and jump to related data. For example, you may have an array of temperature values displayed and you want to show more information when the user clicks on one of these values. It is currently possible to pop-up a chart with the historical data using FD. It's even possible to do it such that only the user that clicked sees that pop-up. But it becomes difficult to go much beyond that. Arguably at some point it's no longer a dashboard and more of a general web app, but it would be nice to be able to continue using widgets and link them to flows and nodes in NR. Food for thought and innovation...

To wrap this up: FlexDash as it stands is not going away but it is unlikely to continue on a path trying to replace the functionality offered by the official dashboards. It is likely to be less tightly integrated into NR 'cause implementing that takes more effort than the dashboard itself and ends up limiting the dashboard due to the limitations in Node-RED.

I'd love to hear the feedback and suggestion you, FlexDash supporters and users, may have!

[smcgann99]

Hopefully some of your innovations will find their way into Dashboard 2.0 at some point.

[BartButenaers]

Hi Thorsten,

I didn't see this one coming a few months ago, until suddenly the new dashboard was announced...

But sadly you hitted the nail correctly a few times in a row above: contributions of custom widgets, overall contributions to your core, and so on...

To be honest I have a very bad feeling about this: first of all Flexdash is a very nice piece of software, which should have been used a lot, until VueJs would ever go end of life in the far future... And secondly because I have been been begging you continiously - at the time being - to support UI nodes in Flexdash. Because at that moment you were the only one with enough knowledge and energy to offer us something like that. So in fact I am quite responsible for consuming a few months of your limited life on this blue planet. I owe you my apologies for that

To be honest, when @joepavitt implemented last week a way to contribute custom UI nodes to dashboard 2.0, I thought to myself: I am very glad that I don't have enough free time right now to start migrating my UI nodes, because how ungrateful would that look towards Thorsten after all he did for me and this community. I had very mixed feelings about that...

So I really, really, really hope you at least had enough fun developing the Node-RED integration for Flexdash, so at least you didn't do it all for nothing.

When I look at your list of improvements above (compared to the old official dashboard), I also hope at least your brilliant innovative ideas will ever find their way to the new official dashboard. Then at least you didn't develop it all for nothing. It would be useful if some Flexdash users can register issues in the new dashboard Github repo, to explain which Flexdash functionality they would like to see arriving one day in the future. Not that Joe or others will be able to implement it all at once, but then at least it won't be forgotten...

And my last request for you . We had a lot of interesting (private) discussions about server side state management, to allow developers to implement ui nodes more easily. Because developing the svg-ui node was very difficult, due to the msg replay mechanisme in the old dashboard. If you have a bit of energy left, I would appreciate a lot if you could write down your innovative ideas in this issue. I tried to make some advertisement for your way of working, but my old brain has leaked most of the information you passed on to me. So if you still remember the concepts you had in mind (and the advantages and disadvantages), it would be very kind of you could write them down for discussion. That would really be helpful for me and other contributors! Then finally you can get rid of me in your life

To end my long story: an awesome thank you from Belgium
And I wish that Flexdash keeps growing outside of Node-RED!

Bart

BTW. Now I have to work hard the entire weekend for my daily job, to do a large complex migration. So (again) no Node-RED developments for me in the next days