Has anyone tried (or had any success) with any of the AI engines converting a Dashboard 1 to Dashboard 2 system ?
Craig
Short answer: don't do it.
Long answer: please don't do it.
Since the current AI offerings are just very clever autocomplete machines, they do not really have any awareness that they might add properties to the flow JSON , that should not be set, are depreciated, or combinations of properties that are just not right.
2 Likes
And
Don't migrate. Make new one. A better one.
6 Likes
OK you have convinced me !!
Roll the sleeves up and time to learn
Craig
I support that fully ... dont migrate !!
Its is a great learning experience and you get a chance to overhaul and improve your old flows.
I ended up changing almost every flow due to the new opportunities DB2 opened up
1 Like
I hope dashboard v1 remain active for a long time because v2 is very very difficult to set up, for me of course. In my home automation I have 14 dashboard and only one is v2 version, made with many tears and many, many, many attempts.
I will never succeed to convert the other dashboards with gauge, button, image, tabulator, switch, ecc. ecc.
Sorry for the outburst but I'm really angry/depressed for this new dashboard.
1 Like
Sorry to hear that @Giamma - what concerns and issues have you had? Dashboard 1.0 has officially been deprecated.
Alas there's still not a great deal of beautiful and inspiring v2 home dashboards on show in the forum. You are not the only one finding it hard to build a satisfactory dashboard.
2 Likes
@Giamma have you raised issues in the github repository specific to the problem areas?
Perhaps you could list your top 3 pain points that prevent you migrating?
While it is not identical to DB1, setup is pretty much the same (create a page, create a group, configure the widget!)
It's actually even less than that as Dashboard 2.0 automatically creates the pages, groups, etc. for you. You just have to drag on a widget
That is/was same for DB1 until the understanding came that the beauty of the thing is hidden behind the CSS.
It's same for DB2. What is a great improvement is the responsive layout which you need to get used to maybe but simplifies your life a lot if you understand what you can build on top of it.
Yes, the design of core widgets may or not suit for you. But wasn't it same for DB1? And what did we do then? We did build one.
Well I have shared a lot of stuff how to achieve something if you don't like the defaults.
I'm still here. But not much of talk about the topic cos ... yeah.
1 Like
I too agree to the most points what the OP said, however, i kept on pushing my boundaries of patience and skill, and converted one of my dashboard to D2.
am still running DB1 too, where i find it somehow 'simpler' and is still not broken, so not going to alter it, BUT parallely a completely new dashboard design is on the drawing board.
Wish I had some more ready to adapt templates in the forum, even if we find some, there is a lot to modify and hence the resistance.
So I have now DB1, DB2 and UIBUILDER dashboards running simultaneously!!
1 Like
In the next few days I'll try to convert one dashboard v1 in v2 version and I will explain where I am in trouble ....
1 Like
I agree @hotNipi , the responsive layout[s] are potentially a big improvement. They also introduce the complexity of how individual widgets and the dashboard as a whole interact on different screen sizes. CSS styling to beautify individual widgets can not contribute much here.
We do see examples where people have used a template with their own HTML in preference to a standard widget. I wonder why - is it because folk find the widgets too difficult to use?
A dashboard page offers 4 different "layouts" and there is the mysterious "theme" option which has a profound effect on page appearance.
I think there should be more guidance/examples available for each option and when it is appropriate.
Perhaps I'll try and put together a demo flow...
I just wonder when will be the time when somebody starts building the dashboard for himself correct way.
- Map your needs.
- Draw it with pen and paper.
- Analyse outcome.
- Reiterate. (may take more than one go)
- Build with basics
- Tweak where needed
- Enjoy
Wow - i think i stirred up a bit of a hornets nest here !
My 2c worth - i cam to Node Red from the Arduino environment - i had heaps of sensors and relays etc as part of my (mainly) home heating and cooling system - and the interface to all of this was LCD panels the mighty 4 x 40 panels that were easy to tie into Arduinos.
I am a networking and security guy and have almost Zero design abilities and less than zero aesthetic ability.
So when i discovered Node Red and how easy it was to build a simple (but ugly) UI by dragging and dropping stuff around i was amazed and was up and running very quickly. Over time i used more of the fancier widgets as the developers and community (HotNiPi and Bart are two that stand out to me) brought them up - but they were never works of art after i had finished with them - but were functional for what i wanted - and the WAF was reasonable also when displayed on tablets around the house.
I went onto the Freelancer platform a few years ago and for $50 managed to get someone to smarten up what i had and set it up with different coloured icons based on values etc - which to me was a huge step up - i coudl read what they had done and understand it - and even over time have been able to change it and manipulate it - without really ever coming to total grips with it.
Each time i try and make a start on DB2 is feels overwhelming - as it feels that there is a LOT more background UI knowledge that is needed/assumed to achieve something reasonable.
Obviously some of that is because it is new and some of that is because the community has asked for new features from @joepavitt on an ongoing basis.
I wonder how many newbies coming to Node Red we are losing though because this first step feels too big ?
Craig
2 Likes
Really appreciate the candid feedback Craig. Would love to know examples here, the primary workflow is identical across the two, so I'm mot sure what we've made worse?
The only thing I can think of, is the shortage of third-party widgets on top of the core collection?
Not many I would say. Any "newbie" will consider DB2 as the dashboard for Node-RED, they would not know the difference between DB1 and DB2. Sorry for the nitpicking 
The win - for most newbies - would outweigh the effort since they have a need: I want to build a dashboard with dials and knobs for my home automation. That is still achieved faster using DB2 than setting up a secondary environment using some other - probably SaaS - product which then accesses their internal network or some external MQTT broker etc.
Are there any good third party products that emulate what DB2 does? DB2 is not just displaying the data but also allowing interaction with the flows - this makes it difficult to use a third party solution.
Disclaimer: I've not used DB2, only DB1. I was personally really impressed by the ease with which I could setup a "admin" interface for my system (using DB1). Admin interfaces are used by company admins/customer-support/analysts to access the system and perform specific tasks required by their respective roles. These aren't customer facing interfaces.)
I've recently converted my D1 dashboards to D2. Yes it took time and there was a learning curve
But overall I'm very pleaded with the results
A real bugbear on D1 was when looking at a screen for a particular period and then another user decides to select a different period and all users screens update. Useful in a class setting perhaps but not during the production meeting 
D2 much better in this regard
Tab layout in D2 is much better than groups in D1 and works well on all screen sizes unlike D1
Flows for charts are simplified as I can feed Database results directly into the widget without the need for a change node
May not be inspiring or beautiful but here are a couple of screens
D1
I tended to lump everything together on 1 page
D2
Tab layout allows for a more logical layout
5 Likes
I like your use of bar charts @gerrybartley, can I suggest you showcase this in the Show your projects category, for the benefit of others?
Perhaps make a simplified flow including:
An inject node simulating your database query output, reduced to show ~10 bars on the charts.
The three charts Module PPM, Module Performance and Module Status.
1 Like
