Using Icons - The facts, nothing but the facts

Configuration Frontend
21

Idealy, HA would store your used-elsewhere icons and offer up those as suggestions first. There would be a search text box linked to MDI’s database with the results selectable and automatically converted to the required lowercase syntax ready for insertion. Instead of seven, obscure, not at all obvious steps to add an icon, we would have, one, at most, two steps to icon nirvana…

Am I right to imagine a fluent coder could knock that off in a couple of hours tops? It certainly seems a simple undertaking to those unsure how such matters get done, in fairness

22

The recent appointment of a UX-designer to this infuriating but rewarding project should start to blur that line over the next year of updates. I wonder if his suggestion box has runneth over yet? One can well imagine :face_with_monocle: :face_with_hand_over_mouth:

1 Like
23

That wasn’t why the meltdown took place. The trigger was the utter lack of bozo-safe instructions for adding cute little pictures to my Lovelace UI. That “icon” text box had been staring promisingly, but unhelpfully at me from my first HA install attempt. When I went looking for simple instructions I found only jargon-laced assumptions about things I should know, apparently, but didn’t. The meltdown occurred when I couldn’t find intelligible explanations in a reasonable timeframe. That the confusion was over a question of text case is certainly embarrassing, but for whom should it be embarrassing? I am but a single noob, limited in experience with beta software glitches, in his late forties arm wrestling with an uncooperative UI. HA has a small army of programming wizards who can’t even make a noob-proof icon implementation.

This is typical of the HA experience and it needs to be rectified or surely more meltdowns will follow.

You may sneer from your tower of complete HA understanding, but we weren’t all watching its construction from the moment ground was broken, were we? This platform needs tidying up for all the come-latelies, which is most of the rest of us, eventually, btw or it will remain a product for semi-professional use only, rather than what it promises to be. The solution to uncooperative corporate skullduggery and the fulfillment of the IoT’s initial promise of everything from everywhere talking amongst themselves without the cracking of skulls along the way.

Don’t you agree?

1 Like
24

Check this out.

In configuration.yaml


panel_iframe:
  mdiindex:
    title: "MDI Icon"
    icon: "mdi:vector-square"
    url: "https://cdn.rawgit.com/james-fry/home-assistant-mdi/efd95d7a/home-assistant-mdi.html"
  iconindex:
    title: "More Icons"
    icon: "mdi:share"
    url: "https://mdi.bessarabov.com/"

This will add a couple icon sites your sidebar. With the second one both of them having the proper format ready to copy.

Not all icons on there, but i find almost everyone I need is there

1 Like
25

Well then, aren’t you a splendid chap? Cheers for that. I shall rejoice suitably for all the Johnnie-come-latelies when this has been made redundant in the UI.

2 Likes
26

Oh, I do hope you’re right!!!

It’s little frustrations like this which demonstrate a larger issue with HA, which Matthew_Malpeli has explained above far more eloquently than I could.

The whole thing feels like an adolescent attempt to prove that the “in crowd” is so much cooler than the rest of you mugs. We have our own secret codes, our own jargon, and if you’re not in the club we’re not going to share them with you.

I see the UX person going one of two ways. They can come to HA with a fresh perspective and call out this BS, or they can try to impress us all, and build up their stature in the design community, with a cutting-edge, shocking and incomprehensible new design.

I’m not a betting man, so I won’t lay odds on which way it goes. But I think it’s clear what I’m hoping doesn’t happen.

2 Likes
27

I am far from having a complete understanding of HA, I’d still put myself in the rookie category actually and let’s be fair to home assistant it’s got placeholder/greyed-out text in the exact format that you need to write it in under a very clear “icon” heading.

I agree with you that ultimately having mdi icons as you have suggested would be the perfect situation. I have spent hours frustratingly designing custom icons and then deleting them because the home assistant system isn’t capable of comfortably handling anything that isn’t mdi (even though there’s lots of SVG icon frameworks out there). There is a lot of UI and other stuff that could be improved to make it more user friendly.

But right now I am absolutely thrilled that massive functional updates like the recent energy one are held to a higher importance that further simplifying what is, frankly, a very simple task. I’m sorry you struggled, but honestly you only have your own frustration blinding you to blame.

28

You should understand that you’re coming into a 10 year old product and only 2.5 of the years have been shifting towards UI. The other years were 100% yaml. Your demographic was not the target audience. The target audience has been shifting in recent times but…

If you have this notion about everything being in the UI, then maybe HA is not for you right now. Come back in 3 more years, but for now it’s in a transition and this is what we have to work with.

Getting upset when someone gives you an answer is not how you should handle a forum. This is an open source package, no one here is paid to support people. You will not get nice PR friendly responses. You’re going to get straight answers with no fluff.

If you read the FAQ, the first point states that this is not a help desk. There’s a reason this is posted. Because there is no help desk. There is a group of people answering questions. Yes, some of those people are jerks. Yes, some of those people do not know how to respond without sounding condescending. These are things you and other users will have to deal with. Until there is a paid individual supporting the community, don’t expect PR responses. If you feel wronged by a post or it doesn’t meet the code of conduct, simply flag the post and move on.

Even with a UX designer, there are 1900 integrations. When the UI initiative started, there was 900 and the transition to UI has moved maybe ~150 integrations from yaml into the UI in these 2.5 years. Even if the UX designer decides to make major overhaul changes to the UI, I wouldn’t expect integrations to pick them up immediately. Remember, this is open source. So your comments like this:

require someone to want to add what you want in. I can tell you, there is a large group of people who don’t want anything in the UI. These are the people who have been here from day one and those people are the people who know how to code. You’ll have to convince them through feature requests, or try to add it yourself.

Also, comments like that aren’t helpful especially if you don’t know the framework. I’ve never developed for the frontend, so I’m not going to make assumptions on how long it will take. You shouldn’t either.

As for the topic of icons

It’s all lowercase and copy/paste from the website:

image

image

mdi:account-box
1 Like
29

Maybe you’re a glass half empty kind of guy, but the developers have never been against you. I don’t understand all the distain towards them you toss around in all your posts.

30

/me didn’t realize until now that HA was such a stable, mature and easy product that people could get infuriated about the UI to change an icon :joy:

1 Like
31

The forums will always be littered with complaints about shifting towards or away from UI configuration. Just look at yaml-gate 2019.

32

I respect your insight, knowledge and helpfulness on these forums, and so that stings a bit. But, fair enough, if that’s how I come across.

I spent a career in IT. I call 'em as I see 'em. HA is a great project. It’s big, it’s complicated and it’s never going to satisfy everyone 100%. The developers are doing a great job. I’ve really enjoyed playing with it for 2-1/2 years so far, and fully expect things to continue to get better all the time.

That said, there are some deep-rooted attitudes condoned here which reflect negatively on the project as a whole, and if left unchecked, will limit it’s reach to just a relatively small group of insiders.

It’s a barrage of simple little things. 2021.8 broke the Honeywell integration. There’s an open problem report. The developer is working on it. Great! I backed out the update and am monitoring for the fix. But other people keep stumbling into this problem because no-one bothered to update the “breaking changes” list.

I could go on and on. I know it’s not a commercial project, and the devs are mostly volunteers. But that’s no reason they shouldn’t consider how end users experience the product. I get it. That’s not the fun part. But it’s part of the job. I’ve worked with some great developers over the years. They’re the ones who sweat the small stuff, and actually listen to the end users.

33

Unintended breaking changes (Bug) aren’t listed… if that were the case, you’d have to search through hundreds in the list each release. I understand you’ve worked with many developers, but expecting perfection from 1000 open source developers is unrealistic. Especially when only ~20 of them do the bulk of the work and manage the other 980. I’d lower your expectations and follow the release threads when things break.

34

When you’ve spent countless hours extracting raw IR codes from two multi-component sound and vision systems with two blasters that function differently when learning and retrieving codes, followed by yet more hours creating scripts for every available IR function (this was after several aborted attempts that went ary for a variety of reasons) and you then attempt to strike an elegant display solution in Lovelace and you remember that innocent icons text box that would surely help your design ambitions out better than the default, ugly vector scroll could possibly, and it’s not nearly as intuitive or informative as to be instantly helpful, I don’t care who you are. You roll up your sleeve and shake your fist angrily at whomever made such a useful program available for free with it’s instructions and feature set so impenetrable and disorderly for the layman that he needn’t have bothered. You do this impulsively not because you want to fill the world around you with your pent up frustrations, though that is a welcome side-effect at times, or to be quietly pitied by the more informed, but because it’s jolly infuriating in almost every instance along it’s steep and often unresolvable learning curve.

This was just one minor irritation too far for this dumbass. Spleen venting is only a matter of time for all who enter this rabbit hole. At least some had a good laugh. A triggered noob or two makes the world go round…

I know it’s entirely irrational to expect a program that is potentially infinite in its complexity, designed by volunteers of varying competency, to behave like a trademarked, third generation of a first to market product that went through scores of test audiences and came out with a party thrown by a multinational worth trillions in market value which was broadcast live and available in the pockets of billions of subscribers, but here I am. Venting spleen without restraint or self awareness.

I am the Xiaomi IR blaster implementation’s missing json ir codes file and I’m mad as hell.

To err is human, but to program bugless code through reverse engineering is divine…

35

I
Yes, but it defaults to hass : script and hass has no icon repository, not an mdi : wotever and as I, and no doubt billions of other potential users have never encountered an mdi : anything in any relatable circumstance, I had to find out. It shouldn’t have been more painful than a simple web search or just a page on home assistant’s site, but here we are. It’s roadblocks like this, whether made from straw or armoured tanks, that will dog this undertaking for years to come.

I guess the question all of the old skool HA boffins have to ask themselves is this: Do we want HA to have a small or a large tent vision? The potential that’s here is infinite and as a tool in a low energy near future I would argue that it’s essential, but the audience numbers will remain small and, therefore, easily ignored by the industry it seeks to tame so long as nobody is tidying the code (for gen pop usage) as they go. Many of the integration limitations can be eliminated once your Googles and Samsungs take notcie (hats off to Tuya for diving in, btw) but until they can no longer afford to ignore HA they will see little point in devoting resources to getting it right.

To paraphrase a line from an old baseball movie, if you build it, they will come, but not if the directions send you into despair over something simple.

36

So, imagine you’re an intrepid explorer, wading into foreign territory for the first time, and the thing you’re looking for is mdi “something”

You see that entry next to “import” directly below account-box that you helpfully highlighted? That’s what I copied and pasted. When it did not work, I went back to find a different mdi “something” but could not.

The natve tracker guiding the expedition starts sniggering at this point… But what I did was perfectly consistent with the information I had received up to that point. A text search box where I could type my “something” would have not only cut down the seven steps required to add all future needed icons to just one step, but also make the platform a nice place for noobs to find themselves in. As it is, you have a rabbit hole waiting there. Plug the rabbit hole and everybody wins.

Any volunteers?

37

I think you have never done any front-end development ?

38

What gave it away? :face_with_hand_over_mouth:

39

The fact you think it can be done in a couple of hours tops. Even for a fluent coder, it can take a couple of hours to grasp how thing are working if they are not familiar with the framework. And than they haven’t started coding yet.

40

As I’ve said in a few other places:

There are usually three “wants” from most people: cheap, easy, powerful

you can usually get two out of the three.

  1. easy and cheap but is basically useless beyond the mere basics

  2. easy and powerful but that comes with a hefty price tag for all of the time and effort invested to make it that way. But even then powerful and easy almost NEVER go hand in hand so “powerful” is usually limited.

  3. cheap and powerful which uses the work of volunteers to do the hard work but the end result is pretty powerful and useful.

I’m guessing you can figure out where HA stands in that list.

I personally don’t think that HA will ever be accepted by the masses and will get stuck at about the “above average techie” consumer. Which is where it seems to be now. And I’m OK with that seeing as WE ARE the helpdesk.

There may be some additional advancements made in the UI but you will still have people coming on here saying “This should be easier. Why isn’t it easier? Apple (or google or microsoft, or etc.) can do it. Why can’t HA? I don’t have all day to sit around and figure out everything that this awesomely powerful but free software can do for me.”

Or it will continue to go inexorably down the “let’s dumb it down for the mass consumer market so they will stop complaining that the really cool things HA can do now are too hard since they won’t be available any longer.” Trust me, the world doesn’t need another app that you are able to configure from your phone.

or maybe the UX dev will perform some miracles. :man_shrugging:

Don’t get me wrong I have no problems criticizing HA when things go down a dumb path but complaining when HA doesn’t curate a complete list of available MDI icons in a drop down menu isn’t that. And if the list is just a wall of text how will that be helpful? you will also need a picture of every icon as well.

That won’t get unwieldy at all… :roll_eyes:

There are examples of configuring icons in the docs. Just type “icons” into the HA docs search bar and click on any of the top results there and it shows you the format it’s expecting.

2 Likes