Ending the support threads?

[johu]

Even though I just declared the forum an autocracy I now decided to ask "the people" what to do.

Now, Damien liked to organize his support efforts in single, page long threads, one per product. These threads are incredibly long and contain an awful amount of redundant questions and thus redundant answers. I don't think many people are going to read through them. Damien himself thinks they don't really work as intended (first read, then ask).

For the SDU boards I locked the support thread 2 years ago and it seems to work fine.

So I'd like to lock all these threads and if issues arise a new thread with a descriptive title (Obama would say: hope) is created, provided there isn't a similar one already. This will make it easier to research existing issues and also display them in the similar topics view.

What do you think?

[Bigpie]

1 issue 1 thread might be better once the solution/board/software is mature enough

[arber333]

Hm...
Maybe an intro post with structured ToC list would help.
Then any issue would be declared hopefully with an answer and this would allow us to find it quickly whilst keeping the single thread.
This would mean some admin work editing the first page whenever a new item is added. But i could argue each addition would mean a change in first page. One should simply check there to see if there are any news .
Also we could add a timestamp with the ammended links so it would be easy to see progress.

[Ev8]

The support threads are a wealth of information, the longer they get though the more diluted the relevant information becomes…. People become less likely to read through and ask already answered questions, so the problem becomes self sustaining… I’m all for locking the support once and teething issues with new hardware are sorted, and questions past this point would have to be stand alone posts I guess if something relevant crops up it could be added to the support thread later by whoever developed the board maybe but this is obviously extra admin which I think we are trying to avoid right!

[Ev8]

Dare a suggest a two tiers support structure in the forum, normal forum members get access to all the usual hardware dev threads but oi board specific support is open only to members who have purchased hardware… thinking out loud and not sure how open soruce this sounds…

[catphish]

I much prefer one thread per issue. It makes it much easier to follow!

[nkiernan]

I think there is benefit in single thread per issue for cleanliness, but would benefit from some form of linking to a master topic for example. Its the trouble of organising so much data and allowing freedom to word things yourself, so duplication is not always easy to find/handle.

Is there any way to consider something like support/aftersales structures that use set dropdowns to force new questions/topics to the correct location and also make everyone use the same terminoligy to aid searches just up to the point the single thread gets allocated its 'home' on the forum...giving somehting like a FAQ library. Automation to force more granularity sort of thing?

[johu]

Thanks for your input on this.

Since it's 10 yes votes and no no votes I think the trend is pretty clear.

I don't yet want to segregate shop customers from self-builders.

The threads will of course remain in place, maybe even as a sticky. I'll make sure they all point to a wiki page that can be updated by anyone as opposed to the opening post that can now only be edited by me.

[tom91]

I would agree to doing one issue per thread.

Largest downside is someone just creating threads that are duplicate issues. This can lead to needing to needing to do more intense moderation.

Would it be possible for more "Senior" forum members to have a sections where they can do everything and new members cannot create threads or posts until they reach a certain level?

[Bigpie]

I would also encourage a support initial post template to get information, such as software version, hardware version, observed behaviour, expected behaviour & steps already taken to address.

[MattsAwesomeStuff]

One of the roles I generally end up playing in most organizations I volunteer at, is to be something akin to an Archivist. Almost all communities have these same struggles.

I'll state something that most places get wrong: Discussion forums are a great place to DISCUSS topics. NOT to find information. Anyone who says otherwise tends to be bad at putting themselves in other's shoes.

They're horrible at finding information:

- It's a massive amount of information to read.
- You don't know if the answer is there.
- You don't know the answer is there even when you're reading it most of the time.
- You don't know how to recognize your own question that was asked before, because you lack context.
- You don't know when to stop looking or how hard to keep looking.
- You don't know if the information you're reading is already outdated or corrected later.
- You have to slow down and pause to try to understand everything, in order to understand whether or not that's the answer to your question, you can't skim, which means you're not just reading everything, you're trying to learn everything. To people who understand it all, that's easy, you already know it. To someone who understands none of it, that's hopeless. It's like learning a new language by reading the dictionary.

The people who think that forums are good places to find information are kind of like... my dad used to tell me to go find a screwdriver or something in the garage. Maybe a tool I don't even know what it looks like really, that he loosely describes. He knows it's there, he knows where it's likely to be, and he'll instantly recognize it when he scans past it because he put it there. To him, "Go look in the garage" is a short recollection task, where everything reminds him of the 3 possible places he put it, and ignores the other 99.9%. To me, it's aimless wandering. If he'd just have a place the screwdriver belongs and put it there, I could find it. But that requires something better than a garage full of meta-data, where everything is sorted by chrono-importance - the more important and more recently he used an item, the closer to the top and middle of the garage clutter it will be. I lack this meta-data, so this is useless to me. I eventually give up, he gets pissed off, finds it in 3 seconds, and then complains that I'm too lazy. People who tell others to "search the forums" with a sneer, are just like my dad (my dad's fine, he didn't do it often).

And, I know that when I tell someone "there was a thread about this last year sometime, go search for it", that they'll spend quite literally 100x as much time looking as it would take me to recall it, and to recognize that thread even if I'm not at the answer, because it jogs my memory that yep, that's the right thread, it's a bit later on.

I've always said, if a community has to nag "This has been asked a million times, use the search function", what you're actually doing is picturing an amalgamation of all previous answers. If you go looking what you'll actually find is a bunch of unanswered dead ends, a bunch of half-answers, some wrong answers, some reference to earlier answers, and a bunch of people saying it's been asked a million times just use the search button. If you had to cherry pick a single thread that answered it well, you might be surprised to find that no such thing exists. The real answer is "If this is asked often enough to annoy anyone, then this is the wrong place to ask it. Write an article/add to the wiki/create an FAQ."

And you HAVE TO keep it current. The answer MUST be, ALWAYS "The most recent information is in the wiki/whatever". As soon as that's not true, and the community has slacked, then no one uses the wiki, because even if they read it they're going to post to ask if it's still the most recent info. There has to be a system in place and confidence that it's the best place for answers.

"It's on my github" is not an answer either. That might as well be in a black hole. And surprise, nothing's ever documented for anyone else to properly interpret either. "It's on my github" is like the napkin scribbles you wrote when you were drunk. it maybe makes sense to you, usually not anyone else.

That said, most of the time questions aren't "questions" you can just state. They're not "where does pin 21 connect to?" with an answer and a "I'll update the wiki with that." They're whole situations that takes 12 back and forth posts to describe. "My inverter isn't working, here's my parameters, it's doing this but not that". How the hell you would ever summarize that up into an FAQ... I dunno. And, most of the time you don't even know *when* the process is complete and can be reviewed and summarized. It's just an ongoing journey of "It's sort of working now, we'll see" and then further tweaking.

...

What I do like about keeping all the support in one place in one thread, is that it cleans up the forums. People suck at labelling threads, "I've got a problem here...", "Can Someone Help Me Out?", etc. It's a useful archive to no one. I hate going looking for a thread and seeing 50 short threads about nothing. It's as bad as seeing a thread with 400 posts. At least the 400 posts are in the same area.

If I go to the Outlander sub-forum for example, most of the top 15 threads are all the main threads I'd ever be looking for. The reverse engineering of each component. Soon as that becomes random questions, I'll never be able to get a picture of what's being worked on.

What I don't like about keeping the support in one thread is that no one's going to read it and it's impossible to find your question.

What I've done over on the DIY EC community a few places is written sticky threads, and if people have information to contribute they make posts, and after a few days I incorporate those into the top post and then (because I'm an admin), I delete their comments. So the thread stays fresh and the information in the top post stays up to date and there's no musing. The thread having more than 1 post is basically just a list that says "Matt hasn't cleaned this up in a while".

I think that would be a good model for the support threads. Basically, every question asked should be able to be condensed into the info on the wiki or an FAQ, or an article, or something. If it can't, it's probably useless anyways. Two problems: 1 - Someone actually has to do the work, and 2 - Pesky admin access to purge posts.

...

Discussion threads are wonderful for DISCUSSING development. It's great for people staying in touch with the process. It's like a scrap book. That's a great use, and 400 posts isn't a problem there, because it's how the design evolved.

...

The only exception I've seen to this, is on a forum that organized the forum by the repair manual chapters for the car. If you have a question about brakes, you go to the 4.1 Brakes section. And you're expected to look up info in the past. Whatever problem you have, you're expected to find the thread that discusses it, dig it up, and post to it if it doesn't answer your question. It has a very active moderator group who are moving and merging threads and posts every day, keeping the technical part of the forums (versus the projects or chatter), well groomed.

That's for a 50 year old car. That doesn't really work with anything new or anything being developed or tested.

...

Also in general, engineers suck at documentation. Since the dawn of time. There wouldn't be so many questions needing answers if developers did fewer things that they documented better. But that's not fun or interesting. It's dumbing yourself down to the average person who would use that info.

I think DIY EV adoption is less than 1% of where it could be, if things were just documented better. You do all the hard work, but stop before getting any of the results.

...

If this had an easy answer, we'd have already done it. Archiving information in a way that is useful is hard, even when you have people who want to do it.

[dadiowe]

Thanks Matt, you have pretty much summed up my frustration with trying to find information. I am constantly going through pages of articles I have read trying to find that little gem of information.

Your suggestion regarding keeping the important information at the top seems a great solution.

Unfortunately I am just an aged motor mechanic with a bit of information picked up through life trying to create my dream car. I don't have the programming or electronic skills to help you guys or any skills in moderating forums.

What I do have is a BMW Z4 with a Zombieverter running a GS450h Gearbox and an outlander charger (working on the charging). I am also retired so I have the time to try updated firmware in my set up.
Please let me know if I can help in any way.

[johu]

Yes this comes up time and again. Yes I suck at documenting or more nicely put I don't enjoy documenting. It is a clear expectation of mine towards the community that they jump in at stuff that I don't like doing. I supply schematics and software, the community adds the rest. And this is already working to a degree and as anything can always be approved on.

When I closed the support threads I noticed in none of them the matching wiki article was even linked. Now it is. I believe in incremental improvements, that was one.

Also many problems are very generic. The last topic in the LDU thread came down to Dave Jones "Thou shall measure voltages". So in addition to the existing FAQs (like here https://openinverter.org/wiki/Inverter_FAQ) a more generic one could help. An EE basics page on how to use your multimeter up to even some more specialized topics such as "why don't I see the parameters on the wifi interface". Luxury item: pictures of commonly used boards with dots where to find 3V3, 5V, 12V etc.

Might put that up as an open task.

EDIT: made a stub https://openinverter.org/wiki/Electronics_Basics

[0tik]

MattsAwesomeStuff
Oh my god thank you so much for describing the source of all the sad, negative emotions and thoughts, lost time, and above all - feelings of hopelessness and frustration. I am not talking about single support threads on openinverter forum but about searching information in general.
Having to go through 40 page thread just to find answer after reading 26 pages of thread is really soul-crushing experience. And I only did that because I gave up searching everywhere else. Thought "might as well read this trashpile again from start to end". And there it was. The answer I needed so much...

A more expanded FAQ for people who want to start a project based on VCU would be really helpful to everyone new too. VCU and more data on parameters of inverters and engines. I understand that it's not easy to test the limits of every hardware
combination but if people made posts like "hi. I have this engine+inverter+battery volt+amps and it works good with XXkw power limit." Building something on my own would be so much easier with knowledge of what people built and works for them.

[mark]

Great post, Matt. I think there's a lot of gatekeeping in some of these communities, DIYEC more than here, for sure. Whether conscious or unconscious, it's not easy for someone to stumble upon the community and bootstrap themselves up to a working level of knowledge. We don't prioritize making it easy for new people to find and understand the projects. The most knowledgable people are wisely using that knowledge to keep the projects moving forward.

In my 20+ year technology career, I've worked as the graveyard shift help desk operator, the support and operations technical lead, a infrastructure and web developer, a sysadmin, and a domain architect. Due to the breadth these different roles, it forced a bit of perspective on me, regarding how to write documentation that was both useful to the end users as well as valuable to the more experienced technical staff. When running the SOC, we would look for similar problems that were generating a significant percentage of our support tickets. We would then write documentation that covered as much of that class of problems as we could, and keep it updated as we learned more.

End users typically start their investigation with symptoms. I've done X and I'm seeing Y results, where do I go next? I think it would be useful if we could put together a "common issues with openinverter" wiki. For example, this is what it looks like if you have a loose connection, bad wiring, interference, etc. These are the symptoms you'll see if your parameters are way off. This is how to tell if your a hardware component is malfunctioning or non-op.

I've been looking at the wiki, and thinking about how to improve the Getting Started section, specifically. I think it deserves its own page and the Main page should be more of an overview of the project with links to some major categories. Even that first page of the wiki is a bit like drinking from the firehose. I think the Electronics Basics is a great addition. Things like How to Build your OpenInverter/DIYEC/Electronics toolkit/test bench/workshop with examples of which tools work and why would be a great help to people. For example, you're going to need at least two bench power supplies, and they're going to need to provide X volts DC and Y amps, and here's a few examples from aliexpress/digikey/amazon. You're going to need test batteries, etc.

I'll leave it at that for now. Understanding systems and documenting them are a decent portion of my work life, and I could ramble (even more) about this for quite a while if I let myself.

Suffice to say, I'm happy to help write documentation, or however I can contribute.

[Bratitude]

i think the open task page was a really great addition, maybe we can take it one step further with bounties? almost like personal support, but its open to whom ever to do the work.

we always talk about " this should be in the wiki" or that the forum is no a great place to find info.

a motivation for documentation.

[johu]

i think the open task page was a really great addition, maybe we can take it one step further with bounties? almost like personal support, but its open to whom ever to do the work.

You mean community funding for whoever takes it on? Makes sense.
Any ideas how to organize it?

[MattsAwesomeStuff]

You mean community funding for whoever takes it on? Makes sense.

I've proposed bounties in other communities in the past, but I've never seen it be necessary, or even for it to work.

There's already a bounty system in place: Pay Johannes for an hour of his time.

The problem with rewards is that the kind of people who can answer the questions effectively, aren't going to be swayed by the (small) amount of money they could get paid. If they want to help, they'll help for free. If they don't want to help, the bounty won't encourage them.

And, it does create a bit of a nastier competitiveness out of the community, commodifies everything, and starts to make people think "Well why should I help for free?"

It's a delicate balance, can sour the mood of the community.

If the community was 100x the size that it is, that might work better, because of economies of scale.

Just my two cents.

[johu]

Yeah I see the point in that.

Just remember that I did actually pay (currency: boards) people who did open tasks, but those tasks weren't really very community relevant, they were production optimizations that mainly benefit me who makes larger production runs.

[Bratitude]

I've proposed bounties in other communities in the past, but I've never seen it be necessary, or even for it to work.

There's already a bounty system in place: Pay Johannes for an hour of his time.

not for answering questions, but for open task. like filling the wiki, code dev , hardware dev, work review, etc.

You mean community funding for whoever takes it on? Makes sense.
Any ideas how to organize it?

the like button was a good addition, what about a tip bot?

for bounty's, unless we go full crypto, some one will have to be a oracle. this may just result in more work in the end, but may be as simple as " job is completed, here's the pay out"

ill look into what options are available.

maybe once a month there's a scheduled open task "update x topic wiki page"