What we should be using is something that allows the definition of a list of ailments and then having everyone specify which ones they had and of which kind.

Having all of this done through "free text" fields without a format is a complete waste of time.

It may be that there's already that kind of sites available but I have my doubts.

Unfortunately I'm not a webdev so I can't be of great help right now..

Split this for this own thread.

What would be useful fields to have?
eg

Affected by:
About me:
Current setup:

No. I was thinking of having a website where you could say what problems you had, modeled like a graph.
Something that would allow us later on, as we have more users, to automatically run plots over and stuff like that.

Each user can, for instance, have a location, an age, a profession, etc.
Then each user has associated a set of conditions.
He knows he has those conditions because he has a set of symptoms, as well a severity of symptoms.
He also knows that he can improve some of the symptoms with a set of workarounds (could't come up with a better name).
He can also list what has he tried and why. And what seemed to work and what didn't. And why, etc.
What other migraine triggers does he have? Food? Noise? etc
The user also has a set of non-migraine-related conditions (bruxism, hay-fever, broken, septum, etc) that maybe is actually migraine related. But we can only know after we start having tens or hundreds of people putting their data.
etc, etc.

I think at least in the beginning we would allow the users to input their stuff in textual form and we would then create this node structure for them, but over time we would start noticing that the structure was solidifying and probably users could add their stuff completely autonomously. But I digress. We currently have so few users that we can get away doing things manually.

You may think this is too much stuff. Most people would start off by just putting the most obvious and simple stuff. That would already be OK. But most people here are also desperate, so maybe they will take the time to sit down and write down everything they can think of. Just looking to the data of other users may make them realize that they also share some of the conditions. Maybe then patterns start to quickly emerge. I don't know. Think of all the information that these ~100 users already put here. It's a lot of info, but unfortunately it's completely dispersed and in a terrible format (written text). We need something that makes running queries over it easy and straight forward, and that makes visualizing it also easy (what's the % of people here with cluster headaches, when did the most people start noticing these problems, etc).

This is a cool project on its own but unfortunately for the following months I most likely won't have any time to develop such a site. Ideally it would be tied to this forum, so we could easily jump from here to there and vice-versa.

  • JTL replied to this.
  • KM likes this.

    __528491__ No. I was thinking of having a website where you could say what problems you had, modeled like a graph.

    That's an interesting idea.

    Sounds like a survey
    Limesurvey has something like this.
    If you're interested in molding a survey to gather data I can set something like that up and link it here.

    • JTL replied to this.

      Slacor Hmm. I disagree. Surveys are usually "one time" and what I think @__528491__ is alluding to is a more "constant" project.

      Hmm... well without a "purpose" (social media, streaming content, etc) there isn't much out there where you login to simply "put data in fields"
      I have thought of this idea (with a entirely different subject and purpose) but I find the hard part to be entirely in the UX part.

      My thinking, without coming up a entirely NIH approach:

      • Use survey software with persistent tokens, allowing for updating of responses
      • Have very loose, long text questions that would be delimited by newlines
      • Consistent publishing of results in JSON / CSV formats.

      Use survey software with persistent tokens

      This is not logging in, but re-clicking on a email link which is not what I would prefer.

      • JTL replied to this.

        Slacor I think you hit the nail on the head with the UX issues.

        a year later
        dev