There are plenty of habit trackers out there, obviously, so this has really been an exercise in figuring out how to integrate a tracking app as seamlessly as possible into a daily routine, while generating a dataset that has the potential to give meaningful personal insights. The jury's still out on that.
Testers, collaborators, and feedback of all kinds welcome.
Is there any chance you could make a video that shows how the app works? The screenshots look nice, but especially with an app that is 'always there' a video showing how it works in practice is important.
The Github repo mentions a REST api that can be used to push arbitrary data into the system. Are there any docs around this yet?
If you're familiar with python webapps, you can see api.py:TrackingAPI, which takes a date and a simple JSON data param.
Let me know if that looks like it fits your need, and if not, I'd be curious to hear more about how you'd like such an API to work.
I'm working on a similar project myself but I welcome innovation in the space.
I can't wait to poke around this a bit more. It also uses the same stack I'm planning to use on my next personal project (Django + React).
Me and my friend were working on a very similar project for the last 10 months, and we still hesitate to publish it (also, it is written in Scala, thus limiting potential audience). Below, I provide more information about our work, some insights discovered during development, and future plans. I hope it could help other projects too.
The project is focused on outsourcing memory, providing smooth, most of the time automatic, time-tracking, being hackable and keeping personal data private.
Key features needed for that are different integrations (we use JIRA, Trello, Hubstaff, Trakt.tv, Apple Health, sleepasandroid, Beddit); constant location tracking; convenient input (currently, Telegram bot with NLP capabilities, suggestions and autocompletion).
Eventually, it turned out, that chatbot interface works like a charm for actions like forwarding links to read them later, managing tasks in JIRA (when chatbot sends you updates, and then you can use your messenger to comment on or manage tickets; also works well with time-tracking), aggregating notifications.
We also have some analytics: daily briefs, time tracking summaries, metric-based reports. Another features include global search, flashbacks, GPX track export (to use with Lightroom, for example), history-based notifications. Future plans include: finance tracking, delivery tracking, ReSpeaker integration, newsfeed aggregation.
Is there anything public on the web about your project at this point?
Sadly I tried to log in with my gmail account and it just returns me to the Login page again.
I'm not going to migrate from Org-Mode and Emacs as I'm too invested there but I find Flow Dashboard a very good project for people who are new to Quantified Self and doesn't want to spend dozens/hundreds of hours setting up their own system or don't need that level of detail/features.
Anyway, these are some features which I miss:
- Being able to clock inside tasks. Also expected time to completion of the task so you can adjust future predictions.
- Child tasks. At least projects should have them. Common tasks may just have checklists.
- Good editor and markdown support for the journal.
- Integration with some finance/budgeting app like YNAB or Ledger.
- Work without Google AppEngine. My journal shouldn't be a hack or a search warrant away. Too sensitive.
- Being able to schedule the tasks and habits. Maybe integration with Google Calendar, but privacy again..
I wouldn't attempt using more than once tracker, and I'm not keen on tracking personal matters. I prefer the freedom of taking action on personal matters in response to how I feel and what's happening around me. A dynamic organic order, (or permissible lack of discipline).
TLDR: why are we relying so heavily on cloud based search engines to organize hyperlinks? Local bookmarking tools could be so much better!
I've often thought it'd be great to have some kind of browser extension that used search queries to automatically tag and categorize bookmarks and threads of browsing history. Then, instead of having to manually put bookmarks into "folders", or think of tags, they'd be organized smartly based on the tree of searches and links that they are related to.
Here's a more concrete example: Say you searched for "curry recipes". You click through a number of different pages, finding one you like. When you bookmark it, bookmark would carry that "curry recipes" metadata, and be more searchable in that way.
A more complex example. Say you search for "gravitational force", and end up reading part of a blog post, then clicking through to another page and another. The third page is really interesting. So you bookmark it. Since the original search of "gravitational force" was the seed of finding this third page, that query would, by default be associated with the bookmark, even if it's not directly related.
This could be extended with all kinds of interfaces. For example, your browser could display a kind of "tree" of recent search queries at all times, which could be pruned, cleared, or added to at any time. A new tab would by default inherit the old tab's tree. Bookmarking would flow naturally from this, as a way of snapshotting a given thread of browsing.
Another, more complicated option would be to bookmark actual threads or trees of links themselves. A cluster of related wikipedia pages could be bookmarked together and traced to the search query that initiated their discovery.
If a page was implicated in multiple bookmarked clusters, it'd be easier to find it in your local bookmarks.
The upcoming annotation standard could end up being a great tool for extending this idea: https://hypothes.is/blog/annotation-is-now-a-web-standard/
Isn't this just a variation of the old pagerank algo?
That is to say, yes, but I think I'd like the proposed solution so much more.
I'm working on a simpler/cuter solution for habit tracking and quantification (http://willyoudidyou.com) but i think flow dashboard looks great for hackers and quantification enthusiasts that want total control and to create their own integrations.
From the local devserver instructions: >gcloud components update
Looks like there isn't?
But it is stored in the cloud?
Only google login? Guess I won't be trying it out then.
A number of people have commented that Google login only is a barrier, so will definitely be adding another signup method.
The source is on GH with instructions to spin up your own instance if you prefer.
If I can't export the data (I see nothing about it) and it's not stored on my device I think it is a stretch to say that I own my data. I get that it might not the first feature to implement, but for the future that is something that at least I would expect.
And what does it offer beside what any other tracking software offers? I still don't know after reading the page. Looks like the same thing a thousand others have done before.
Awesome app though. Good features and aesthetics
It seems like Google's APIs for tasks/keep/goals, etc are in flux right now, which is not ideal, but I will be looking into an integration to remove that redundancy.