Username mentions are now live.
A person who thinks all the time has nothing to think about except thoughts, so he loses touch with reality and lives in a world of illusions.
— Alan Watts
An interesting quote to keep in mind while designing Shimatta. Shimatta's goal is to capture thoughts based on your daily experiences and learnings. It shouldn't actively force you to think about stuff that you could write down just for the sake of doing so. But what if a thought occurs but you're not in the habit of writing them down or you forget doing so?
A solution for this could be to let the user setup individual reminders. Reminders that will be sent at a moment the user thinks is the best time to reflect on his thoughts. Be it in the evening or the next morning those thoughts occurred.
Also just reminding a user "what was on your mind today?" might be to generic. Everyone is different, everyone does different things during the day. I for example stumble across interesting articles or programming libraries during the day which I might want to keep in mind. So my reminder should ask me:
"Have you found any interesting links today? What's interesting about them and how will you be able to implement them in the future?"
Of course, those are just thoughts about improving how to write down thoughts. And as the initial quote said, just thinking without doing detaches you from reality. I don't know if the described solution above is useful unless I built it and tried it out.
Well, I'm looking into adding a recurring payment integration with Stripe for the third time. Already tried it twice on two other projects with no/part success. I'm always stumbling upon something new while developing and I haven't found a guide which shows a nice and clean complete overview of what's required for building a subscription service. It's not that I couldn't build a working prototype with the information available. More that I want to have a clear vision of how the system has to look instead of building the walls and realizing that some parts of the foundations are missing and wasting time refactoring everything five times. I did some research today and combined with my experiences from the last two trials I think I got a pretty good overview together (for now). I've thrown everything into a list and split it into different iteration steps which I'll tackle one after another. If everything works out as imagined I'll write up a complete guide on how to integrate Stripe subscriptions into a Rails application. So others can benefit from my struggle. So long.
Oh, btw: Shimatta know supports public and private posts.
Hint: You can save posts as "drafts" by making them private, updating them and publishing them later on.
The current plan is to add a simple "keyword" search to Shimatta which returns posts that contain the search query in their content. On a later stage (which means when we have more content from different users) it might be interesting to make search more "content aware". One possible solution could be by using a bayesian classifier.
It was introduced under a different name into the text retrieval community in the early 1960s and remains a popular (baseline) method for text categorization, the problem of judging documents as belonging to one category or the other (such as spam or legitimate, sports or politics, etc.) with word frequencies as the features.
Humans think all the time. And about 99% (this is pure speculation, but would be interesting to find out if this measurable) of our thoughts are lost. While some thoughts are worth losing sometimes some really cool and valuable ones get lost as well. I think we can do better. In a few years we might be having technology that can tap into our brains and extract our thoughts but until that we need an alternative. Why? Thoughts are the origin of most if not all human inventions. By crafting and combining thoughts we generate new ideas which lead to new innovations. With the internet we've gained instant access to millions of other people and their thoughts. A few of those thoughts are already captured. But they are scattered across the internet and most of them are probably not of high value. Imagine if we could bring high value thoughts together in one place and connect them in a way that let's us generate new ideas based on thoughts of the entire human race. Of course this is moon shot thinking. But I think worth considering. And that's what the long term goal of Shimatta is about.
One issue that I have with blogs, twitter or facebook is that it triggers a kind of anxiety in me. Every time when I want to post something I get a feeling of like "is this good enough?", "does this make any sense?", "what will other people think about it?". For me those mediums do not encourage me to write down my honest feelings and thoughts without putting a big amount of time in it.
I want to address this problem in Shimatta but I'm not yet completely sure how to do this. Of course one option is to allow private posts, but I'm thinking more about the public ones.
I think the initial message what Shimatta is and does and how the user will be triggered for writing his thoughts down will be essential to solve this problem. If the user thinks a post will be read by all his friends and hundreds of strangers it will be much harder to publish something instead when a user is thinking "let's write this down quickly for later usage" and not giving a shit if everything is fully understandable or has grammar mistakes and typos.
Another part is the community aspect. Writing down thoughts doesn't require a follower count, like count or read count. Because it doesn't improve the quality of a thought in itself. So the social "proof" pressure can be removed.
The last way that currently comes to my mind to encourage users to post more is to teach them. A strong habit building onboarding strategy is key here. All in all I think it's a problem worth solving and most importantly I think it's doable.
This is an excerpt of a conversation with @sirupsen
Regarding merging commonplace book and daily blogging: I’m thinking it’ll come down to a good search feature. At the moment you just have one feed and tags are basically the only thing to categorize your posts/thoughts with. A search which finds posts not just based on matching keywords but also on recognizing relevance based on the content of a post might be quite helpful in that matter. That’s what mostly bothered me with Evernote and sometimes caused me not finding relevant notes again because I messed up or forgot the keywords.
I'm currently working on cleaning up the backend of Shimatta and rearranging some things for future updates regarding private posts. More soon.
Here are some things that are in the pipeline for Shimatta:
- Private posts
- Auto linking @username (but only the ones that exist)
- Shimatta API (closed for now)
- OS X app for faster posting (that's why @rodrigo's feed is full with OS X and iOS development stuff)
- Referencing other notes
- Auto expanding content from external links (I'll probably create another service for this with a public API)
- Emoji support :+1: