It's been 2 weeks without progress. Let me share my bad excuse with you.

[BrutalCoding]

Quick update

I've hit a roadblock 2 weeks ago, which I made a ticket for myself back then here:
#24

So, it's been 2 weeks now and nothing has changed. Why? You might ask.

Balancing priorities

I can't find a good balance between work, life and free dev time. It's that simple. I'm working on it in a few ways like: reducing (or simpliying) my goals for ShadyAI and freeing some mental space with auto-deliveries of products at home (groceries, house-hold items that can run out etc).

Life & free dev time

I've learned since starting this side project (ShadyAI) that I've realistically only got a couple of hours every week before I feel guilty (e.g. my responsibilities as a dad). Sometimes, I ignore that guilty feeling and continue with my project (ShadyAI in this case).

A recent example was yesterday eve when my kid asked whether I finished putting together his new bicycle today and I replied with: "No sorry, it's dark and too late now for papa to fix your bike. I'll do it tomorrow.". It's true, it was dark. But I should've finished his bike during the day instead of spending my time learning NixOS and its quirky syntax in this case.

Work

I just removed lots of texts I don't want to overload you with boredom, but work takes a significant portion of my time. Think: 8am - 4pm & most of the time I tend to overwork 1 to 4 hours after. This is just a bad habit I guess, I just try to tackle more todo's than what I've agreed to with my colleagues, otherwise I see them lingering around for many more months.

What's next

Whether I like it or not, I need a tokenizer that works in Dart. What I'll do is attempting to get the official HuggingFace Tokenizer library (Rust) ported to Dart.

As I've mentioned earlier, I tried this alraedy a few weeks back and I failed. The reason I failed is because I rushed through the documentation (Rust to Dart) instead of actually reading it.

This time around however, no deadline. It's meaningless anyways if I put it down and not make it.
I can tell you that I'd like to have it done today, but that's not going to happen.

Here's what I think I really need to get a chat app prototype:

1. About 4 hours full focus to have the HuggingFace Tokenizer in Dart. Ideally I do this in a different repo so I can publish it as a Dart plugin for everyone to use.. but I'll skip that due to my time. It'll end up in this ShadyAI repo directly.
2. About 2 hours to look into usage examples of other projects (language doesn't matter). Having the HF Tokenizer in Dart is not helpful if I don't know what to do with it, right?
3. Another 2 hours to apply theory in practice. Just imagine me bashing my keyboard for 1 hour and 50 minutes, where the last 10 minutes is used to celebrate the minor bugfix that made it all work.
4. UI time: 1-2 hours for a working chat UI, probably looking really horrible. But this is important to get the right look and feel of what I want to achieve.

Done. I'll share what I have once I got that done, no matter how bad the UI turns out. Whether I spend 2 hours or 8 hours on the UI, I know that I'll spend hours making changes to it anyways over time. Might as well start with the fastest & worst UI.