Tweet

2024/03/30 21:19

Now I can generate blog post directly from joplin thanks to joppy, which is a python version of Joplin API. Finally I did something useful for myself using python. Old posts should not be just dead archives of past thoughts. I should review and improve them after some period.

2024/03/30 23:06

If I could simply use Obsidian or VS Code to edit the files, why should I configure this mess? For the sake of open source? I’m tired…

2024/05/28 12:22
I dived in to the source code of joppy, which is not complicated at all. It’s just making requests to joplin’s native API. Scrutinising other’s work is really the best way to improve.

05/01/2025 11:23
It’s fine to make a copy for long posts, but it’s clumsy to copy quick thoughts like tweet. Maybe I should improve this.

2025/03/24 22:41

Syncthing proved too sluggish for syncing my Joplin notes—Android’s transfer throttling turned every load into a test of patience. The solution? A self-hosted Joplin server via Docker. Now, I can seamlessly edit my blog posts straight from my phone, without the wait.

I finally finished rearranging this site! Now the gallery section shows my drawing in pleasant grids, the tweet section directly loads the contents of my tweets, and the homepage is no longer a messy feed but a clean introduction and table of contents (I move the feed to a separate page). I become more familiar with how hugo manages the page templates and of course, learnt a lot about editing html and css. I am very satisfied. Hmm… what should I do next?

The code was very cryptic… how did I finished this thing? Maybe I rearranged the code from the original template?

It’s quite limiting to using this template. The way how the CSS files organised is very confusing.

Actually, It’s not that hard to understand after I read through them.

Tailwind css looks tempting because it reveals the problem of overly structured and scattered CSS classes. Don’t predict which classes can be reused, only reuse it if you saw it repeated many times. But it might not be a good thing to make edits without understanding the structure of the site, so I’m doubtful.

https://scriptraccoon.dev/blog/tailwind-disadvantages

Not enough… this site is not interesting enough.

tags:#iw
I just want to make a menu that sticks on the pages using CSS, how come this could be so difficult? Even ChatGPT gave me the wrong answer.

2024/06/25 03:44
Problem Solved. The key is to make the max height of the panel be 100vh minus top padding minus bottom padding.

tags:#iw
I finally migrated my operating system to linux.

Originally, I wanted to do a small-scale test on my old desktop computer, but my beloved ancient machine has failed me and refused to boot up. I have no choice but to install linux on my laptop’s spare drive. It’s very simple since I’ve tried that before on a virtual machine, but on bare metal I need be careful and not to accidentally wipe my original drive when formatting disk space.

Surprisingly, everything software I need on Windows has an linux alternative. I also tried other interesting stuff like a tiling window manager (qtile is the only one that works out of the box). I could have finished all the work earlier but I wasted a lot of time browsing because there are no ready-made net blocking software on linux. Finding the key to lock the Pandora’s box is really tough.

It’s hard to ask for help. If you procrastinating until the last hour and realise you can’t solve the problem on time, the shame you feel when typing the text message to ask someone for help will makes you throw the phone away and panic alone.

If we rely on the internet to answer every small question, we miss the chance to slowly build trust with others with these questions. It is very tempting to quickly obtain the knowledge without having to admit our ignorance. Is it really the cause? Only trust begets trust, but I don’t have to trust others when asking those easy questions, it doesn’t matter whether they could answer them.

On some extreme occasion, if I am really scared about the immediate danger, I will draw on everything in sight to save myself without a second thought.

Maybe I just don’t believe that I could get meaningful help. If I go ask about homework, instead of discussing the the answer together, they will finish the work alone and share their answers.

On the filp side, if someone ask me for help, I can act immediately. I may feel annoyed sometimes, but I like the feeling to be prompted to do something.

The hairdresser grumbled that my demand was unclear, although I had been using the same vague words and the outcomes were somewhat consistent. How misfortunate. At least I learnt their lingo.

If I don’t post anything, I can’t feel my progress. If I can’t feel my progress, I get stuck on changing the same text. In term of writing novel, should quantity be favoured over quality? Is there a way to quantity progress if I can’t predict when and how the project is going to be completed?

I finally finished rearranging this site! Now the gallery section shows my drawing in pleasant grids, the tweet section directly loads the contents of my tweets, and the homepage is no longer a messy feed but a clean introduction and table of contents (I move the feed to a separate page). I become more familiar with how hugo manages the page templates and of course, learnt a lot about editing html and css. I am very satisfied. Hmm… what should I do next?

After giving up on YouTube, Zhihu and games, mindless searching had become my only time drainer. I thought of a way to set a soft search limit by opening ten search tabs upon browser launch. Also, I removed the search function of the omnibar. That means I run out of searches after the ten tabs have been used. What an elegant solution.

2024/04/07 03:37

It was a good idea but I can’t really cling to it. What a shame.

Tweaking new productivity tools may not really pay off. Still, I tried Super Productivity and ActivityWatch and they both seem promising. How will the data be useful if I’m trying to gamify my life? I shouldn’t worry too much as long as it is automatic.