Puran
TIL Photos Reading Weekly Current Shelf Search Stats
  • In total, I cleared out nearly 100 books—some I’d read, some I hadn’t. Lately, I’ve even lost the urge to buy new ones, let alone sit down and read.

    → 11:56 PM, Jul 28
    Also on Bluesky
  • I tried out the new ChatGPT Agent experience. Overall, it didn’t quite wow me like Claude Code did—probably because it’s constrained within a web browser. Still, the visual transitions between its desktop and activity views were striking.

    → 11:55 PM, Jul 28
    Also on Bluesky
  • 2025-07-28

    On yet another rainy day in Beijing, I dove into ideas around making AI Build support custom LLM API keys—but Semantic Kernel remains limited to OpenAI. I also explored ChatGPT Agents (visually slick, but less interactive than Claude Code).

    A sudden insight hit while unwinding with Xiaoyan: our Agentic GIS might align beautifully with Claude Code’s Subagents model. Fired up, I explored it late into the night.

    Meanwhile, my brother called—Hamer’s swimming over 20 meters now. She’s growing fast.

    → 11:51 PM, Jul 28
    Also on Bluesky
  • 2025-07-27

    We spent the entire day decluttering—mainly tackling two balconies and tossing out loads of unused items. Rearranging furniture and switching to fresh IKEA curtains completely transformed the space, making it feel lighter and calmer.

    I sold 55 books online for ¥168 after shipping costs and gave the rest to our building’s cleaner. We cleared nearly 100 books in total. It was exhausting but deeply satisfying—more joyful than a day out.

    We replaced the old, heavy, dull-gold ones with a new set: soft white sheers inside and a light turquoise outer layer. Once we hung them, the whole room felt brighter and more refreshing—we immediately wished we had done it sooner.
    → 10:38 PM, Jul 27
  • 2025-07-27

    We spent the entire day decluttering—mainly tackling two balconies and tossing out loads of unused items. Rearranging furniture and switching to fresh IKEA curtains completely transformed the space, making it feel lighter and calmer.

    I sold 55 books online for ¥168 after shipping costs and gave the rest to our building’s cleaner. We cleared nearly 100 books in total. It was exhausting but deeply satisfying—more joyful than a day out.

    We replaced the old, heavy, dull-gold ones with a new set: soft white sheers inside and a light turquoise outer layer. Once we hung them, the whole room felt brighter and more refreshing—we immediately wished we had done it sooner.
    → 10:38 PM, Jul 27
  • 2025-07-22

    I started the day chasing ideas to improve my handwriting app—especially for pinyin and stroke parsing. Found some solid resources, but work came calling.

    Spent the day refining the Identify logic with better parallel handling. After dinner, I updated the app using those references—progress made, but it’s still rough (and Claude API credit is running low). I was really happy to finally discover the clever loading word indicator design of Claude Code.

    A pop-up message with Chinese text appears over a background featuring various technology logos such as Apple, AWS, GitHub, and others.
    → 11:38 PM, Jul 22
    Also on Bluesky
  • TIL: Hanzi related

    Render Hanzi writer character data, stroke animations or stroke quizzes. Hanzi Writer

    • chanind/hanzi-writer: Chinese character stroke order animations and practice quizzes

    skishore/makemeahanzi: Free, open-source Chinese character data

    • Learn Chinese one stroke at a time

    HSK levels related

    • Word Lists - HSK东西
    → 11:25 PM, Jul 22
    Also on Bluesky
  • TIL: Claude Code's loading messages

    I was attracted to the word loading indicator when Claude Code is working on a response. For example, jiving, whirring, channelling, conjuring, pontificating, concocting, discombobulating…

    After searching from the web, I found this topic Claude Code’s loading messages which resolved my curiousity.

    …found out that Claude Code will actually call the API and ask Claude Haiku to generate this single word based on your input. It does this as you type so it’s ready to go when you submit (and it will make many calls for each prompt).

    For those interested, here’s the full system message for the request. It will send you input in a separate user message.

    Analyze this message and come up with a single positive, cheerful and delightful verb in gerund form that's related to the message. Only include the word with no other text or punctuation. The word should have the first letter capitalized. Add some whimsy and surprise to entertain the user. Ensure the word is highly relevant to the user's message. Synonyms are welcome, including obscure words. Be careful to avoid words that might look alarming or concerning to the software engineer seeing it as a status notification, such as Connecting, Disconnecting, Retrying, Lagging, Freezing, etc. NEVER use a destructive word, such as Terminating, Killing, Deleting, Destroying, Stopping, Exiting, or similar. NEVER use a word that may be derogatory, offensive, or inappropriate in a non-coding context, such as Penetrating.

    This is what the official docs have to say about it:

    Haiku generation: Small creative messages that appear while you type (approximately 1 cent per day).

    → 11:13 PM, Jul 22
    Also on Bluesky
  • 2025-07-21

    Xiaoyan was drained from a full day of pricing negotiations, but I saw it as a sign that deeper collaboration might be brewing. We shared dinner and laughs tonight. Hamer called saying she missed me. She’s now learning to swim at the swimming pool.

    At work, I wrapped up bug fixes and cross-testing ahead of schedule and resumed using KanbanFlow.

    A plastic cup filled with a dark liquid is placed on a surface scattered with white and green flowers or petals.
    → 11:13 PM, Jul 21
    Also on Bluesky
  • 2025-07-20

    We caught the midnight screening of The Stage by Chen Peisi—a film adaptation of the stage play I had long wanted to see. It was funny at first, then slowly grew emotionally gripping, reminding me of Farewell My Concubine. One standout was Feng Xiaotong’s portrayal of Yu Ji—subtle, moving, unforgettable.

    Earlier in the day, we met friends for lunch and checked out their new NIO EV. After a ride demo and some good food, we ended the day with great company and a powerful film.

    A person wearing traditional Chinese opera attire and makeup is seated in a dimly lit room.
    → 11:06 PM, Jul 21
    Also on Bluesky
← Newer Posts Page 18 of 112 Older Posts →
  • RSS
  • JSON Feed
  • Lucky
  • Archive
  • About