As the @DataCollie UI gets more complex, more components need to stay in sync. I'm really feeling the power of data-driven GUI design now.
Early on I didn’t apply it well enough, and the cost is showing. Time for a refactor.
#BuildInPublic
#DataCollie Day by Day
Still pushing forward on the Table Designer.
Currently improving the DDL script generation feature.
#BuildInPublic #MySQL
#DataCollie Day by Day
Lately, I’ve been spending too much time on the Table Designer.
I think I need to wrap up this module as soon as possible,even if it’s not perfect yet.
Otherwise, without the excitement of new features,
I’m afraid I’ll lose my momentum.
#BuildInPublic
#DataCollie day by day
Just finished a few updates for the Table Designer's context menu:
Name incrementing now works across different designer instances seamlessly.
If you delete a newly created meta object, the sequence number resets so the next one stays in order. No more gaps in the naming!
#DataCollie day by day
Tonight I added context menu to the table designer.
#BuildInPublic
When #DataCollie started, I built an eventbus based on greenrobot lib. It worked at first, but eventually needed a rewrite.
This weekend I rebuilt it from scratch: zero dependencies, clean, maintainable, with publish / subscribe and filtering support.
#BuildInPublic
Before the release, embedded fonts will be added to ensure consistent rendering across platforms.
Performance in large-scale data processing and analysis has always been a key focus for me. I hope to get your feedback and guidance after the release.
#DataCollie day by day
Tweaked text and icon colors tonight — much easier on the eyes now. No more eye-searing brightness.
Embedded font tuning coming later for true cross-platform consistency.
#BuildInPublic
i use codeAreaBox
#DataCollie daily progress
Implemented DDL retrieval, formatting, and syntax highlighting for MySQL View definitions.
#BuildInPublic
#DataCollie daily progress
Tonight I implemented reading MySQL view metadata, including several key configuration properties, and completed the UI display.
#BuildingInPublic
#DataCollie Daily Progress
Recently finished MySQL view data querying and view metadata preview.
#BuildInPublic
#DataCollie Day by day
Refined the context menu of Table nodes in the data navigation tree to better reflect the current.
#BuildInPublic
#DataCollie Day by day
Guided by Leo’s Interaction Design Philosophy, I decided to make a subtraction:
- Remove the toolbar from table panel.
- Remove New Table from the table right-click context menu.
#BuildInPublic
#DataCollie Day by day 🧩
Removed the DDL preview toolbar. Actions moved to the context menu.
Why:
1. A cleaner, calmer UI
2. Non-frequent actions should stay out of sight
3. One action, one clear entry point
When the work starts to feel like subtraction,I know I’m getting close.
#BuildInPublic
#DataCollie Day by day 🚀
Got the Trigger UI done in the table designer today, with trigger metadata now loading properly.
#BuildInPublic
While everyone is busy building thin GPT wrappers, I’m engineering the professional workstation for the next decade of AI data.
I’m feeling a real sense of urgency for #DataCollie.
The future is clearly in Vector DBs, visualizing similarity and vector search is the must-have for AI infrastructure. Yet, here I am, stuck on MySQL.
It’s a race against time, but I know where I need to be. Get it done.
#DataCollie daily progress
Tonight, I added support for loading trgger metadata into table structure designer and showing it in a table pane.
#buildinpublic
#DataCollie daily progress
Just added support for loading table foreign key metadata directly into the designer form.
#buildinpublic
#DataCollie Day by day
Yesterday, i added a new tab view that lets you browse foreign key constraints at the schema level.
Hopefully, it makes the db structure esaier to understand and reduces mental overhead.
#buildinpublic
#DataCollie Day by day
This week was hectic, but I still wrote a bit of code every night. no broken context.
Finally finished the foreign key design form in the table schema designer tonight.
#buildinpublic
#DataCollie One Step a Day
It’s Saturday. I woke up earlier than on a workday and felt that “being woken up by a dream” again.
In just one morning, I implemented:
1. Editable Column Name field in the table.
2. An Add button that inserts a new Index Column.
#DataCollie Daily Progress
Worked overtime until 8 p.m., got home at 9, and spent 2 hours finishing the feature that loads index metadata into the Index Designer form. Time is like… well, if you squeeze it a bit, you can always find some. #buildinpublic
#DataCollie daily progress
Progress has been slow these past few days. However, I’ve rebuilt the UI and interactions for the Index Designer.
#buildinpublic #indiedev
#DataCollie Daily Progress
Today I refactored the field designer form in the MySQL table schema editor. Switched to a different layout container and finally fixed the occasional dropdown misalignment issue on HDI screens and dual-monitors.
#buildinpublic #indiedev
Got home late tonight, didn’t code, just reviewed what I’ve built so far. Looking back, I’ve done something meaningful past year. the best thing about being a developer? You can lose yourself in your own product, no midlife crisis, no emptiness, no loneliness. #DataCollie
For vector databases, GUI client tools are not just productivity aids. It don't just need visualization; the depth and approach of their visualization directly determine how efficiently users can leverage this powerful AI-native infrastructure.
Vector DB clients differ from traditional RDB ones — they run similarity searches and interpret scores, not exact matches. It’s about semantic exploration, not logic filters. The next wave of DB tools starts with native vector support. #DataCollie #buildinpublic
