By coincidence, another issue arose illustrating this very problem. See #7519 and the end of thread Rename query fields when renaming a page title.
And yes, the timing is a pure conincidence - I saw the closing of the issue above only a few hours after having posted the above post.
I’ll quote myself from the github issue:
In the larger picture, I thinkLogSeq has fallen into a trap, that many software projects have seen (not with a nice ending): Core functionality is lacking, therefore the part of the user base consisting of dedicated long-term users is small, the user turn-over is large with many new users coming in but leaving after a short time. Creating new shiny features does bring on new users - but again, they don’t stay long. Large user turn over → many new registered users and Github stars are rising, which is interpreted as a positive sign, when it is actually a sign of a problem. The team thinks all is well and continues creating the new shiny features and ignoring the basic functionality and usability. It’s an evil spiral pointing upwards, but it consists of nothing - it’s like an inflation bubble.
I have seen too many software projects failing because of this effect. It would be a shame if a tool with the potential of LogSeq took this route.
This needs to be fighted somewhat similar to inflation: don’t keep inflating, instead go back to the basic mechanism of the economy/software and do the groundwork to make an actual, stable improvement. The fundamental, unsexy, stabilizing groundworks. To win back a larger base of dedicated, long-term users.
Emil Eifrem, founder and CEO of Neo4j, with Neo4j being somewhat unmatched on the combination of company valuation and user satisfaction the last 10 years, has his “one single advice” on creating a successful software project that comes to mind:
“As a software tool company, you can never, ever, over-index on the satisfaction of your actual, core, daily users”.