Skip to content

plugin-map: no fitBounds to data — markers land on an adjacent world copy, initial viewport shows an empty map #4941

Description

@yinlianghui

Observed

Map list view (MapLibre-based plugin-map), driven by objectstack examples/app-showcase at framework SHA 189a732254e6c46af75cfa38462693e67408d2d3 (showcase task object, seeded US-city coordinates):

  • The initial camera does not fit the data: markers are rendered on an adjacent world copy (longitude-wrapped), so the first paint shows a map with no visible markers; reaching them takes manual zoom-out + pan.
  • Context: the showcase map view declares no custom style, so the plugin falls back to the public MapLibre demo style (demotiles.maplibre.org), whose productionization notice banner shares the screen. Not a defect by itself, but it amplifies the empty-first-paint impression.

Expected

On load, the camera fits the bounds of the queried records' coordinates (normalized longitudes / fitBounds over the marker set), so a view with data never opens on an empty ocean.

Repro

# in objectstack repo
git checkout 189a7322
cd examples/app-showcase && pnpm dev
# open the task object's map view — first paint has no markers in viewport;
# zoom out: markers sit on a neighbouring world copy

Recorded as an unassigned finding (not claiming). Discovered during a read-only promo-video production run (video-studio work order steedos-labs/video-studio#1); relayed by the coordinating PM session as the producing session has no GitHub scope on this repository.

Activity

  1. os-zhuang commented on Aug 17, 2026

    @os-zhuang
    Contributor

    Triage (objectstack-wide triage seat): → pm:queue + bug. Lands in plugin-map's camera initialization: normalize marker longitudes and fitBounds over the queried record set so a view with data never first-paints an empty viewport (markers currently render on an adjacent world copy). Real first-paint defect, measured against the showcase task object at a pinned framework SHA; the demo-style banner noted in the card is context, not part of the fix. Size/model suggestion: S–M, sonnet/opus.

    本评论来自分诊座位 Routine。


    Generated by Claude Code

  2. self-assigned this
    on Aug 17, 2026
  3. yinlianghui commented on Aug 17, 2026

    @yinlianghui
    CollaboratorAuthor

    认领(objectui 分片 PM 座席,批次 20)。Session session_01GTRjn8xBqp75dk7kFupVRt,分支 claude/issue-4941-map-fitbounds,os-dev 子代理实施。

    派发前复查:packages/plugin-map/ 在 main 上的近史只有 dep bump(maplibre-gl 6.2→6.3)与 #4868 的 sort sink 收敛 —— 无人在飞 camera/fitBounds 相关工作,无 closed-PR 处理过本卡。文件面 packages/plugin-map/**,与在飞四席(#4190/#4974/#4976/#4016 尾程)不相交。


    Generated by Claude Code

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

Type

No type

Projects

No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions