Proposal: narrowly scoped account order API

未关闭
#73 0 条评论 0 个 reaction 已指派 0 人 在 GitHub 查看

@big-bay 已经在做这个了。

开始于 2026年11月12日。

  • #74 来自 @big-bay —— 未关闭

评估

难度
5/5
预计耗时
一周以上
新手友好度
25/100
Issue 类型
功能
描述清晰度
描述清楚
活跃度
活跃
技术栈
javascript
领域
api

调研方向

Read the proposed accountOrder.getOrder() and accountOrder.setOrder(accountIds) contract and the linked contribution description for the implementation, schema, example, documentation, and tests; the payload does not name their file paths. First assess whether the repository wants this experimental API and its permission model. Done means maintainers resolve the scope and permission questions and agree on a draft; the proposal says a pull request will link back here.

由索引模型根据 Issue 内容生成。

描述

Use case

Mail add-ons can read the account order through accounts.list(false), but the documented API has no corresponding write method. CSMailSort needs to let users explicitly arrange existing accounts while retaining Thunderbird's native account list and default account.

Proposed narrow experiment

  • accountOrder.getOrder(): Promise<string[]> — IDs of the same mail accounts exposed by accounts.list(false), in native order.
  • accountOrder.setOrder(accountIds): Promise<void> — persist a complete permutation of the current mail account IDs.

The draft uses Thunderbird's native MailServices.accounts.reorderAccounts. It validates the complete membership before writing, reads back the result, flushes preferences, and attempts to restore and verify the previous order if a write or persistence step fails. A failed recovery is reported distinctly. Non-mail accounts keep their visible slots; the native method preserves hidden accounts. The default account is not changed.

There is no window patching, replacement folder pane, folder ordering, grouping, message access, account creation, arbitrary preference access, network access, or startup reapplication in the experiment. The two methods require accountsRead; like all experiments, installation still requests full unrestricted access. A core write permission, if appropriate in the future, is outside this proposal.

Contribution

I have prepared an MPL-2.0 implementation, schema, read-only example, documentation, and dependency-free tests against the actual parent implementation. The repository lint and nine test groups pass, including invalid/stale membership, permission denial, excluded/hidden accounts, no-ops, default-account preservation, and write/readback/recovery errors. Native testing is being recorded separately; the mock tests are not a native UI compatibility claim.

The intended development target is Thunderbird 160. A pull request will link back here with the code and validation limits.

Feedback requested

Would this scope and complete-permutation contract be suitable for a draft in this repository? In particular, is the separate accountOrder namespace appropriate, and should a future core version use a dedicated write permission?

This is a proposal for discussion, not a claim of acceptance into Thunderbird core or approval for an ATN listing.

主要语言
JavaScript
星标
33
派生
18
PR 合并指标
30 天内没有已合并 PR

环境准备

  • 没有 Dockerfile 或 Docker Compose 文件
  • 没有 Pull Request 模板
  • 阅读贡献指南

从这里开始

  1. 先读完整个 Issue,再读项目的贡献指南。
  2. 在 Issue 下留言说明你要接手 —— 这能避免两个人做同样的事。
  3. Fork 仓库,在一个分支上完成修改。
  4. 提交 Pull Request,并在描述里引用这个 Issue 编号。

thunderbird/webext-experiments 的其他 Issue

查看 thunderbird/webext-experiments 的全部 Issue

相似的 Issue

更多 JavaScript Issue

把新 issue 发到你的邮箱

精选适合新手参与的 GitHub issue 摘要。