This is an automated email from the ASF dual-hosted git repository. mengw15 pushed a commit to branch add-monthly-meeting-2026-07-28 in repository https://gitbox.apache.org/repos/asf/incubator-texera-site.git
commit 98790d84c8657620620a3d8c6a48dec02496d13b Author: mengw15 <[email protected]> AuthorDate: Wed Jul 29 14:15:11 2026 -0700 Add July 28, 2026 monthly meeting minutes --- content/monthly-meetings/2026-07-28.md | 121 +++++++++++++++++++++++++++++++++ 1 file changed, 121 insertions(+) diff --git a/content/monthly-meetings/2026-07-28.md b/content/monthly-meetings/2026-07-28.md new file mode 100644 index 00000000..779639a3 --- /dev/null +++ b/content/monthly-meetings/2026-07-28.md @@ -0,0 +1,121 @@ +--- +title: "July 28, 2026" +date: 2026-07-28 +weight: 1 +type: docs +toc: false +hide_feedback: true +description: > + Monthly sync meeting summary for July 28, 2026 +--- + +# Texera Apache Incubation – Monthly Meeting Minutes + +**Date:** July 28, 2026 +**Participants:** +* **Texera Committers:** Meng Wang, Xuan Gu, Ali Risheh, Chen Li, Jiadong Bai, Yicong Huang, Kunwoo Park +* **Texera Contributors:** Eugene Gu, Carlos Ernesto Alvarez Berumen, Yang Zhang, Kary Zheng, Tanishq Gandhi, Matthew Ball, Ryan Zhang, Ranjani Veena Belavadi, Sarah Asad, Suryaa +* **Apache Incubating Mentors:** Ian Maxon +* **Guests:** Haochao Ma + +--- + +## 1. Project and Community Development + +### Last Sync Meeting +- The previous sync meeting was held on June 30, 2026. The team reviewed v1.2.0-incubating release progress and major developments in role-based access control, the one-command local development stack, and the Hugging Face operator. + +### Community Activity +- July remained highly active, sustaining the May–June peak. Issue activity stayed strong, reflecting ongoing feature development and planning. The figures below cover July 1 through July 27. +- Excluding bot-authored activity, the project recorded 339 merged commits on the main branch, and 369 issues were opened. +- Counting all merged pull requests, including bot-authored changes and the release/v1.2 branch, 415 PRs were merged in July. Test-focused PRs were the largest group at 222, or 53%, followed by dependency bumps at 73 and fixes at 58. 20 PRs, or approximately 5%, introduced new features. +- PR size trended smaller compared with June. The mean merged PR shrank from 376 to 262 changed lines, and the median from 183 to 176. The share of PRs above 1,000 changed lines fell from 6% to 3%, and PRs of 500 to 999 lines fell from 11% to 8%. The team reiterated the goal of keeping feature PRs small and splitting large feature work into smaller, reviewable PRs. + +### Community Discussions and Contributor Onboarding +- The dev@ mailing list had 303 emails across 45 threads in July, including design RFCs, release votes, and announcements. +- GitHub Discussions saw 27 new threads, most of them in the Ideas category with 15, followed by General with 10. +- The project welcomed 5 first-time contributors in July. + +### Contributor Activity in July +- There were 27 total contributors in July, including 5 new contributors. + +--- + +## 2. v1.2.0-incubating Release Status + +- The v1.2.0-incubating release remains in progress. +- Release timeline: + - The release/v1.2 branch was cut on June 3. + - RC1 was prepared on June 12 and later canceled. + - RC2 was prepared on June 24, and one issue was identified during validation. + - RC3 passed the PPMC vote on July 15. + - The RC3 IPMC vote was opened on July 23, and a license header issue was identified. + - RC4 was in preparation as of July 27. +- The team thanked everyone who helped review, test, validate, and vote on the release, with special thanks to Xuan Gu, Yicong Huang, Xinyuan Lin, and Jiadong Bai for preparing the release. + +--- + +## 3. Major Developments + +### Loop Operators +Texera added its first control-flow construct for iterative workflows. Recent work included: +- Adding loop operators for iterative control flow inside workflows (#5700). +- Carrying loop-bookkeeping columns through materialized State (#5900). + +### Kubernetes and External Storage +Object-storage backends are now configurable for cloud and Kubernetes deployments. This work concerns the deployment-level storage backend and is separate from per-user warehouse storage. Recent work included: +- Making app-tier object storage configurable via storage.s3 (#5932). +- Supporting external S3 for LakeFS and Lakekeeper, with MinIO now optional (#6295). + +### Hugging Face Operator +The Hugging Face operator lets users run pretrained models from the Hugging Face Hub directly inside Texera workflows. Recent work included: +- Adding task-aware field visibility and preview to the property editor (#5568). +- Adding media output rendering in the result panel (#5675). + +### Config Service and Feature Flags +- Configuration continues to be consolidated behind a dedicated microservice. The site_settings API was moved from texera-web to config-service (#6116), giving a cleaner separation of runtime settings from the web tier. +- The goal of this effort is to make features configurable from the UI. + +### Python Notebook Migration Tool (Proposal #5959) +A tool to migrate Jupyter notebooks into Texera workflows. Recent work included: +- Adding the notebook-migration orchestration service (#5262), which drives AI-assisted notebook-to-workflow generation through the LLM proxy. + +The next step is an embedded Jupyter panel in the workflow editor. + +### Operator Result Cache (Proposal #5880) +The operator result cache reuses cached operator output so that re-runs skip already-computed regions. Recent work included: +- Adding the operator_port_cache storage table (#5967). +- Clearing the legacy cache path ahead of the MVP rewrite (#6126). +- Adding cache-reused operator state and cached-region statistics (#6729). + +The MVP is in progress under epic #5881. Next steps are scheduler planning, a cache panel, and cache invalidation. + +--- + +## 4. Quality, Testing, and Benchmarks + +### CI and Integration Testing +- A new platform-integration CI job boots every service against real infrastructure (#6274). + +### Testing and Coverage +- Testing continued to improve in July. 222 of the 415 PRs merged in July, or 53%, were test-focused. +- Project coverage improved to 78.94%, up from 58.44% at the previous sync meeting. + +--- + +## 5. Ongoing Work and Next Steps + +### User-Provided ML Models (Proposal #6616) +This effort will bring user-trained models into Texera workflows, alongside the Hugging Face operator. Planned work is tracked in: +- Model metadata tables, file storage, and path resolution (#6496, #6497). +- Model management API and UI (#6498, #6499). +- Using and sharing models inside workflows (#6500, #6501). + +The proposal is in design and early implementation. + +### Multi-Source Authentication (Proposal #6716) +- This effort will decouple login from the user record so that a single account can be used with multiple identity providers, for example Facebook. +- Today the user and login records are tightly coupled. Under the new design, one user can sign in to the same account through multiple vendors. + +### Other Ongoing Efforts +The team will continue advancing workflow runtime observability, an admin dashboard for computing units, workflow templates, Python-to-workflow compilation, per-user warehouse storage, Kubernetes autoscaling and idle cleanup, public workflow moderation and reporting, and Java 21 and Python upgrades.
