Repository navigation
Release: v0.8.0 plan and checks (was v0.7.1) #203
Description
Goes live only on a v* tag (curator tags; a tag overwrites the live site; no staging deploy).
Merged since v0.7.0: locus-map winner / set-aside marks and Mucoromycotina order / family (#171), logo (#175, #176), site wording, raw-reads pages for waves 1 and 2 with clone groups and by-study counts, idiomorph labels (output only, beside raw names), label registry and dedupe rule.
Needed
- CHANGELOG entry and release notes.
- A decision per open v0.7.1 item (4754_b80_PM_combined matMi: recorded span starts with a stop codon (off by one?), no CDS or protein #135, Flag polished gene models with a frameshift or internal stop as defective (pseudogene) #151, Revisit the Basidiomycota bLocus 10 kb intron cap: trade-offs and alternatives (follow-up to #92, #153) #154, Held out of v0.7.0: what each item needs to be ready for the next release #155, Run-to-run variance: same code and inputs give different polished gene models for equal-scoring copies #161, Rust genomes on the BFD suppress list are not searched (7 called in v0.6.0) #165-Landing page: show the Fola and A. fumigatus read-typing results like Coccidioides (charts, per-strain tables) #172): ship or move to v0.8.0; are Basidiomycota: 102 pheromone-receptor loci admitted through the B43 record's pheromone proteins #166-Basidiomycota: 14 MAT-family loci outside Tremellomycetes rest on STE3 plus PAN6 or RPL39 #170 (Basidiomycota) known issues again?
- Landing page: show the Fola and A. fumigatus read-typing results like Coccidioides (charts, per-strain tables) #172 (Fola and A. fumigatus on the landing page): blocking or not?
- Version stamp check: the version written in reports must equal the tag (lesson of detection_report records matpredict_version 0.6.1 whatever the code is #139).
git diff v0.7.0 <tag> -- src dbshows only intended changes (labels are output-only).- A local site build and a headless browser check of the species and overview pages; withheld species must be visibly marked before tagging.
- Rollback = re-tag.
Activity
hyphaltip commentedon Oct 29, 2026
Status 2026-10-28
git diff v0.7.0 main -- src dbis empty: labels, pages and notes are output or documentation only, so v0.7.1 is a site/documentation release on the v0.7.0 data.date-releasedin CITATION.cff; tag.hyphaltip commentedon Oct 29, 2026
Label work decided in #201 is part of v0.7.1: #210 (registry_version, frozen codes), #211 (label_status and marker), #212 (About wording, downloads), plus PR #208 (explorer switch and genome-table labels, which had been merged into a stacked branch and not main). Draft prep PR #207 should be merged after these.
[-]Release: v0.7.1 plan and checks[/-][+]Release: v0.8.0 plan and checks (was v0.7.1)[/+]hyphaltip commentedon Nov 2, 2026
v0.8.0 is released: tag
v0.8.0onmain(5d6ce82), release published 2026-11-02. The Pages deploy and the Docker release workflows succeeded; the live site shows "v0.8.0 (a2933f7, October 2026)" with 22,979 BFD genomes searched, 90.7% called and 27,023 loci, and the imageghcr.io/stajichlab/matpredictcarries the tags0.8.0,0.8,latest.The definitive run, the comparison against v0.7.0 and its attribution, and the known issues are in the changelog ("The definitive run", "Known issues"). Follow-ups: #227 and #228 (v0.8.1), and the timing benchmark for the 3 to 4 times longer wall time per genome (cause not yet known; calls are unaffected).