Tags: david-priest/GateLab
Tags
Rewrite the brand card in plain prose, and say what GateLab does not … …do (#108) (#33) The card read as machine-written: em-dashes several to a paragraph, bold-label bullets, and a worked example nobody asked for. It is the first thing anyone reads about the app. Three changes: - Plain paragraphs throughout. What each thing is, then why it matters. - The FlowJo anecdote is gone. Quoting one of our own gates moving from 43.5% to 80.5% of its parent made a general design point sound like a grievance about a specific file, and it put a number in front of a reader who has no way to check it. - A new paragraph on scope, which was the real gap: GateLab applies ONE population hierarchy to every checked file. FlowJo keeps a separate population tree per FCS, and Cytobank can tailor a gate's coordinates per file. Someone arriving from either will look for that and should be told plainly that it is absent, and that it can be added if it is wanted. No version bump. This is copy. Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
Shorten the curved-edge note, and let it be dismissed (#31) Folded into v0.6.0 rather than cut as a patch: the release is minutes old and this is the same feature's presentation, not a change in behaviour. Two lines of standing explanation pushed the plot down and kept explaining something you only need told once. Now a single sentence with a dismiss button, remembered across sessions; the full explanation stays on the Gate edges control's tooltip. Port of GateLab-dev #106. Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
Release v0.5.0 — open FlowJo workspaces, and choose the compensation … …matrix (#29) Port of GateLab-dev #102 and #103; the tree is identical to dev master. .wsp becomes a file type GateLab opens: it resolves the workspace's samples against the files it is given, asks which strategy to import where a sample holds several, and asks which compensation matrix to use where the FCS and the workspace disagree. Marked beta — exercised against two real workspaces, both FlowJo 10.9/10.10 on BD instruments; other vendors and older versions are untested. The compensation choice can change where fluorescence gates land, which is the point: a FACSDiva export stores the acquisition matrix while the operator's adjustment lives only in the workspace, and GateLab previously used the former without comment. Polygon outlines now follow their true transformed boundary rather than chords between transformed vertices. Gating and exports are untouched. Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
PreviousNext