Speakr is a personal, self-hosted web application designed for transcribing audio recordings
Python
Mobile transcript: sticky speaker pills now actually stick — was iterating over the flat processedTranscription.simpleSegments and rendering each segment as its own scrollable block, which meant each segment was its own sticky containing block; once that segment scrolled past, the speaker pill went with it instead of pinning to the top of the scroll area until the speaker changed; now uses the same processedTranscription.simpleSegmentRuns structure desktop has used since the run grouping shipped — consecutive same-speaker segments are collected into a .speaker-run container with the speaker-tablet (or speaker-tablet-group when timestamps are on) as the first child of the run, and the existing CSS at styles.css line 1680 makes that first child position: sticky; top: 0 so the pill stays visible for the entire run; the inner segments retain their click-to-seek and double-click-to-edit handlers plus the active-playing highlight, with originalIndex used in place of the plain enumeration index so the audio sync still maps correctly back to the flat segment list
Murtaza Nasir826ca3707996
diff --git a/templates/components/detail/mobile-transcript-panel.html b/templates/components/detail/mobile-transcript-panel.htmlindex 8f913560..de266ff7 100644--- a/templates/components/detail/mobile-transcript-panel.html+++ b/templates/components/detail/mobile-transcript-panel.html@@ -126,17 +126,33 @@<!-- Transcription content (show regardless of processing state if transcription exists) --><div v-if="selectedRecording.transcription && !processedTranscription.isError">+ <!-- Speaker-run layout: consecutive same-speaker segments+ grouped under one .speaker-run so the speaker pill at+ the top of the run can stay sticky for the WHOLE run.+ Was iterating over the flat simpleSegments before,+ which meant each segment was its own sticky scope and+ the pill disappeared as soon as the first segment+ scrolled past — the "sticky pills don't work" symptom. --><div v-if="!processedTranscription.hasDialogue || transcriptionViewMode === 'simple'" class="transcription-simple-view"><div v-if="processedTranscription.hasDialogue">- <div v-for="(segment, index) in processedTranscription.simpleSegments" :key="`seg-${index}-${segment.startTime}`" :class="['speaker-segment', { 'active-playing-segment': currentPlayingSegmentIndex === index }]" @click="seekAudioFromEvent" @dblclick.stop="openAsrEditorAtSegment(index)" :data-start-time="segment.startTime" :data-segment-index="index">- <template v-if="segment.showSpeaker">- <div v-if="showTimestampsSimpleView" :class="['speaker-tablet-group', segment.color]">- <span class="speaker-tablet-time">${ formatTimestamp(segment.startTime) }</span>- <span class="speaker-tablet-name">${segment.speaker}</span>- </div>- <div v-else :class="['speaker-tablet', segment.color]">${segment.speaker}</div>- </template>- <div class="speaker-text">${segment.sentence}</div>+ <div v-for="(run, runIndex) in processedTranscription.simpleSegmentRuns"+ :key="`run-${runIndex}-${run.startTime}`"+ class="speaker-run">+ <div v-if="showTimestampsSimpleView"+ :class="['speaker-tablet-group', run.color]">+ <span class="speaker-tablet-time">${ formatTimestamp(run.startTime) }</span>+ <span class="speaker-tablet-name">${run.speaker}</span>+ </div>+ <div v-else :class="['speaker-tablet', run.color]">${run.speaker}</div>+ <div v-for="(segment, segIndex) in run.segments"+ :key="`seg-${segment.originalIndex}-${segment.startTime}`"+ :class="['speaker-segment', { 'active-playing-segment': currentPlayingSegmentIndex === segment.originalIndex }]"+ @click="seekAudioFromEvent"+ @dblclick.stop="openAsrEditorAtSegment(segment.originalIndex)"+ :data-start-time="segment.startTime"+ :data-segment-index="segment.originalIndex">+ <div class="speaker-text">${segment.sentence}</div>+ </div></div></div><div v-else>