[2026-11-21] eagle: personal/projects/personal-os/personal-os-day-history.md personal/projects/personal-os/personal-os-evidence/2026-11-19/follow-up-live-acceptance.md personal/projects/personal-os/personal-os-production-acceptance.md personal/projects/personal-os/personal-os-project-management.md personal/projects/personal-os/personal-os-remediation-plan-2026-11-18.md personal/projects/personal-os/personal-os-user-scenarios.md personal/projects/personal-os/project-management-rules.json work/zoom-transcripts/2026-10-06 Dominik Kapusta's Personal Meeting Room 2026-10-06 15 12(GMT+6 00) NtgNJ-Y-Qxy5sMGMjngukw/transcript.json work/zoom-transcripts/2026-10-06 Dominik Kapusta's Personal Meeting Room 2026-10-06 15 12(GMT+6 00) NtgNJ-Y-Qxy5sMGMjngukw/transcript.md work/zoom-transcripts/2026-10-06 Jacek Łyp's Zoom Meeting 2026-10-06 20 44(GMT+6 00) gUhtLoqRTfSsTbZsann73A/transcript.json

This commit is contained in:
Alexey Martemyanov committed 2026-11-21 03:36:41 +06:00
1 parent 0a6b07f558
commit 07b614012d
13 files changed
+3609 -11

No files matched your search

@@ -10,6 +10,7 @@
- Для черновика статуса заголовок обозначает тип события и задачу. Прямые ссылки открывают коммит и цель Asana. Возврат открывает то же событие, из которого получен черновик. [Рабочая проверка FaviconStore](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#черновик-статуса-faviconstore) - Для черновика статуса заголовок обозначает тип события и задачу. Прямые ссылки открывают коммит и цель Asana. Возврат открывает то же событие, из которого получен черновик. [Рабочая проверка FaviconStore](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#черновик-статуса-faviconstore)
- В интервале с установленной связью видны результаты и действия той же задачи за день, каждое со своим временем. Переход к событию и возврат сохраняют выбранную карточку. Для интервала без связи список не показывается. [Проверка Xcode](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#результаты-по-задаче-из-интервала-xcode) - В интервале с установленной связью видны результаты и действия той же задачи за день, каждое со своим временем. Переход к событию и возврат сохраняют выбранную карточку. Для интервала без связи список не показывается. [Проверка Xcode](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#результаты-по-задаче-из-интервала-xcode)
- Повторное открытие Asana-задачи берётся из события `marked_incomplete` с подтверждённым автором. В карточке видны точное время, событие и задача. PR показывается при единственной явной ссылке в описании. Текущий статус задачи не меняет прошедшее событие. [Проверка PR #7784](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#нумерованный-ответ-и-повторное-открытие-pr-7784) - Повторное открытие Asana-задачи берётся из события `marked_incomplete` с подтверждённым автором. В карточке видны точное время, событие и задача. PR показывается при единственной явной ссылке в описании. Текущий статус задачи не меняет прошедшее событие. [Проверка PR #7784](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#нумерованный-ответ-и-повторное-открытие-pr-7784)
- Совпавшее событие повторного открытия из дневной сводки и зеркала Asana показано один раз. Сохранённые ссылки на прежний формат эпизода открывают ту же смену статуса. [Проверка CPM](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#одно-событие-повторного-открытия-cpm)
- В раскрытии события видны что сделано, решения, оставшиеся вопросы и выходы. Результат может быть коммитом, ревью текущего HEAD, изменённым документом, действием «Рабочей очереди» или созданной Asana-задачей. Точный комментарий, фрагмент встречи, PR, журнал и документ открываются из этого же блока. - В раскрытии события видны что сделано, решения, оставшиеся вопросы и выходы. Результат может быть коммитом, ревью текущего HEAD, изменённым документом, действием «Рабочей очереди» или созданной Asana-задачей. Точный комментарий, фрагмент встречи, PR, журнал и документ открываются из этого же блока.
- Для каждого возникшего действия показаны состояние, источник, точный следующий шаг, ожидаемый результат и целевой артефакт. Показывается разница между предложенной задачей, сохранённым действием в «Рабочей очереди» и реально созданной записью Asana с её ссылкой. Если одно действие подтверждено в нескольких источниках, хронология показывает его один раз с нужными связями. Ссылки открывают предыдущие версии и дату, на которой действительно произошли события. - Для каждого возникшего действия показаны состояние, источник, точный следующий шаг, ожидаемый результат и целевой артефакт. Показывается разница между предложенной задачей, сохранённым действием в «Рабочей очереди» и реально созданной записью Asana с её ссылкой. Если одно действие подтверждено в нескольких источниках, хронология показывает его один раз с нужными связями. Ссылки открывают предыдущие версии и дату, на которой действительно произошли события.
@@ -1469,3 +1469,18 @@
- Первичная разметка Asana подтвердила Bartek в реплике Bartosz, Alex M и Michal в реплике Bartek. Рабочий E5 за 62,376 секунды опубликовал цитаты с именами, причину удержания и точный переход в каждый комментарий. При открытии и перезагрузке на ширинах 1280 и 390 px проверены имена, отсутствие сырых профильных URL и команд отвечать. В подтверждённом запросе от Anh Do остались конкретные шаги. Контрольный пример без подтверждённого имени показывает «участник Asana». [Реплика Bartosz](https://app.asana.com/0/0/1219967170800092#comment-1220640466680864) · [Реплика Bartek](https://app.asana.com/0/0/1219967170800092#comment-1219967333892523) · [Открытый запрос](http://127.0.0.1:4317/#/task-queue?action=asana-story%3A1220705961663640) · [Рабочий E5](http://127.0.0.1:8000/run/cb26d7d5-c46e-4e29-a046-a701e0338228?workspace=personal_os) - Первичная разметка Asana подтвердила Bartek в реплике Bartosz, Alex M и Michal в реплике Bartek. Рабочий E5 за 62,376 секунды опубликовал цитаты с именами, причину удержания и точный переход в каждый комментарий. При открытии и перезагрузке на ширинах 1280 и 390 px проверены имена, отсутствие сырых профильных URL и команд отвечать. В подтверждённом запросе от Anh Do остались конкретные шаги. Контрольный пример без подтверждённого имени показывает «участник Asana». [Реплика Bartosz](https://app.asana.com/0/0/1219967170800092#comment-1220640466680864) · [Реплика Bartek](https://app.asana.com/0/0/1219967170800092#comment-1219967333892523) · [Открытый запрос](http://127.0.0.1:4317/#/task-queue?action=asana-story%3A1220705961663640) · [Рабочий E5](http://127.0.0.1:8000/run/cb26d7d5-c46e-4e29-a046-a701e0338228?workspace=personal_os)
- После обновления портала проверены 16 правил сопровождения, неподтверждённый слот iOS Kick-off и три ближайшие встречи. Времена O‑B Weekly, Career Advisor и Apple Weekly совпали с Asana на «Сегодня» и «Неделе» после открытия и перезагрузки в двух размерах окна. Ссылка на условия из действующего шаблона Midmortem по-прежнему возвращает HTTP 404. Доступ к почте не использовался. [Kick-off](https://app.asana.com/0/0/1220371473869152) · [Шаблон Midmortem](https://app.asana.com/0/0/1203886245376707) · [Недоступные условия](https://app.asana.com/0/0/1206092209398857) - После обновления портала проверены 16 правил сопровождения, неподтверждённый слот iOS Kick-off и три ближайшие встречи. Времена O‑B Weekly, Career Advisor и Apple Weekly совпали с Asana на «Сегодня» и «Неделе» после открытия и перезагрузки в двух размерах окна. Ссылка на условия из действующего шаблона Midmortem по-прежнему возвращает HTTP 404. Доступ к почте не использовался. [Kick-off](https://app.asana.com/0/0/1220371473869152) · [Шаблон Midmortem](https://app.asana.com/0/0/1203886245376707) · [Недоступные условия](https://app.asana.com/0/0/1206092209398857)
## Одно событие повторного открытия CPM
- До правки живая история возвращала два события с датой, временем и ссылкой Asana `1220556402018013`. Сводка и зеркало использовали разные идентификаторы для той же смены статуса. Первичная Asana подтвердила три личных события 16 ноября: закрытие в 13:14:50, повторное открытие в 14:29:23 и закрытие в 14:43:46 по Бишкеку. [Первое закрытие](https://app.asana.com/0/0/1220422645861900#comment-1220554650704489) · [Открытие](https://app.asana.com/0/0/1220422645861900#comment-1220556402018013) · [Второе закрытие](https://app.asana.com/0/0/1220422645861900#comment-1220556557507896)
- После правки каждая из трёх смен статуса встречается один раз по своему идентификатору. После открытия и перезагрузки при 1280 и 390 px сверены порядок, точные секунды и исходные ссылки. Прежний адрес эпизода PR #7784 от 11 ноября открывает ту же смену статуса. На контроле проверены отсутствие одного источника и несовпавшее время. [История CPM](http://127.0.0.1:4317/#/work-history?date=2026-11-16&episode=asana%3Astory%3A1220556402018013) · [Прежняя ссылка PR #7784](http://127.0.0.1:4317/#/work-history?date=2026-11-11&episode=asana-reopened%3A1220418204432230)
## План связи DRI и PA по шаблону
- Asana подтвердила пустое поле Project Advisor у iOS Top Priority, открытую дочернюю нить и полный список без комментариев. Действующие шаблоны спрашивают, нужен ли регулярный ритм, по какому каналу общаться или как связываться по необходимости. Версия 16 показывает назначение Advisor и план связи отдельными шагами. [Проект](https://app.asana.com/0/0/1220371472571114) · [Нить](https://app.asana.com/0/0/1220371386845570) · [Шаблон](https://app.asana.com/0/0/1207685708094386) · [Стейкхолдеры](https://app.asana.com/0/0/1207268731049844)
- Рабочий E5 завершился за 51,959 секунды. При 1280 и 390 px дважды подтверждены шаги, ссылка нити, разные подписи источников в карточке и правиле, и пустой фиксированный интервал. Контрольный путь с уже отправленным запросом сохраняет ожидание. Существующее активное поручение по закладкам и контекстные комментарии Shift снова сверены после обновления портала. [Рабочий E5](http://127.0.0.1:8000/run/ba95022b-5fc1-41b2-aa5d-c064244e4eb2?workspace=personal_os) · [Карточка](http://127.0.0.1:4317/#/task-queue?category=project-management&action=project-management%3Aadvisor-thread%3A1220371472571114%3A1220371386845570)
- В доступном общем README AoR речь идёт о поддержке. Шаблон AoR описывает назначение зоны. Требование Running или тег macOS App в этих источниках отсутствует. Проектный поиск по Running не дал совпадений, а связанное в README руководство поддержки вернуло HTTP 404. Сохраняется проверенное размещение в Incoming. [README](https://app.asana.com/0/0/10148630051897) · [Шаблон AoR](https://app.asana.com/0/0/1207925924931741) · [Руководство](https://app.asana.com/0/0/1182794228567306)
## Текущий фокус и граница недели
- В субботнем обзоре 21 ноября прежний заголовок «Главное на неделе» относил к текущей неделе проект со стартом в понедельник 23 ноября. Asana подтвердила `start_on`, а Daily Status — выбор дочернего Kick-off как 1‑Must и проекта как Top Priority. При 1280 и 390 px после открытия и перезагрузки видны один блок «Текущий фокус», дата старта, обе ссылки и отдельный «План до конца недели». Контрольный путь с обновлением даты сохраняет фокус ссылки. [Неделя](http://127.0.0.1:4317/#/weekly-review) · [Daily Status](https://app.asana.com/0/0/1202477079846579) · [Проект](https://app.asana.com/0/0/1220371472571114)
@@ -246,6 +246,9 @@
136. C-S-S в рабочей очереди показывает согласованный порядок, ожидание и две рабочие ссылки. Старый запрос Shane снят после проверки ответов и текущего Top Priority. При 1280 и 390 px дважды открыты карточка, ссылка на решение и исходное обсуждение. [Проверка](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#согласованный-порядок-c-s-s-при-паузе) 136. C-S-S в рабочей очереди показывает согласованный порядок, ожидание и две рабочие ссылки. Старый запрос Shane снят после проверки ответов и текущего Top Priority. При 1280 и 390 px дважды открыты карточка, ссылка на решение и исходное обсуждение. [Проверка](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#согласованный-порядок-c-s-s-при-паузе)
137. В очереди снят запрос Bartek по пунктам #2–#5. Открытые PR #7784 и анализ SERP сохранены. В истории показано повторное открытие PR-задачи 11 ноября в 14:26:37 с точными ссылками. Оба пути дважды проверены при 1280 и 390 px. [Проверка](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#нумерованный-ответ-и-повторное-открытие-pr-7784) 137. В очереди снят запрос Bartek по пунктам #2–#5. Открытые PR #7784 и анализ SERP сохранены. В истории показано повторное открытие PR-задачи 11 ноября в 14:26:37 с точными ссылками. Оба пути дважды проверены при 1280 и 390 px. [Проверка](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#нумерованный-ответ-и-повторное-открытие-pr-7784)
138. Удержанные комментарии Shift показывают Bartek, Alex M и Michal по исходной разметке Asana. Состояние и прямые ссылки сохранены, указания ответить сняты. Подтверждённый запрос по закладкам оставлен с конкретными шагами. При 1280 и 390 px проверены открытие и перезагрузка. [Проверка](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#имена-и-контекст-в-удержанных-комментариях) 138. Удержанные комментарии Shift показывают Bartek, Alex M и Michal по исходной разметке Asana. Состояние и прямые ссылки сохранены, указания ответить сняты. Подтверждённый запрос по закладкам оставлен с конкретными шагами. При 1280 и 390 px проверены открытие и перезагрузка. [Проверка](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#имена-и-контекст-в-удержанных-комментариях)
139. У техдизайна CPM 16 ноября показаны два закрытия и одно повторное открытие. Прежний адрес эпизода PR #7784 продолжает работать. На живом портале обе даты дважды проверены при 1280 и 390 px. [Проверка](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#одно-событие-повторного-открытия-cpm)
140. Версия 16 предлагает согласовать Advisor для iOS и способ связи. При 1280 и 390 px дважды сверены поля проекта, нить и две разные ссылки на первоисточники. Шаблон не назначает обязательный созвон или интервал. [Проверка](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#план-связи-dri-и-pa-по-шаблону)
141. Недельный обзор отделяет текущие назначения Asana от плана до конца недели. В живом обзоре 21 ноября начало iOS Top Priority 23 ноября и обе ссылки проверены дважды при 1280 и 390 px. [Проверка](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#текущий-фокус-и-граница-недели)
## Как закрывается проверка ## Как закрывается проверка
@@ -66,7 +66,7 @@ Milestone **D18-201** добавляет в «Рабочую очередь» о
В [«Рабочей очереди»](http://127.0.0.1:4317/#/task-queue?category=project-management) доступны 16 исходных правил и редактор. Можно добавить правило, выбрать стадию и категории, указать условие, инструкцию, период и критерий закрытия, а также приостановить или восстановить исходное правило. Текущий расчёт сохраняет версию правил. Изменение настроек запускает новый расчёт. При конфликте двух вкладок сохраняется уже записанная версия. В [«Рабочей очереди»](http://127.0.0.1:4317/#/task-queue?category=project-management) доступны 16 исходных правил и редактор. Можно добавить правило, выбрать стадию и категории, указать условие, инструкцию, период и критерий закрытия, а также приостановить или восстановить исходное правило. Текущий расчёт сохраняет версию правил. Изменение настроек запускает новый расчёт. При конфликте двух вкладок сохраняется уже записанная версия.
В рабочей версии 14 у каждого из 16 базовых правил подключено хотя бы одно автоматическое условие. Для Running, обязательных тегов и общих порогов Midmortem ещё требуются первичные правила. Сценарий на отдельном наборе подтвердил снятие предложения при паузе, возвращение после восстановления и отказ устаревшего запроса. Допуск предложения в оба плана проверен на копии. В рабочем журнале решение пока отсутствует. В рабочей версии 16 у каждого из 16 базовых правил подключено хотя бы одно автоматическое условие. Для Running, обязательных тегов и общих порогов Midmortem ещё требуются первичные правила. Сценарий на отдельном наборе подтвердил снятие предложения при паузе, возвращение после восстановления и отказ устаревшего запроса. Допуск предложения в оба плана проверен на копии. В рабочем журнале решение пока отсутствует.
Правило «Обновить статус проекта» проверяет действующий Top Priority после даты старта. При заданном интервале оно показывает последний статус владельца и до трёх работ, завершённых после него. Без настроенного интервала, при неоднозначном обсуждении и до начала работ напоминания нет. Статусы AoR остаются отдельной проверкой. [Проверка CPM и рабочей версии 10](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#регулярное-обновление-статуса-проекта) Правило «Обновить статус проекта» проверяет действующий Top Priority после даты старта. При заданном интервале оно показывает последний статус владельца и до трёх работ, завершённых после него. Без настроенного интервала, при неоднозначном обсуждении и до начала работ напоминания нет. Статусы AoR остаются отдельной проверкой. [Проверка CPM и рабочей версии 10](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#регулярное-обновление-статуса-проекта)
@@ -102,4 +102,6 @@ Milestone **D18-201** добавляет в «Рабочую очередь» о
Проверка пустого Project Advisor использует состояние проекта и дочернюю нить DRI ↔ PA. Закрытие Kick-off оставляет выбор Advisor или ожидание ответа в очереди. Закрытие проекта снимает шаг. Сверены рабочий iOS Top Priority и контрольные копии iOS и C-S-S. [Доказательства](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#project-advisor-после-закрытия-kick-off) Проверка пустого Project Advisor использует состояние проекта и дочернюю нить DRI ↔ PA. Закрытие Kick-off оставляет выбор Advisor или ожидание ответа в очереди. Закрытие проекта снимает шаг. Сверены рабочий iOS Top Priority и контрольные копии iOS и C-S-S. [Доказательства](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#project-advisor-после-закрытия-kick-off)
Действующий шаблон DRI ↔ PA предлагает выбрать частоту и канал либо договориться о связи по необходимости. Пока Advisor iOS не выбран и в нити нет обсуждения, Personal OS предлагает записать этот план вместе с назначением. Названия ссылок различают раздел о стейкхолдерах и шаблон нити. [Рабочая проверка](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#план-связи-dri-и-pa-по-шаблону)
После назначения Advisor правило сверяет заданный интервал с последним подтверждённым обновлением DRI. Если обновлений ещё нет, отсчёт возможен от подтверждённой даты начала проекта. Короткая реакция период не сбрасывает. При неполной нити или свежем содержательном комментарии без однозначной отметки статуса напоминание удерживается. В рабочем правиле интервал не задан. [Контроль CPM](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#настраиваемый-интервал-dri--pa) После назначения Advisor правило сверяет заданный интервал с последним подтверждённым обновлением DRI. Если обновлений ещё нет, отсчёт возможен от подтверждённой даты начала проекта. Короткая реакция период не сбрасывает. При неполной нити или свежем содержательном комментарии без однозначной отметки статуса напоминание удерживается. В рабочем правиле интервал не задан. [Контроль CPM](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#настраиваемый-интервал-dri--pa)
@@ -837,7 +837,9 @@ Zoom используется для транскриптов состоявши
### D18-171 · Дата начала приоритета в «Неделе» ### D18-171 · Дата начала приоритета в «Неделе»
- «Главное на неделе» показывает подтверждённое начало по Asana, если приоритет назначен на будущую дату. Дата и исходная ссылка видны при раннем и полном чтении. Изменение даты сохраняет фокус ссылки. Отметки 1‑Must и Top Priority берутся из подтверждённого расчёта. [Проверка](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md) - Называть раздел текущим фокусом, поскольку подтверждённые 1‑Must и Top Priority могут начинаться в следующей неделе. На обзоре 21 ноября сверить начало проекта 23 ноября в Asana, исходную ссылку и отдельный план до конца недели.
- «Текущий фокус» показывает подтверждённое начало по Asana, если приоритет назначен на будущую дату. Дата и исходная ссылка видны при раннем и полном чтении. Изменение даты сохраняет фокус ссылки. Отметки 1‑Must и Top Priority берутся из подтверждённого расчёта. [Проверка](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#текущий-фокус-и-граница-недели)
### D18-172 · Учёт времени обслуживания ### D18-172 · Учёт времени обслуживания
@@ -1017,6 +1019,10 @@ Zoom используется для транскриптов состоявши
### D18-201 · Сопровождение работ после загрузки источников ### D18-201 · Сопровождение работ после загрузки источников
- Исправить предложение DRI ↔ PA для действующего iOS Top Priority. Шаблон требует согласовать ритм и канал связи или план связи по необходимости. Проверить, что карточка не требует созвона или фиксированной частоты без договорённости.
- У ссылок в предложении DRI ↔ PA показать названия источников. Сейчас одинаковая подпись «Правило сопровождения» скрывает разницу между разделом о стейкхолдерах и шаблоном нити.
- Версия 16 показывает шаг согласования Advisor, план связи и две различимые ссылки. У текущей нити iOS отсутствуют комментарии и согласованный интервал. [Доказательства](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#план-связи-dri-и-pa-по-шаблону)
- Для приостановленного C-S-S сверить вопрос Shane, ответ владельца, подтверждение Shane и решение автора проекта изменить порядок. Убедиться, что выбранный первым проект совпадает с текущим Top Priority, а работа Shane остаётся открытой. Снять старое поручение об ответе и показать у паузы две точные рабочие ссылки. При возобновлении проекта или изменении решения вернуть обычную проверку. - Для приостановленного C-S-S сверить вопрос Shane, ответ владельца, подтверждение Shane и решение автора проекта изменить порядок. Убедиться, что выбранный первым проект совпадает с текущим Top Priority, а работа Shane остаётся открытой. Снять старое поручение об ответе и показать у паузы две точные рабочие ссылки. При возобновлении проекта или изменении решения вернуть обычную проверку.
- Рабочий C-S-S показывает согласованный порядок и две подтверждённые ссылки. Старый вопрос Shane доступен в исходной задаче. [Доказательства](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#согласованный-порядок-c-s-s-при-паузе) - Рабочий C-S-S показывает согласованный порядок и две подтверждённые ссылки. Старый вопрос Shane доступен в исходной задаче. [Доказательства](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#согласованный-порядок-c-s-s-при-паузе)
@@ -1051,6 +1057,9 @@ Zoom используется для транскриптов состоявши
### D18-202 · Графическая хронология работы и возникших задач ### D18-202 · Графическая хронология работы и возникших задач
- Объединять повторное открытие из сводки и зеркала по точному событию Asana. Проверить три последовательных смены статуса техдизайна CPM 16 ноября, сохранение старого адреса эпизода 11 ноября и работу через один доступный источник.
- В истории 16 ноября подтверждены два закрытия и одно повторное открытие CPM. Прежняя ссылка эпизода PR #7784 продолжает открываться. [Доказательства](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#одно-событие-повторного-открытия-cpm)
- Показывать подтверждённое повторное открытие Asana-задачи на времени исходного события. Проверить PR #7784 от 11 ноября с точной ссылкой на смену статуса. Чужое закрытие и текущий статус задачи не создают события открытия. - Показывать подтверждённое повторное открытие Asana-задачи на времени исходного события. Проверить PR #7784 от 11 ноября с точной ссылкой на смену статуса. Чужое закрытие и текущий статус задачи не создают события открытия.
- В истории 11 ноября виден результат в 14:26:37 и три ссылки на событие, Asana-задачу и PR #7784. [Доказательства](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#нумерованный-ответ-и-повторное-открытие-pr-7784) - В истории 11 ноября виден результат в 14:26:37 и три ссылки на событие, Asana-задачу и PR #7784. [Доказательства](personal-os-evidence/2026-11-19/follow-up-live-acceptance.md#нумерованный-ответ-и-повторное-открытие-pr-7784)
@@ -361,6 +361,8 @@
### U13.1 — неделя ### U13.1 — неделя
- В субботнем обзоре открыть текущий Top Priority, запланированный на следующий понедельник. Дата Asana видна в разделе текущего фокуса. По ссылке открывается этот проект. План до конца недели остаётся отдельно.
- Вход и предусловия: Проверенный граф с дедлайном, общим предшественником, условной цепочкой, циклом и отклонённым подготовительным шагом. - Вход и предусловия: Проверенный граф с дедлайном, общим предшественником, условной цепочкой, циклом и отклонённым подготовительным шагом.
- Действия через обычный путь: Открыть каждый день недели; вручную сверить порядок, human/review лимиты, окна и резерв по исходному графу. - Действия через обычный путь: Открыть каждый день недели; вручную сверить порядок, human/review лимиты, окна и резерв по исходному графу.
- Ожидаемый результат: сроки, зависимости, human/review бюджет и резерв соблюдены на каждом дне, включая условную цепочку, общий предшественник, цикл и отклонённый подготовительный шаг. - Ожидаемый результат: сроки, зависимости, human/review бюджет и резерв соблюдены на каждом дне, включая условную цепочку, общий предшественник, цикл и отклонённый подготовительный шаг.
@@ -795,6 +797,8 @@
### U24.5 — Регулярное обновление проекта и DRI ↔ PA ### U24.5 — Регулярное обновление проекта и DRI ↔ PA
- Открыть действующее предложение iOS при пустом поле Advisor. Сверить шаблон DRI ↔ PA, поле проекта и пустую нить. Шаг должен предложить согласовать Advisor и план связи. Допустимы асинхронный канал и связь по необходимости.
- В раскрытии предложения и правила открыть раздел о стейкхолдерах и шаблон нити. Подписи различимы до перехода, обе ссылки открывают точные источники.
- На копии активного CPM взять заданный интервал и последнее содержательное сообщение вида Status Nov/17 из карточки проекта. Проверить порог в день возникновения действия, недавние подтверждённые работы и прямую ссылку на сообщение. Новый статус снимает действие. Реакция, комментарий другого автора, неполное чтение, будущий старт и пауза не используются как основания для просрочки. - На копии активного CPM взять заданный интервал и последнее содержательное сообщение вида Status Nov/17 из карточки проекта. Проверить порог в день возникновения действия, недавние подтверждённые работы и прямую ссылку на сообщение. Новый статус снимает действие. Реакция, комментарий другого автора, неполное чтение, будущий старт и пауза не используются как основания для просрочки.
- Настроить частоту и проверить свежий содержательный комментарий, старый комментарий, реакцию и отсутствующую нить. На копии добавить нужное обновление. Карточка закрывается при свежем доказательстве правильной нити; неизвестная частота выводится как требующая уточнения. - Настроить частоту и проверить свежий содержательный комментарий, старый комментарий, реакцию и отсутствующую нить. На копии добавить нужное обновление. Карточка закрывается при свежем доказательстве правильной нити; неизвестная частота выводится как требующая уточнения.
@@ -866,6 +870,8 @@
### U25.4 — Работа над PR и точное ревью ### U25.4 — Работа над PR и точное ревью
- В истории 16 ноября открыть техдизайн CPM. Сверить два закрытия и одно повторное открытие с точными часами и ссылками Asana. Открыть прежнюю ссылку на эпизод PR #7784 11 ноября. Снятие одного источника не удваивает остальные события.
- Открыть историю 11 ноября и найти повторное открытие Asana-задачи PR #7784. Сверить автора, секунды, ссылку на событие и переход в PR. На 9 ноября закрытие другим участником не показывается как личное повторное открытие. После перезагрузки проверяется тот же эпизод. - Открыть историю 11 ноября и найти повторное открытие Asana-задачи PR #7784. Сверить автора, секунды, ссылку на событие и переход в PR. На 9 ноября закрытие другим участником не показывается как личное повторное открытие. После перезагрузки проверяется тот же эпизод.
- Открыть PR-эпизод, HEAD, комментарий и следующий шаг. Проверить роли автора и ревьюера, состояние обсуждения и источник. Просмотр CI, чужое ревью и более поздний HEAD не становятся выполненным личным действием на предыдущей версии. - Открыть PR-эпизод, HEAD, комментарий и следующий шаг. Проверить роли автора и ревьюера, состояние обсуждения и источник. Просмотр CI, чужое ревью и более поздний HEAD не становятся выполненным личным действием на предыдущей версии.
@@ -1,5 +1,5 @@
{ {
"version": 14, "version": 16,
"rules": [ "rules": [
{ {
"id": "project-registration", "id": "project-registration",
@@ -102,12 +102,16 @@
"stage": "active", "stage": "active",
"title": "Согласовать Project Advisor и связь в DRI ↔ PA", "title": "Согласовать Project Advisor и связь в DRI ↔ PA",
"condition": "До начала или во время активного проекта проверяются Project Advisor и назначенная нить.", "condition": "До начала или во время активного проекта проверяются Project Advisor и назначенная нить.",
"agent": "Проверь поле Advisor и нить DRI ↔ PA. При заданном интервале сопоставь его с последним подтверждённым обновлением DRI или датой начала проекта, если обновлений ещё не было.", "agent": "Проверь поле Advisor и нить DRI ↔ PA. Сверь согласованную частоту и канал обновлений либо план связи по необходимости. При настроенном интервале сравни его с последним обновлением DRI или датой начала проекта, если обновлений ещё не было.",
"completion": "Advisor указан, нужная нить однозначна, актуальное обновление содержит результат, вопрос или блокер в требуемом окне.", "completion": "Advisor указан, в нити записан план связи. При настроенном интервале актуальное обновление содержит результат, вопрос или блокер.",
"sources": [ "sources": [
"https://app.asana.com/0/0/1207268731049844", "https://app.asana.com/0/0/1207268731049844",
"https://app.asana.com/0/0/1207685708094386/f/" "https://app.asana.com/0/0/1207685708094386/f/"
] ],
"source_labels": {
"https://app.asana.com/0/0/1207268731049844": "Роль Advisor и стейкхолдеры",
"https://app.asana.com/0/0/1207685708094386/f/": "Шаблон нити DRI ↔ PA"
}
}, },
{ {
"id": "project-update", "id": "project-update",
@@ -15,3 +15,294 @@ source: "https://us01mynotes.zoom.us/NtgNJ-Y-Qxy5sMGMjngukw"
- **15:12:26 · Alex Martemyanov:** Yeah, no problem. - **15:12:26 · Alex Martemyanov:** Yeah, no problem.
- **15:12:29 · Dominik Kapusta:** Райт. Хауинг? - **15:12:29 · Dominik Kapusta:** Райт. Хауинг?
- **15:12:34 · Alex Martemyanov:** I think I'm doing pretty good. - **15:12:34 · Alex Martemyanov:** I think I'm doing pretty good.
- **15:12:37 · Dominik Kapusta:** Майс?
- **15:12:41 · Alex Martemyanov:** And I'm just doing a lot of stuff. Uh, recently… So, uh… Mainly, mainly, mainly what data priority is saying, the CPM. And, uh, release maintenance this week. to Liberace. It's gone. But I'm trying to go through the crashes that we have at least.
- **15:13:10 · Alex Martemyanov:** Identified several web kits, probably related to macOS 27 update.
- **15:13:15 · Dominik Kapusta:** Угу?
- **15:13:16 · Alex Martemyanov:** And, uh, trying to investigate deeper. For the top priority, the CPM crash, we started receiving the data, the attribution data and fixes results and well, surprise, surprise. Nothing helps. And the attribution data shows like regular normal initialization flow failure. Uh, and uh… nothing. So, yeah, we are…
- **15:13:48 · Dominik Kapusta:** Даже? Салют, Фикс Ди Донт Хэлп?
- **15:13:53 · Alex Martemyanov:** The fact that this didn't help, uh, the fix… I mean, the… the… Uh, the fixed, uh… The one that, uh, was released, that put to the release, it's not active yet. Uh, the… background process crash one. But… Probably… probably it's not the case, actually, because the pixel data does not identify the ground terminations. So… Well, it is the kind of… Lab reproduced issue. But it's probably not the case that users are experiencing and the cases. Actually, it is mitigated by WebKit itself after we release its navigation cycle, so it was only a prelude standard. special conditions that we are not… they are not proof, basically, so probably I'm gonna revert the fix.
- **15:14:58 · Dominik Kapusta:** Окей?
- **15:15:01 · Alex Martemyanov:** And, uh… Yeah, there is some… Mmm… So, one fix, uh, was about the tab management, and, uh… It didn't show up as a…
- **15:15:18 · Dominik Kapusta:** Угу?
- **15:15:22 · Alex Martemyanov:** Something actually affecting a lot of users, and those cases were probably some single ones, if they were… Uh, at all. I mean, it's not the… Thousands that we are receiving. And, uh… At least we've, uh, leaned… Cleaned up this piece. And, uh… know the conditions. Uh, there is one hypothesis that I'm working on about the WebKit initialization race condition that was fixed. Uh, in macOS 27, but we are still receiving the reports from macOS 27, but there is… there is not a strong…
- **15:16:11 · Dominik Kapusta:** Ооо?
- **15:16:15 · Alex Martemyanov:** Uh, proof. that the fix… It was actually released in macOS 27.0. Uh, so, I… Started. I found out the… Well, actually, I've… Mainly with the help of codecs, but we managed to find the relation between macOS versions and. matching webkit releases. So I'm gonna publish it to Apple Documentation and talk about it in the Apple Weekly today.
- **15:16:52 · Dominik Kapusta:** Угу? Угу?
- **15:16:56 · Alex Martemyanov:** Hopefully it could help somebody else in the investigation, in getting fit issues. Uh, the relationship between macOS versions and WebKit releases, so we could… so we could know exactly which bug was fixed where, because there is a fix in our web extensions system, and for sure it was there, uh…
- **15:17:08 · Dominik Kapusta:** Оу. Окей? Оу, найс?
- **15:17:19 · Alex Martemyanov:** before Mac OS 27. But, I still didn't finish this research and, have no, you know, proof that the fix. that the issue is affecting us, but it could. And it looks like a very good candidate, uh, to fix it, and, uh, if it's… If, uh, if it's proved, uh, it means the actual issue is found, and, uh… It's gonna be fixed.
- **15:17:48 · Dominik Kapusta:** Найс. Гай.
- **15:17:49 · Alex Martemyanov:** But if not, I will… I'll keep the research, probably send an update to the project, maybe a bit more time. Thank you. And keep trying to. To find the issue… maybe we'll have another iteration of some… Excel data attribution. Uh, especially having the most cases in this, like, normal workflow. So basically, we're not receiving the, uh, the response to the init event, and everything seems fine. No crashes, no… leakages, no…
- **15:18:28 · Dominik Kapusta:** Та?
- **15:18:29 · Alex Martemyanov:** Reloads, just regular launch, regular load, and… Silence. Those are, in most cases. Uh, but at least, uh, like, no data is still a data. Yeah, at least other cases are mostly filtered out. So, uh, another iteration is going on. Uh, yeah, I am investigating the crashes. that we have in Sentry. In scope of… Reduce maintenance. Um…
- **15:19:04 · Dominik Kapusta:** Гм?
- **15:19:06 · Alex Martemyanov:** So luckily everything else. Seems… and a good… And… And, uh, yeah. made some more efforts to. keep the VS Code migration project going. Um, so yeah, you've seen that, and, uh, I actually opened the draft PR, uh, that already works in VS Code, and, uh, yeah, I'm… I actually work in this branch and these changes and it just works. You just open the project file, and it loads, and indexes, and builds, and debugging in there, and yeah, it's… it's working fast, and, uh, there is version manager and everything else, so compared to Xcode, it's, like, really… step forward.
- **15:19:53 · Dominik Kapusta:** Найс?
- **15:19:58 · Alex Martemyanov:** Especially having all these bugs in Xcode 27. I tried using it, it's unusable, basically. compared to Xcode 26. Its window blinks, it hangs.
- **15:20:10 · Dominik Kapusta:** Интенько?
- **15:20:10 · Alex Martemyanov:** If you try to render some diffs for… A bit larger files, it just hangs, as usual.
- **15:20:16 · Dominik Kapusta:** Угу?
- **15:20:17 · Alex Martemyanov:** I didn't crash yet, but anyways, I don't find it really usable product.
- **15:20:21 · Dominik Kapusta:** Интростинг?
- **15:20:24 · Alex Martemyanov:** So hopefully it could. It could be a good product to show the commitment. And yeah, I made some measurements. Of course, I, uh… I understand it includes all these extra build steps like VPN and, DBP targets. And, uh, several extra steps that XcodeBuild includes and VS Code does not. But… Compared to the… Regular testing builds that we are… it's like… Those who don't use DBP or VPN are using it, like, twice speed up.
- **15:21:07 · Dominik Kapusta:** Yeah, yeah, yeah. Um… Dots. Cool. Overall, that's cool. I'm also wondering… If we could use it. For AI agents in the cloud, for instance, like to speed them up. Uh, when it comes to… when it comes to, like, debugging issues, you know, like, I don't know, fixing crashes, things like that. We're… we don't do it at all right now, but just…
- **15:21:31 · Alex Martemyanov:** And…
- **15:21:41 · Dominik Kapusta:** just thinking about possible future users of this. Of course, Xcode build is the default build system, and so on and so on, but… If we can speed up a build. drastically by using Swift Package Manager essentially, right?
- **15:22:06 · Alex Martemyanov:** Well, it's the next step. I already started another work 3D from this, from VS Code main migration, so there is some corners.
- **15:22:14 · Dominik Kapusta:** Mm-hmm.
- **15:22:17 · Alex Martemyanov:** present. First of all, there is the files cleanup. Files by target cleanup is not complete yet. I just made it work by moving the files that are not included in the main browser target, but there is still some work to move. To the right places, to… make the shared packages used by several targets. It's like a… And… ideal picture that we'd like to see in the future, I think. But it's even blocking this stage. And the next stage is the tests. So first of all, the tests are not migrated at this step. It's only the main app. And, uh… but the next step is already made, uh, about the unit tests. So basically, those should not be tied to the app. Because we are supposed to be unit testing and not integration testing, so the app delegate was cut, the, uh, dependencies were moved away, and, uh, all these tab loading, uh.
- **15:23:11 · Dominik Kapusta:** Yeah, of course, of course.
- **15:23:21 · Alex Martemyanov:** There were many dependencies. added to the tab since the last screen up. It was made to unit tests, and, uh, they added some complexities. But, uh, so currently, at my next work three, uh.
- **15:23:28 · Dominik Kapusta:** Mm-hmm.
- **15:23:36 · Alex Martemyanov:** Unit tests are working under Swift build and Swift test. Integration tests are not touched yet, and they are kind of broken from if you run from VS Code, but they are still run from the Xcode target as before. Uh, so that's the state, so it's just, uh, the… Beginning of our… Clean up project, I'd say, and. I think it's worth doing it and. Just to keep, um… To keep the technical debt cleaned up. So yeah, I'm not sure yet about the integration tests, but uh… If… If it works out… Again, in the perfect world, there should be a delegate mock. in the test that's required. And there should be a window mocks in the test that are testing integration inside a window. And supposedly nothing should stop us from using Swift testing in the integration test as well.
- **15:24:40 · Dominik Kapusta:** Mm-hmm.
- **15:24:49 · Alex Martemyanov:** But of course, it would be much more work to handle all of that.
- **15:24:56 · Dominik Kapusta:** Mm-hmm.
- **15:24:57 · Alex Martemyanov:** But, uh, yeah, uh, speaking of, uh, the general speed up. I'm not sure if it would be the real case here, unless we want to… move further and, migrate the unit tests to Swift, to Swift tests. Essentially making them parallelized. Uh, if they don't depend on the main thread, because most of them do at the moment.
- **15:25:24 · Dominik Kapusta:** Mm-hmm.
- **15:25:32 · Alex Martemyanov:** But if, uh, if you just build the whole build with the DPP and VPN targets and so on, it's just… Essentially the same amount of work. Maybe there is a speed up, like… 2 to 5 seconds, but it's not, uh…
- **15:25:48 · Dominik Kapusta:** Mm-hmm.
- **15:25:49 · Alex Martemyanov:** It's not very acceptable.
- **15:25:51 · Dominik Kapusta:** You know. um… All right, that sounds… that all sounds good. There's one thing that I wanted to bring your attention to, which is, uh… how you're running this work, uh, these days, and what happened on the Apple Weekly recently, and that proposal that you had, uh, that was… mentioned in the Apple Weekly and then quickly closed. Uh… Let me bring it back. Code initialized, code data model, those… Yeah, okay, it's essentially a no-op, because Xcode 27, uh… handle that. Uh…
- **15:26:52 · Alex Martemyanov:** Yeah, so generally, I think it was a pretty good discussion happening there, yeah.
- **15:26:57 · Dominik Kapusta:** Mm-hmm.
- **15:27:01 · Alex Martemyanov:** So we… had a small talk with Sam and Bartek, basically, in the tech design, and so general agreement was that, yeah, it's worth doing, maybe, at some point, but we have a really good validation of. The migration is working and nothing breaks. So that would include testing on all the supported operating systems and migration prompt. I have some of the major browser releases, but… Again, I think it was doable, and maybe it's actually worth doing at some point, and maybe there is still some migration issues that we may experience from the older browser versions that we. Okay. As I remember, there's, uh, not… not much related to actual database migration, and it was mostly relying on the Core Data native migration system that.
- **15:28:10 · Dominik Kapusta:** Hmm, yeah.
- **15:28:11 · Alex Martemyanov:** It was only recently updated to some smart… migration techniques. In the recent, uh… I'll go to service assigning. So, uh, I won't tell it's, like, 100% closed, but yeah.
- **15:28:28 · Dominik Kapusta:** Mm-hmm.
- **15:28:30 · Alex Martemyanov:** using good judgment that it just works, and, uh, Expo 27 supports it, uh, building without problems, uh, just requiring a minor cleanup of the fact.
- **15:28:33 · Dominik Kapusta:** Hmm. Mm-hmm.
- **15:28:41 · Alex Martemyanov:** used to compile it in runtime, towards just letting it live as it is, along with the files that are not currently moved to shared packages and so on. So just to…
- **15:28:52 · Dominik Kapusta:** Mm-hmm.
- **15:28:54 · Alex Martemyanov:** So just to reduce the amount of attention of other team members and my attention as well.
- **15:29:01 · Dominik Kapusta:** Mm-hmm. Yeah.
- **15:29:02 · Alex Martemyanov:** in order to keep getting things done and not getting things perfect.
- **15:29:12 · Dominik Kapusta:** Fine. One thing that I wanted to raise here is… You're doing this… Some people… could say for yourself, essentially. Because, uh, we aren't really… that much upset with Xcode, and people are just using Xcode and not trying to find, uh… alternatives. Uh, probably not as much. As they would… Be interested in it. Before Xcode MCP? Uh, but now, when… when agents can use Xcode MCP, it's probably not that much of a deal-breaker, uh, to have to continue using, uh, Xcode. And, uh… It seems to me, which can be seen also in some comments from the team, that. that people are not on board with… not… I mean, they are not as excited about it as you are, and they are asking questions, and one of them was from Juan on your project. Uh, ask last week, where he outlined a few… Doubts? And…
- **15:30:33 · Alex Martemyanov:** Yeah, and basically, I answered that question recently. Provide… updating the description, and uh… providing the benefits, and uh… metrics of, the build times.
- **15:30:49 · Dominik Kapusta:** Okay, that's good, but… Well, are you aware that… No one gets a notification about that. So, if Juan asked a question and you put some… Additional description in the task.
- **15:31:14 · Alex Martemyanov:** No, I mean, I did done sorts, didn't I?
- **15:31:20 · Dominik Kapusta:** I, I can't see your comment.
- **15:31:23 · Alex Martemyanov:** Mmm…
- **15:31:24 · Dominik Kapusta:** Well, you mentioned updated the description, adding the benefits of the project, that's fine, but it's…
- **15:31:24 · Alex Martemyanov:** Here's the findings.
- **15:31:32 · Dominik Kapusta:** Uh… how to say that? It could have been done more, uh… I mean, it's not clear that you're addressing Juan's doubt here. You could… you could have, as well, just ping Juan and say… and write, like. here you are, here's how things look like. You… you could steal, like, other command below your command, and uh… And ping one and mention that the benefits are now in the description. You can say I mentioned that in my previous comment, but maybe that wasn't too clear, so please have a look. Just to be sure that…
- **15:32:15 · Alex Martemyanov:** Mm-hmm. Oh, yeah.
- **15:32:20 · Dominik Kapusta:** That they got the message, because you're doing a good job, but it may be misunderstood or not understood by some people on the team.
- **15:32:33 · Alex Martemyanov:** Yeah, probably let me just use this Asana answer feature. Under the comments, and…
- **15:32:40 · Dominik Kapusta:** Yeah, that makes sense.
- **15:32:41 · Alex Martemyanov:** Mention the… the update.
- **15:32:44 · Dominik Kapusta:** Yeah, okay. Um… no. But also, it is something that… The team is overall uncertain about when it comes to… Complexity of your changes? And we may be in a sort of a bubble, the two of us, because I've reviewed a lot of your code, and I'm used to big changes, and I'm just reviewing them, and I understand what is changing most of the time. And if I see a, like, that pull request recently, which had 8,000 lines of code added, 17,000 removed, I didn't freak out, because I saw that the localizable strings. file, so that's all good, and you've verified with me that it's just… deleting some strings that were never used, so that's… that's good. But… the rest of the team… Maybe just not… maybe, like, more sensitive overall, and, um… And it's important to… To keep them calm, essentially, to…
- **15:33:55 · Alex Martemyanov:** And maybe should I, uh, have another… another talk during today's Apple Weekly? And, uh, go through the steps that were already completed and, uh, the ones that… remaining… Uh, kind of justifying the, uh, the path forward, and for those who are kind of hesitant or don't want to switch to… give an explanation of what is changing and the changes, not not as big as it.
- **15:34:30 · Dominik Kapusta:** Yeah.
- **15:34:30 · Alex Martemyanov:** May… as it may sound.
- **15:34:32 · Dominik Kapusta:** Yeah. I think this… I think this is a good idea. Essentially… Making your work… more… approachable… by the rest of the team. So. If you have the time. to prepare some short talk about this on… for today's Apple Weekly, that's great. Um, but… But… Ideally. Ideally, make it so that… It shows a clear improvement. For those, also for those that… aren't looking to switch from Xcode to VS Code. Is there? Any improvement? Actually, I think there is, right?
- **15:35:31 · Alex Martemyanov:** Yes, so, yeah, basically edited to the… to the project benefits and section. So, basically, there is a fixed list of the tasks completed.
- **15:35:36 · Dominik Kapusta:** Yeah, in that table, right? Mmhm.
- **15:35:42 · Alex Martemyanov:** Like the code separation and clean up overall.
- **15:35:46 · Dominik Kapusta:** Mm-hmm. No.
- **15:35:49 · Alex Martemyanov:** Moving to… Six packages, unifying the compile flags, dropping storyboards, and…
- **15:35:55 · Dominik Kapusta:** Mmhm.
- **15:36:00 · Alex Martemyanov:** Moving to the UI test instead of the sandbox test tool. Dropping dead files and unifying asset access and core database models access after the discussion.
- **15:36:07 · Dominik Kapusta:** I don't know.
- **15:36:16 · Alex Martemyanov:** Um, they're localizable, clean up, and, uh, so basically nothing changed in that, but, uh, they are now. targeted to a bundle, which is, like, again, the unifying of the codebase, and maybe give a short glimpse about the tests and the further cleanup that remains.
- **15:36:24 · Dominik Kapusta:** Yeah. Mmhm.
- **15:36:35 · Alex Martemyanov:** after the project is dropped in. Again, there is, uh… Um, so we… Essentially, we eliminate the, uh, project file merge conflicts, because, uh, if we use SPM targets, they are just… Placed into the folder and match the target.
- **15:37:02 · Dominik Kapusta:** Mm-hmm.
- **15:37:03 · Alex Martemyanov:** What else? The dependency structure. Like, for some routing for… for storyboards and related crash reports when the app is removed. And… So, basically, those are little… Steps. that are improving this sequentially. And, uh… Probably should not be highlighted as a, like, hundred percent CS code migration required for everybody in the in this room.
- **15:37:35 · Dominik Kapusta:** Mmhm.
- **15:37:35 · Alex Martemyanov:** But general improvements that also bring the VS Code compatibility and as basically adding another build target and another workspace file that should not be. Uh, wouldn't… won't be required to manage manually, because it just works.
- **15:37:58 · Dominik Kapusta:** Yeah, um… Yeah, yeah, that would be… that would be good, um… There's one… more thing that… Many people on the team would be interested. in I believe and that is the iOS app.
- **15:38:25 · Alex Martemyanov:** Somebody needs to volunteer, I guess. Well, I was, uh, thinking… I was considering to…
- **15:38:27 · Dominik Kapusta:** Yeah.
- **15:38:33 · Alex Martemyanov:** to take the iOS app after this project is complete, finished, and done, and closed. Uh, so I have, like, some time slots to… to make some iOS, uh… Impact. And the first thing to do about it is the navigation framework migration, of course, which was discussed, and I think it's being discussed, but nobody really. And he's approaching to it. Another thought I did have is the… the further Swift packages separation, because currently we have around 1,700 code files in the main target, and it takes really long to index and to build. And… Actually parallelizing it and separating the responsibilities is… May sound like a good idea. Of course, not as a single step, and not at this point in time. But maybe some some things could be moved out and unified again between iOS and macOS code base.
- **15:39:50 · Dominik Kapusta:** Mm-hmm. Mm.
- **15:39:50 · Alex Martemyanov:** For example, UI components, like we have a lot of shared components. A lot of third. Mmm… Components in the design system, and we have a lot of. macOS unique components that can be moved to some package, for example, the buttons. address field as a separate package that could be like tested and so on and so on. Uh… no. So. I don't have a roadmap for it or a concrete plan.
- **15:40:27 · Dominik Kapusta:** No.
- **15:40:27 · Alex Martemyanov:** I don't know, but… Probably it would be a good AR maintenance project. As well. Just worth, uh…
- **15:40:36 · Dominik Kapusta:** Mmhm.
- **15:40:39 · Alex Martemyanov:** reaching at some point.
- **15:40:41 · Dominik Kapusta:** Mm-hmm.
- **15:40:43 · Alex Martemyanov:** Certainly, certainly not, not today.
- **15:40:46 · Dominik Kapusta:** Yeah, yeah, um.
- **15:40:52 · Alex Martemyanov:** Yeah, maybe it's worth mentioning further. Improvement roadmap that just to give. people some thoughts about to to think about the perspectives. So… I'm not really into the iOS code base and can't say I understand how it works in a whole.
- **15:41:14 · Dominik Kapusta:** Hmm. Sure. I agree. I mean, I understand that.
- **15:41:20 · Alex Martemyanov:** So…
- **15:41:24 · Dominik Kapusta:** I understand that for sure.
- **15:41:26 · Alex Martemyanov:** So, probably it's more of a iOS-dealized responsibility, I'd say, but yeah, I'm kinda open to help or to guide, to make some parity if there is anybody to…
- **15:41:31 · Dominik Kapusta:** Mm-hmm. Mmhm.
- **15:41:40 · Alex Martemyanov:** To work on it.
- **15:41:42 · Dominik Kapusta:** Yeah. Yeah, yeah. Uh, yeah, sure. I'm, I'm, you know, I'm trying to…
- **15:41:43 · Alex Martemyanov:** But not something on my plate yet.
- **15:41:49 · Dominik Kapusta:** Trying to make it easier to find… to ensure that the team is okay with this work, because as some of the people mentioned. Didn't. The consequences of mistakes during refactoring can be… can be very severe. Uh… I understand that we've dropped, uh… we've dropped the idea of… of initializing core data models in… in the code, which was probably the biggest, uh… the thing that started everything, really, the entire discussion, because otherwise… otherwise you're just refactoring the code. Uh, but that was… A big thing. So… So, where were we? Um… Just let me just understand it better. Quickly share this.
- **15:43:02 · Alex Martemyanov:** Okay.
- **15:43:04 · Dominik Kapusta:** So, in this table, we have, like, clean build, for instance. Xcode Main is what we have currently in Main, right?
- **15:43:13 · Alex Martemyanov:** Yep.
- **15:43:14 · Dominik Kapusta:** And Xcode SPM, what is that exactly?
- **15:43:17 · Alex Martemyanov:** It's basically the migration that's happening in this branch. So we're moving the main app code to an SPM package.
- **15:43:24 · Dominik Kapusta:** Mm-hmm. Mmhm.
- **15:43:31 · Alex Martemyanov:** And, uh, the three columns. after the main branch is basically V. This is basically the new approach where we use the main target as an SPM package.
- **15:43:41 · Dominik Kapusta:** Mm-hmm. Mm-hmm, mm-hmm, m.
- **15:43:47 · Alex Martemyanov:** So yeah, I highlighted it in the comment before that…
- **15:43:47 · Dominik Kapusta:** Oh.
- **15:43:54 · Alex Martemyanov:** Well, the main targets in SPM. Technology migrated to USPM.
- **15:44:08 · Dominik Kapusta:** So I think, um… I think we should be realistically looking at this column. Maybe also this column, but it would be great to have Xcode build. Starts for main as well. Maybe there are… Closer to this. As well. Could you… could you get them for, like, Xcode build for main? What I'm getting at is… VS Code, realistically, I don't believe people on the Apple team will switch to VS Code. So they will stay with Xcode. But Xcode build is something that agents use. So if you make agents 20% faster to build the app. That's a significant improvement, and that's something that people should… Uh… understand, and should be aware of, that your work is making. Xcode build, 20% faster. This is also CI. I understand. And that… In my opinion, that should be the main selling point of your work.
- **15:45:15 · Alex Martemyanov:** Mmhm. It's maybe not 100% accurate compared to the main branch. Because I suppose it would be the same for the main branch Xcode.
- **15:45:30 · Dominik Kapusta:** I see, okay.
- **15:45:31 · Alex Martemyanov:** So it's just comparing to the to the main branch. And, yeah, I probably should run just a comparison for the…
- **15:45:38 · Dominik Kapusta:** Yeah. Okay.
- **15:45:38 · Alex Martemyanov:** Uh, main branch, and it's called… Mm-hmm. SPM… Xcode build on the main branch, yeah.
- **15:45:49 · Dominik Kapusta:** Mm-hmm. Um… All right, because then we essentially rebuild no changes, rebuild code change in main module. This is for some reason slower.
- **15:46:03 · Alex Martemyanov:** Yeah. There is, so generally, export build, performs, this package resolution thing that is main slowdown during the everyday development, and that's why I.
- **15:46:12 · Dominik Kapusta:** Mmhm. I see.
- **15:46:17 · Alex Martemyanov:** did not really consider, uh, using it in VS Code, like a build system, because it's… essentially, it slows down compared to the main Xcode IDE, and does not give any… any speedups.
- **15:46:21 · Dominik Kapusta:** Hmm. Mmhm.
- **15:46:33 · Alex Martemyanov:** But, yeah, here we can see that this extra steps is giving extra ten. maybe six to ten seconds compared to the main Xcode that does not perform this package resolution every time you rebuild it.
- **15:46:48 · Dominik Kapusta:** I see. Okay.
- **15:46:49 · Alex Martemyanov:** And the VS code speed up is mainly because of the target difference that does not include this VPN and GBP. Yeah, so it's not some code change that causes it. It's not magic.
- **15:46:58 · Dominik Kapusta:** Mm-hmm. Oh, yeah. Okay. Okay. I see. Right.
- **15:47:04 · Alex Martemyanov:** I'm just making another target for local builds of the main browser UI, and nothing else.
- **15:47:14 · Dominik Kapusta:** All right. All right. Right, um… Okay. In this case… I'm… I'm unsure if… if this… It's worth, uh. Pursuing, unless the people on the team… would be interested in trying out VS Code. yeah, I… I realize it's harsh, but just wanted to tell you how I feel about it, and I… and how I believe. people will feel about it. I'm.
- **15:48:02 · Alex Martemyanov:** You mean the SPM migration of the main target?
- **15:48:05 · Dominik Kapusta:** No. If it's… Because a lot of work that you've done for this is great to have overall, like all the cleanups, separating files by target, I like this very much, and it's great to have someone. Especially someone like you work on this and do this. Right. But then… There may be some, um… There is… there is this, like, infliction point at some… somewhere in this project where… The changes are not as… As easy ingestible, easily ingestible by everyone, and where they would be essentially… Questioning. Why are we doing them? So… Uh… so, um…
- **15:49:11 · Alex Martemyanov:** So, generally, the changes to… Move the files. Hmm, from… project file to… project to the ESPM folder, and the events mostly staying in the same place, and uh… accessible from there. So, just continuing the work of, uh… moving the files to SPM targets that were started, so… We're a while already. And basically, the VS Code step, the… I mean, the VS Code compatibility step itself, it's, uh… Not kinda. Related to it, it's just adding a workspace file and an extra target to the Swift package definition.
- **15:50:02 · Dominik Kapusta:** Yeah, yeah, I see. Um.
- **15:50:02 · Alex Martemyanov:** Right. So, I'd say, uh… the… adding the project. The VS Code workspace file itself should not meet any… Issues with the team or whoever, just… if you don't want, don't use it.
- **15:50:24 · Dominik Kapusta:** Hmm, yeah.
- **15:50:24 · Alex Martemyanov:** Uh, but the SPM package migration, it may, uh… It may be a discussion point here.
- **15:50:32 · Dominik Kapusta:** Yeah, so, uh… What would you say to… Just open a tech design about it and lay out clearly. The benefits, the risks. the, um, build time, influence it has on build time, uh. Of course, there are clear benefits. There are clear, very clear benefits, like AppDelegate is not involved in unit tests. That's… the main, most important thing for me, I believe. Yeah, um… But… It feels to me like this needs to be clearly laid out and discussed within a team. And then you can just do it. Because otherwise, you're doing something that has not been truly agreed on, and it's… A significant change to the app.
- **15:51:31 · Alex Martemyanov:** Boom. Well, I remember it was at some point, but probably…
- **15:51:38 · Dominik Kapusta:** It also… yeah, it also rings a bell. Like, I think that at some point there was a tech design on that, but I…
- **15:51:39 · Alex Martemyanov:** It was, um… There was a tech design and it was generally discussed in the meeting.
- **15:51:46 · Dominik Kapusta:** Yeah.
- **15:51:50 · Alex Martemyanov:** the point.
- **15:51:51 · Dominik Kapusta:** Yeah.
- **15:51:52 · Alex Martemyanov:** That's, uh… There was no really real complaints about this migration to perform. Let me find it where it was.
- **15:52:05 · Dominik Kapusta:** Convert all macOS modules to Swift package.
- **15:52:08 · Alex Martemyanov:** Yes, I think it was.
- **15:52:10 · Dominik Kapusta:** Mmhm.
- **15:52:11 · Alex Martemyanov:** Is it…
- **15:52:12 · Dominik Kapusta:** I was a, I was a reviewer there.
- **15:52:15 · Alex Martemyanov:** Yeah. And…
- **15:52:24 · Dominik Kapusta:** But… Yeah, okay.
- **15:52:28 · Alex Martemyanov:** Doing design, yeah, convert. Yeah, so generally, it includes pretty much everyone, and uh… Recommended approach was the. Yeah, it's basically…
- **15:52:40 · Dominik Kapusta:** Mmhm.
- **15:52:41 · Alex Martemyanov:** That one, October 24th, a year ago.
- **15:52:43 · Dominik Kapusta:** Mm-hmm. A year ago… Yeah, a year ago. Em.
- **15:52:54 · Alex Martemyanov:** So… Do you think I should reopen it for… Another round of discussion.
- **15:53:04 · Dominik Kapusta:** I think it would make sense. I'm just having another look. There is a proposal linked there and proposal is called convert all macOS modules to Swift package.
- **15:53:26 · Alex Martemyanov:** Yep.
- **15:53:27 · Dominik Kapusta:** And there were some… some concerns will have essentially different linking. Yeah. I think it would be… Best to resurface this discussion, given that in the original tech design. Oh, okay, so you open the proposal, and then because of the comments on the proposal, you open the tech design, I think, because you're… In the tech design, in the comments, you're linking some comments from the, um…
- **15:54:05 · Alex Martemyanov:** Yeah.
- **15:54:07 · Dominik Kapusta:** mentioning comments from the proposal. Is this 100% accurate? Is this tech design 100% accurate with the work you're doing right now?
- **15:54:26 · Alex Martemyanov:** And… It's, uh… Mostly, in main points. So there is, uh, this, uh…
- **15:54:32 · Dominik Kapusta:** Hmm. Okay.
- **15:54:41 · Alex Martemyanov:** VS Code launcher implementation shipped into the repository step.
- **15:54:41 · Dominik Kapusta:** Yeah, okay.
- **15:54:48 · Alex Martemyanov:** And, uh, using the unsigned loader of, uh, dynamic libraries. So, essentially, it's the same, but, uh… Instead of it, the build script just creates a new PBX project file. a compact one, writes a single main function in C++, and, uh, signs it, and, uh, loads the dynamic library, so…
- **15:55:13 · Dominik Kapusta:** Mm-hmm.
- **15:55:13 · Alex Martemyanov:** Most probably, it would even work for the App Store builds. And in the tech design, it's mentioned that it won't work with App Store builds. Oh. Otherwise, the idea is mostly the same. It's just some implementation detail.
- **15:55:33 · Dominik Kapusta:** It doesn't support sandboxed builds, is that right?
- **15:55:36 · Alex Martemyanov:** Uh, it should, I guess. Because, uh, Sandbox builds… do support dynamic linking and since there is this.
- **15:55:46 · Dominik Kapusta:** You know.
- **15:55:47 · Alex Martemyanov:** Oh. I think they should. The only issue is the signature that should match. For both the app and the library. And since we signed both.
- **15:55:56 · Dominik Kapusta:** But then again. Yeah, the dynamic linking is still just… to make it compatible with VS Code, is that right? Like, Xcode wouldn't use it. Yeah.
- **15:56:09 · Alex Martemyanov:** Yes, so it's an extra target, only for the VS Code user, so it's not the main app target.
- **15:56:13 · Dominik Kapusta:** Mmhm.
- **15:56:16 · Alex Martemyanov:** The main app is compiled to this, static library as, it's mentioned in this schema here, macOS as a static library.
- **15:56:23 · Dominik Kapusta:** Right, yes, yes, yes, yes.
- **15:56:25 · Alex Martemyanov:** And, it's, either the Xcode project that uses it or the dynamic library that's linked.
- **15:56:33 · Dominik Kapusta:** Mm-hmm. Mm.
- **15:56:33 · Alex Martemyanov:** Uh, the king suit. And the dynamic library is just kind of an empty placeholder that loads all the content on the static one.
- **15:56:51 · Dominik Kapusta:** No. All right, well, so the tech design, as far as… converting the macOS app to a Swift package or. Yeah, it's… well, what I wanted to say is that as far as… Removing up delegate dependency from unit tests, which is the clearest benefit here. It's looking good. I… might… I might consider just… just… Opening… this one more time and asking for input if anyone has anything because we really didn't have much discussion on that one and it's been a year. Um… So. Just so it doesn't… I actually have to postpone my next one a little bit. Please give me.
- **15:58:07 · Alex Martemyanov:** Well, uh, I may…
- **15:58:09 · Dominik Kapusta:** Give me a moment.
- **15:58:11 · Alex Martemyanov:** I may redesign the tech design in, uh… In a way, to only include the… SPM migration.
- **15:58:21 · Dominik Kapusta:** Mm-hmm.
- **15:58:22 · Alex Martemyanov:** Without, uh, direct… VS Code. without directly mentioning the VS Code thing and the dynamic library that was essentially discussed already in that tech design and kind of approved. But yeah, just to keep it in the. Insight. Of, uh, everybody, um…
- **15:58:46 · Dominik Kapusta:** Mm-hmm. Mm-hmm.
- **15:58:48 · Alex Martemyanov:** So this step of migration may not go into the package and providing the benefits like tests and code separation. Oh, yeah. and maybe the build times that I'll. measure here.
- **15:59:11 · Dominik Kapusta:** No.
- **15:59:13 · Alex Martemyanov:** And…
- **15:59:14 · Dominik Kapusta:** I think that would be great, I think that would, uh… Help keep people calm about this. Um… Yeah, because even… even if…
- **15:59:27 · Alex Martemyanov:** And probably…
- **15:59:28 · Dominik Kapusta:** Yeah. Go on, go on.
- **15:59:29 · Alex Martemyanov:** Probably it's worth making another roadmap that we discussed today before about the further separation and maybe unifying with iOS as well.
- **15:59:43 · Dominik Kapusta:** It would be… it would be amazing. Yeah.
- **15:59:43 · Alex Martemyanov:** I mean, moving. splitting the target into… into smaller… into smaller pieces.
- **15:59:49 · Dominik Kapusta:** Yeah. I think it will be great.
- **15:59:51 · Alex Martemyanov:** Because, uh… Because currently the indexation and build of 1,700 files is… What takes most, I believe.
- **16:00:05 · Dominik Kapusta:** You know, there's also Apple modularization roadmap. which this… could potentially… fit into.
- **16:00:17 · Alex Martemyanov:** Oh.
- **16:00:17 · Dominik Kapusta:** And that roadmap is… Sounds like something in progress. Since… last year. And it's essentially about breaking down dependencies, but also… Well, it's called Apple Modularization Roadmap, so whatever… Whatever is related to… to Apple modularization.
- **16:00:42 · Alex Martemyanov:** Basically… Basically, yes, it is essentially work in scope of regularization.
- **16:00:44 · Dominik Kapusta:** Could fit there.
- **16:00:50 · Alex Martemyanov:** And sharing code, and… Moving. Uh, moving forward, so… And the roadmap in the VS Code migration project does also have this separation of files to the shared packages.
- **16:01:06 · Dominik Kapusta:** Mm-hmm.
- **16:01:12 · Alex Martemyanov:** between DVP and VPN targets that are remaining shared code in the export project.
- **16:01:20 · Dominik Kapusta:** Yeah, mm-hmm. So, I guess, for now… I have a pull request on me that is looking fine. I'll have to… I just have to build the app and test the VPN.
- **16:01:38 · Alex Martemyanov:** So it's… yeah, the pull request is just… it's just moving files away from the folder. It does not change the… the targets.
- **16:01:43 · Dominik Kapusta:** Mm-hmm, yeah, yeah, yeah, yeah, sure.
- **16:01:47 · Alex Martemyanov:** But the rest, it's not done yet, and the rest is more heavy refactoring, and it's not planned in current iteration.
- **16:01:47 · Dominik Kapusta:** The… Yeah, yeah, and I… And I suggest, if that… Makes sense to pause here, and uh… just present the idea to the team, somehow. If you want… I mean, it's completely fine if you want to draft something and have me review it before you share it with the team. You don't have to do it today, also. But… It also… it also makes sense… To just communicate that. In this. Migration project, by saying… I will prepare a tech design and share it with you soon. That would also make people feel heard and feel like we're not doing something behind the scenes. That could potentially harm…
- **16:02:58 · Alex Martemyanov:** Okay, so…
- **16:02:58 · Dominik Kapusta:** The product.
- **16:03:00 · Alex Martemyanov:** So, do you think it's worth mentioning today's Apple Weekly, just to highlight what… what was done, what is actually… on the plate as of today, and how it's related to the existing modularization effort, and what are the. further plans, and kind of wrap it up with the tech design that I'm going on.
- **16:03:23 · Dominik Kapusta:** I think so. I think that would be great. And that tech design, worst case, it will be open for a number of days, but we'll figure out…
- **16:03:23 · Alex Martemyanov:** Prepare for the for the migration.
- **16:03:35 · Dominik Kapusta:** the right way forward that everyone is happy with.
- **16:03:42 · Alex Martemyanov:** But it probably also can… or should include the further roadmap and the further code separation maybe.
- **16:03:52 · Dominik Kapusta:** Yeah, why not? Why not? Uh, well… As long as the tech design is not getting too big, because if it's too many problems in a single tech design, then it should be multiple tech designs, probably.
- **16:04:08 · Alex Martemyanov:** I mean, it can be like a framework for the roadmap. So first stage is to main article to an SPM package. The second step is to move like UI.
- **16:04:08 · Dominik Kapusta:** But. Mmhm. Mm-hmm. Mm-hmm.
- **16:04:19 · Alex Martemyanov:** The third step is to clean up the DBP and VPN to the shared code.
- **16:04:24 · Dominik Kapusta:** Mm-hmm.
- **16:04:26 · Alex Martemyanov:** Packages and… Just mentioned this. These packages that would be moved. on each step.
- **16:04:31 · Dominik Kapusta:** Yeah, and…
- **16:04:32 · Alex Martemyanov:** So to make it a part of a bigger project in the mobilization effort.
- **16:04:38 · Dominik Kapusta:** And try thinking about iOS, and I understand that iOS can't be done because of blockers, and some of the blockers are probably just… they use storyboards, still, so if they would like to… Follow this… this roadmap, they should… Essentially fix their stuff. Um… If they want to.
- **16:05:00 · Alex Martemyanov:** Well, storyboards are not a big deal really, because I did all this remaining storyboard migration in a single day, just putting it to cloud and asking, okay, do it correctly and don't make mistakes.
- **16:05:06 · Dominik Kapusta:** Thank you. Bye.
- **16:05:15 · Alex Martemyanov:** There were a couple of mistakes, although, but anyways, it's much better than it used to be a year ago, when I had to manually go through the conversion tool and validate every constraint it adds.
- **16:05:16 · Dominik Kapusta:** Yes. Right. Mm-hmm. Mmhm. Mmhm.
- **16:05:27 · Alex Martemyanov:** Because currently, it took only… several dozen of iterations in the cloud. And that's all that it took.
- **16:05:35 · Dominik Kapusta:** No. Okay, okay, sounds great. Sounds great. Thanks for listening to me and for taking action going forward. Yeah, if you want… I can help you prepare the message if you want to just speak at Apple Weekly, be my guest. I'm here to help anyway. And that's it.
- **16:06:01 · Alex Martemyanov:** Yeah, I think if… if we could… if you could join the… the discussion, or maybe support the… the further movement, that could be, uh…
- **16:06:07 · Dominik Kapusta:** Mm-hmm. I'll try. I'll… I'll definitely try.
- **16:06:13 · Alex Martemyanov:** Everything.
- **16:06:14 · Dominik Kapusta:** Yeah. Like, the most important things are… For political reasons, don't make it… oriented around VS Code. But. Mention iOS. And, uh… and that's actually it. And make it essentially clear, like, working in the open, asking for feedback, and so on, so that you could do via tech design. And if you could reply to Juan in his comment on your project, that would also be great.
- **16:06:55 · Alex Martemyanov:** of…
- **16:06:56 · Dominik Kapusta:** Or not to Juan, but to everyone.
- **16:06:58 · Alex Martemyanov:** Yeah, maybe it was just mentioning, highlighting it during the weekly talk.
- **16:07:01 · Dominik Kapusta:** Yeah. Yeah. Yeah, yeah, yeah. Or you can just post a comment that… But yeah, we'll share more at the Apple Weekly today or something like that. Um… fine.
- **16:07:14 · Alex Martemyanov:** Okay.
- **16:07:15 · Dominik Kapusta:** We're done with this, we're over time, but one last thing that we should talk about is the… continuous development check-in. We won't have time for this today, so we can do it next week, even though the deadline is in two days, but that's fine. And I guess we can just keep it as is. Or we could do it… We could do it async. Let's see. I'll try to think about your growth areas. Uh, if you have… Any thoughts? Do you have time to look at your growth areas recently?
- **16:07:59 · Alex Martemyanov:** Mmhm.
- **16:08:04 · Dominik Kapusta:** I think things are going well with the project communication with OE, but…
- **16:08:05 · Alex Martemyanov:** I think it's, uh… It's mainly the project communication, the AOR and team work, the tech designs and…
- **16:08:15 · Dominik Kapusta:** Mmhm.
- **16:08:19 · Alex Martemyanov:** Again, keeping things visible that we are working on.
- **16:08:21 · Dominik Kapusta:** Yeah, I saw your comment on tech designs as well, um… Yeah, I made it… happen that you are assigned to some code reviews for the permissions, UI at least.
- **16:08:38 · Alex Martemyanov:** Yeah, we're working on with Andrea.
- **16:08:38 · Dominik Kapusta:** Permission changes.
- **16:08:41 · Alex Martemyanov:** On it, and yeah, actually we had some… Uh, some changes on these, on his last, um… Pull request, the one that you assigned to me because of all this swizzling that was originally made by me, and yeah, I think we… I basically opened another pull request, refactoring all this stuff and making it lots of easier to digest and…
- **16:08:55 · Dominik Kapusta:** Yeah. Right. Yeah. Okay, that's really great. That's really great.
- **16:09:07 · Alex Martemyanov:** Better documented.
- **16:09:10 · Dominik Kapusta:** Let's… okay, for one last thing, because I have to switch to another call, but… We should request feedback from someone. Steve, Gary… is missing in action. I think he's going on a paternity leave. Now, or next week, or I don't know. I pinged him, I created him a task. And, okay, he's not… He's preparing for paternity leave, anyway. Um, and uh… I created him a task, he didn't respond on the task. I know people ops are chasing people on feedback tasks, so maybe he'll get to it, maybe not. We have some material from the Doug Bill award, so we can… make the case for your good project management, at least. But… some other people that you could request feedback from. Uh, I think we were, uh, talking about that briefly last week. Uh, you mentioned Max. And Mihao as well.
- **16:10:16 · Alex Martemyanov:** Just…
- **16:10:18 · Dominik Kapusta:** Huh-huh. Good. Uh, is there anything in particular that you would like to ask them about? Because this feedback needs to be oriented around some events. So, this is, like, this is your project.
- **16:10:30 · Alex Martemyanov:** So, basically, CPM breakage… CPM breakage work, yeah.
- **16:10:32 · Dominik Kapusta:** Yeah. Okay. Um… All right, uh… I can send out that feedback, actually. And let's see. Good. Okay, um… All right, let's talk more about it next week and that will be it. Fine. Thank you. Thank you so much, and see you at the Apple Weekly.
- **16:11:04 · Alex Martemyanov:** Okay.
- **16:11:10 · Dominik Kapusta:** That's it. Let's talk about that next week as well.
- **16:11:17 · Alex Martemyanov:** Ah, yeah, and I got numbers for Xcode build, just let me quickly check it, if it gets any difference. Uh, for the main, yes, and there is 35-second decline, I mean, increase. improvement over the…
- **16:11:33 · Dominik Kapusta:** Improvement, okay.
- **16:11:35 · Alex Martemyanov:** was the main main branch so yeah.
- **16:11:37 · Dominik Kapusta:** That's good. That's great to hear.
- **16:11:38 · Alex Martemyanov:** 24% on the clean build.
- **16:11:42 · Dominik Kapusta:** Does it build VPN?
- **16:11:45 · Alex Martemyanov:** Uh, yes.
- **16:11:46 · Dominik Kapusta:** Nice.
- **16:11:46 · Alex Martemyanov:** I mean, the full… the full package… the full package rebuilt with export build, just with the main target refactor to SPM, instead of the main project.
- **16:11:48 · Dominik Kapusta:** Yeah, okay, that's great. Cool. Cool, cool. Okay, that's great to hear. Thank you and see you around.
- **16:12:05 · Alex Martemyanov:** Yep.
@@ -15,3 +15,305 @@ source: "https://us01mynotes.zoom.us/gUhtLoqRTfSsTbZsann73A"
- **20:44:49 · JŁ:** Hello. - **20:44:49 · JŁ:** Hello.
- **20:44:51 · Alex Martemyanov:** Yeah, so… But I think… There should be a PR open for the… Privacy config files. For Jacek… - **20:44:51 · Alex Martemyanov:** Yeah, so… But I think… There should be a PR open for the… Privacy config files. For Jacek…
- **20:45:03 · JŁ:** Yeah, just. Just one one thing when when you are fixing Alex this, have you been able to reproduce it or not? - **20:45:03 · JŁ:** Yeah, just. Just one one thing when when you are fixing Alex this, have you been able to reproduce it or not?
- **20:45:10 · Alex Martemyanov:** Yes, there is exact testing steps in the PR. So, basically, you have to skip the… got it. button press, and it puts you to the NTS restore loop. Uh, because the flag is set on when the button is clicked. Uh, so, the next time you open Duck AI, and you restart the app from that point, it will restart to the FHIR onboarding, locking in, uh, the Duck AI burning. So it's not a kind of blocking user experience, but it's a regression of user experience for sure.
- **20:45:48 · JŁ:** Okay, let me check it.
- **20:45:49 · Alex Martemyanov:** Uh, and, uh, disabling, there is no, there is no dedicated feature flag for this, uh, for this exact feature, but this is the whole, uh, tailored DACA AI. Onboarding flow. If we want to disable it, or maybe it's just easier to go with the…
- **20:46:07 · JŁ:** So sorry, if we disable it, what happens?
- **20:46:10 · Alex Martemyanov:** The whole dark AI onboarding vanishes.
- **20:46:14 · JŁ:** Okay.
- **20:46:17 · Alex Martemyanov:** So we learned to the old, uh, onboarding without the Duckinghide part.
- **20:46:20 · JŁ:** Mmhm.
- **20:46:22 · Alex Martemyanov:** So it kind of hurts the Duck AI, uh, product. But if we… if we push it to the hotfix or internal release, it just mitigates it right… right away.
- **20:46:22 · JŁ:** Okay. Yeah, yeah, so I think we need to…
- **20:46:33 · Alex Martemyanov:** So maybe it's easier, it's an easier option not to disable the feature flag. And just push it to the release. If it makes sense.
- **20:46:44 · JŁ:** I mean, uh. Not doing the hotfix or doing the hotfix.
- **20:46:50 · Alex Martemyanov:** I mean doing default fix without changing the feature flag.
- **20:46:54 · JŁ:** Okay.
- **20:46:57 · Bartek Waresiak:** Again… What's the plan? Sorry, I got lost because I just got into the infinite loop on my browser.
- **20:47:05 · JŁ:** Yeah, the… Okay, so like, you mean representation.
- **20:47:09 · Alex Martemyanov:** To release it. Related to 240 iOS release, so these…
- **20:47:14 · JŁ:** And…
- **20:47:18 · Alex Martemyanov:** This week's public release and make an assisted app store review. Right. How does it work?
- **20:47:35 · Hanyu Tang:** Okay, I can reproduce it on me.
- **20:47:41 · Alex Martemyanov:** Hmm? Sorry.
- **20:47:43 · Hanyu Tang:** I can reproduce it on me using the testing steps.
- **20:47:49 · Alex Martemyanov:** Cannot reproduce it. Why? On the broad branch.
- **20:47:54 · Hanyu Tang:** Yeah, I can refer to you soon. Uh, not your branch, on me.
- **20:47:57 · Alex Martemyanov:** Ah, you're good to be list. Okay.
- **20:48:00 · Hanyu Tang:** Yeah.
- **20:48:02 · Alex Martemyanov:** Yeah, it's not even produced, just exactly with the steps. So there is two main points, not pressing the button, and uh… staying on the CAI page when the app is, um… They're too dumb.
- **20:48:24 · Bartek Waresiak:** Again, not pressing which button.
- **20:48:25 · Alex Martemyanov:** The got it button. When there is a… you've got this screen. if you… Turn… turn down the app. If you turn it in the app at that stage, it won't record the correct state for the parallel marking. Then you have to ignore the… Final. the final onboarding button on one more thing that marks the onboarding as completed. and directly head to the address bar and start AI chat. So we detectively, uh… kind of lays over two issues that we stay in the state restoration, in the onboarding state restoration for the FHIR onboarding state, and the onboarding incomplete state. Because you didn't complete the onboarding in a normal way, and uh… instead of. completing the onboarding, the app considers the onboarding not finished and restores it to the FHIR onboarding state, because the DACA AI page was… There, when it was took down. So there is three factors activating this bug. And it should not have a huge effect on the users. Yes, I suppose.
- **20:50:02 · Bartek Waresiak:** Can I share my screen? Because I want to make sure that we are on the same page with this. So, basically… This is my application right now, okay?
- **20:50:16 · Alex Martemyanov:** Yep.
- **20:50:18 · Bartek Waresiak:** This is my application. And… they'll. If I kill it… And open it. I see this.
- **20:50:35 · Alex Martemyanov:** Yes. This is exactly the stage.
- **20:50:39 · Bartek Waresiak:** And… Pressing that… And I'll and starting any new new chat. And leaving this means that every restart I see that.
- **20:50:57 · Alex Martemyanov:** Yes.
- **20:51:00 · Bartek Waresiak:** And this is on the 240.
- **20:51:00 · Jonathan Jackson:** So if you would. So if you leave now, what happens?
- **20:51:03 · Bartek Waresiak:** Like the… And if I leave now…
- **20:51:06 · Jonathan Jackson:** Yep.
- **20:51:07 · Bartek Waresiak:** And come back, I'm back to onboarding.
- **20:51:12 · Alex Martemyanov:** Yep.
- **20:51:20 · Bartek Waresiak:** Can we control this with feature flag?
- **20:51:24 · Alex Martemyanov:** Uh, we can disable the docker, yeah, onboarding.
- **20:51:29 · Bartek Waresiak:** Like, I would say, like, I would disable it.
- **20:51:35 · JŁ:** The the question is disabling. It fixes it for the future users, or if someone is in the loop. It's going to fix it for them as well.
- **20:51:43 · Bartek Waresiak:** I would say eventually it should fix. This is not fully blocking the UI. I can press this. It's like the shitty experience. I can still use the app.
- **20:51:51 · Alex Martemyanov:** Yeah.
- **20:51:53 · Bartek Waresiak:** Right? If I kill this and resurface. it works, but if I, for any reason, end in the… And the flow in this. And kill.
- **20:52:11 · Jonathan Jackson:** So if you're in Duck AI when the app is quit, it brings you back with AI onboarding.
- **20:52:14 · Bartek Waresiak:** Yeah, this is insane loop. So every time I end up in the state that I've used Duck AI, I'm back to this annoying onboarding step.
- **20:52:26 · Jonathan Jackson:** Is that true for everybody who uses Duck AI at this point?
- **20:52:31 · Bartek Waresiak:** No, I think this is true for everyone that has onboarding state that is broken.
- **20:52:38 · Alex Martemyanov:** Yeah, so the ones who did not push the button.
- **20:52:38 · Jonathan Jackson:** So they stopped in the middle of the morning, basically. Okay.
- **20:52:42 · Bartek Waresiak:** Yep. So the friction.
- **20:52:44 · Jonathan Jackson:** So it's only for our new users who we're trying to push towards Duck AI.
- **20:52:48 · Bartek Waresiak:** Yes, exactly. That's the… that's exactly the impression I have. So, my… my intention here, like, intuitive intention would be to figure out.
- **20:52:50 · Alex Martemyanov:** Hmm. Okay.
- **20:52:52 · Jonathan Jackson:** Okay.
- **20:53:00 · Bartek Waresiak:** can feature flag. disable this. What kind of feature flag do we have for us for this?
- **20:53:08 · Alex Martemyanov:** We have, uh…
- **20:53:08 · Jonathan Jackson:** Yep.
- **20:53:10 · Bartek Waresiak:** Oh, I don't. For some reason.
- **20:53:12 · Alex Martemyanov:** Custom product page deck AI onboarding flow.
- **20:53:16 · Jonathan Jackson:** So we would disable the entire onboarding flow.
- **20:53:17 · Bartek Waresiak:** I don't have.
- **20:53:21 · Jonathan Jackson:** For Dec AI.
- **20:53:21 · Alex Martemyanov:** Yeah, try setting the flag.
- **20:53:24 · Bartek Waresiak:** Yeah, I need to log in to…
- **20:53:28 · Jonathan Jackson:** I don't know if that's necessarily better.
- **20:53:32 · Bartek Waresiak:** Not necessarily better what?
- **20:53:32 · Jonathan Jackson:** Because we wouldn't be giving, it sounds like we wouldn't be giving them a decade onboarding at all. Based on the flag name anyway.
- **20:53:41 · Alex Martemyanov:** That's… That depends.
- **20:53:46 · Jonathan Jackson:** Or is it only for people coming from a custom product page?
- **20:53:51 · Bartek Waresiak:** Like, this is the onboarding when you install the app.
- **20:53:51 · Alex Martemyanov:** You know.
- **20:53:54 · Jonathan Jackson:** Yeah, but Alex just shared the… what I think is the feature flag name, right, Alex?
- **20:53:59 · Alex Martemyanov:** On boarding. Hmm.
- **20:54:05 · Jonathan Jackson:** Okay.
- **20:54:06 · Bartek Waresiak:** Like the question is.
- **20:54:08 · Jonathan Jackson:** Do we have a mitigation?
- **20:54:08 · Bartek Waresiak:** This onboarding can effectively set people in this broken state.
- **20:54:14 · Jonathan Jackson:** Right.
- **20:54:14 · Bartek Waresiak:** Plus, we have people in this broken state that basically cannot recover.
- **20:54:20 · Jonathan Jackson:** Yeah. Mm-hmm.
- **20:54:22 · Bartek Waresiak:** So the question is. What's more important, to fix that on the actual users right now or to still keep presenting this onboarding? Potentially bringing in more bleeding.
- **20:54:35 · Jonathan Jackson:** Thank you. Yeah, yeah, yeah.
- **20:54:39 · JŁ:** Yeah, but… So I think we don't know if again, like the question I asked, if. disabling the flag fixes it for like fixes the people that are in the loop. you know, because if if it doesn't, then. I think we should just go with the hot fix as soon as possible, and if it does, then yeah, we could. We could turn the flag for today and still go with a hot fix. Probably like.
- **20:55:04 · Jonathan Jackson:** Yeah, I think we have parallel work streams here. We have one looking into what feature flag we have and what we can do to mitigate as quickly as possible. And we have another looking at what's the right hotfix.
- **20:55:17 · Alex Martemyanov:** No, basically the flags won't fix it. They will only fix it for the new users by not showing the onboarding step.
- **20:55:28 · Jonathan Jackson:** But the users would be stuck still.
- **20:55:30 · Alex Martemyanov:** Yep. who already have this issue, it's not gonna fix it.
- **20:55:34 · Jonathan Jackson:** Yeah. So then, Bartek, your question, I think, is right. Does disabling it and preventing new people from ending up in this position make sense or not?
- **20:55:46 · Bartek Waresiak:** Like, the question is, can we expedite the code fix to… have it rolled out as soon as possible. Like, basically, this is what we need. If we cannot fix this. With…
- **20:56:05 · Jonathan Jackson:** I mean, do we have a sense of what an actual fix looks like yet?
- **20:56:11 · Alex Martemyanov:** Yes, the pull request is open for the actual fix. It's basically resetting this log.
- **20:56:22 · Jonathan Jackson:** Okay. And this is iOS only, is that right?
- **20:56:26 · Alex Martemyanov:** Yes.
- **20:56:31 · JŁ:** Oh, okay, I think…
- **20:56:32 · Hassaan Elgarem:** So I I want to. Yeah, I want to share my screen and show like, I think, Jonathan and Alex, you're in there when we talked about that. But I I came across a different issue. I'm not sure if it's related or not. Let me show you. um… Okay. It's this one. So when going through the onboarding flow and then choosing like one of the pre-filled prompts, I reach the stuck. State. and it takes a very long time like we did it on the call before you joined, and we waited. Maybe I'm not sure, like more than a minute till something actually happened.
- **20:57:25 · Jonathan Jackson:** Fun.
- **20:57:25 · Hassaan Elgarem:** And only the only the only thing that's. Interactive is the Firebolt. Right now.
- **20:57:31 · Alex Martemyanov:** It's basically a known issue if you have a connection loss or a weak connection. It would behave like this, yes, and it's kind of intended behavior.
- **20:57:42 · Jonathan Jackson:** Do you have a reconnection?
- **20:57:42 · Alex Martemyanov:** So, they will fire buttons and… Yes, it's, uh, it's loading. Because it starts loading. Maybe there is, like… I don't know. There shouldn't be CPM compilation, but if it takes to one minute to load, it means that it's, yeah, it's loaded, basically.
- **20:57:55 · Bartek Waresiak:** Do we…
- **20:57:58 · JŁ:** Wait. Wait. It's crazy if it's intended, sorry.
- **20:58:05 · Bartek Waresiak:** Yeah, like, okay, like, I can buy it is acceptable. But… I would only make it acceptable if we have metrics to understand that this is not a problem on a wider scale. Do we have metrics? For how long does it take this initial load?
- **20:58:23 · Alex Martemyanov:** Nope, we don't.
- **20:58:27 · JŁ:** Probably some.
- **20:58:27 · Alex Martemyanov:** Okay.
- **20:58:28 · JŁ:** Someone would be good person to address it like he. He was doing some. And… SLIs. Air.
- **20:58:38 · Hassaan Elgarem:** So my my question here is to Alex like, if this problem is like an existing problem. But if this problem became more frequent, would it trigger more issues? Of the actual, like, loop that we are discussing.
- **20:58:54 · Alex Martemyanov:** Nope, it's just long loading. There is probably nothing to do on the native side about it.
- **20:58:58 · Jonathan Jackson:** Okay. Hassan, was that in a link condition scenario so your network was slow, or was that just a normal load?
- **20:59:07 · JŁ:** Sorry, sorry, I'm not, I'm not, I'm not buying it, like, it's, it's, it's a site, right? What does it mean? It's an intended behavior, and…
- **20:59:09 · Hassaan Elgarem:** No, I'm good.
- **20:59:14 · Jonathan Jackson:** Yeah.
- **20:59:18 · Alex Martemyanov:** We don't show the, uh… We don't show the content until it is completely loaded with all the contents.
- **20:59:26 · JŁ:** Okay, but… but, like, Hassan's on a Zoom here with great connectivity, like, how… I… I don't get it, like… Probably any site would load immediately, like.
- **20:59:39 · Alex Martemyanov:** Maybe there is a signal… But yeah, so we just seen the appearance of the text started after it appeared.
- **20:59:50 · Jonathan Jackson:** Yeah, but it looks like a broken state. It looks like the page has failed to render.
- **20:59:50 · Alex Martemyanov:** So, basically, the loading.
- **20:59:55 · Jonathan Jackson:** And so maybe that's the implementation right now, but I thought it was very broken.
- **20:59:58 · Alex Martemyanov:** No, it's, uh, it's not… it's not the… it's not the rendering failure, it's loading, because, uh, it's…
- **21:00:04 · Jonathan Jackson:** Yeah but the user clicks the user goes to that page on a good connection and they have to wait thirty seconds to a minute for anything to happen.
- **21:00:10 · Alex Martemyanov:** There is a signal.
- **21:00:10 · Jonathan Jackson:** That doesn't look…
- **21:00:11 · Bartek Waresiak:** One moment.
- **21:00:12 · Jonathan Jackson:** Yep.
- **21:00:13 · Bartek Waresiak:** What I would expect to happen when the button is pressed is that the question is transferred to the bubble, chat bubble. is sent to the backend. And we render the question that is being asked. Then we see generating response prompt that is ongoing and this didn't happen.
- **21:00:37 · Alex Martemyanov:** This is logical, but it's not the implementation. It's, uh… it's done this way to…
- **21:00:41 · Jonathan Jackson:** So…
- **21:00:45 · Alex Martemyanov:** show the blank screen until the answer is ready. So when the connection is final and the signal arrives in time, it should render just in time and appear without all this page flickering.
- **21:01:01 · Bartek Waresiak:** Was that raised in ship review and presented in such way?
- **21:01:01 · Alex Martemyanov:** But there was… Nope. during the ship review, there… the issue was not reproduced, and basically it appears during 2 seconds. And yes, it was, uh… this way during the ship review. So there was no bubble, nothing animated, and so on.
- **21:01:21 · JŁ:** I called Fernando. to this meeting. I just, it's hard for me to believe like.
- **21:01:26 · Jonathan Jackson:** Well, so, I mean, it sounds like… it sounds like that's maybe a separate user experience issue, which we should have a task for and address. Hasan, your question is still valid, though, related to the incident here.
- **21:01:29 · JŁ:** But.
- **21:01:30 · Alex Martemyanov:** Thank you.
- **21:01:32 · Bartek Waresiak:** Yeah, yeah.
- **21:01:37 · Jonathan Jackson:** Does that behavior put us into a position where more users are at risk of ending up in this loop state?
- **21:01:47 · Alex Martemyanov:** So there may be an issue with this front-end signal from Duck AI page that we used to start rendering, but as we've seen, the actual animation has started after a significant delay, so it means the signal is.
- **21:01:58 · Bartek Waresiak:** Okay.
- **21:02:01 · Alex Martemyanov:** synchronized with this appearance, this presentation slide.
- **21:02:01 · Jonathan Jackson:** Right. But if that's happening during onboarding, and the entry to the loop is hitting the fire button.
- **21:02:03 · Bartek Waresiak:** I, I have, I have…
- **21:02:07 · Jonathan Jackson:** It's going to encourage more people to hit the fire button than might otherwise.
- **21:02:11 · Alex Martemyanov:** Maybe.
- **21:02:12 · Bartek Waresiak:** I have one question.
- **21:02:12 · Hassaan Elgarem:** hitting the firewalls and and killing the app like killing the app like, if I'm in that state. I would think of killing the app, and that's what like triggers the loop. So.
- **21:02:14 · Jonathan Jackson:** Right. Because they think it's stuck. Yeah.
- **21:02:23 · Hassaan Elgarem:** I think they are related.
- **21:02:24 · Jonathan Jackson:** I don't know that we need a fix for the user experience issue we're talking about, but it is a factor in putting people into that state. From a behavior perspective, I'd buy that.
- **21:02:26 · Bartek Waresiak:** Yeah, like, the… from the perspective on onboarding, when I select Duck AI onboarding, when I select some random prompt. And it loads. When I delete this… When I use the fire button. It should work okay from that point onwards. Or is the condition flux set somewhere else? That's what I don't really follow.
- **21:03:15 · Alex Martemyanov:** The condition set, the flag is, the flag for the fire onboarding is set on the… Prior. Screen. I mean, when do you have this AI answer presented? Uh, it sets the state restoration, the onboarding set restoration plug. to… Awaited. So basically, yes, it may… It may indicate this.
- **21:03:45 · Bartek Waresiak:** So when I got the response.
- **21:03:49 · Alex Martemyanov:** When you get the response, the flag is turned on. And when you… Push the got it button, it's, uh, removed.
- **21:03:59 · Bartek Waresiak:** Okay, so I'm I'm trying to reproduce it right now. And I'm… Question to AI.
- **21:04:10 · JŁ:** In the meantime, Hassan, could you show Fernando the finding that you have as well? Another was was working on the Ki.
- **21:04:19 · Hassaan Elgarem:** Give me a minute to prepare it and I'll share my screen.
- **21:04:24 · JŁ:** Thank you. Okay, I'm going to focus now on the official part, so I will officially start the incident, okay, because… Yeah, I have issues with multitasking. So yeah, so I will leave you like I'll be here. But I will leave you guys to to questions, answers, and more investigation. And just ping me if anything.
- **21:05:03 · Bartek Waresiak:** No, like, it's like, sorry, but all of that just doesn't make sense to me. It's like super easy to trigger for me to enter the loop.
- **21:05:20 · Fernando Bunn:** What it is.
- **21:05:22 · Bartek Waresiak:** Like, I…
- **21:05:22 · Fernando Bunn:** Sorry, I don't know exactly what is going on.
- **21:05:24 · Bartek Waresiak:** Let me reinstall the app and share my screen again, so I will show you what I mean by super easy.
- **21:05:29 · Fernando Bunn:** Okay.
- **21:05:31 · Jonathan Jackson:** While you're doing that, the short version is, if you go through the Duck AI onboarding, it gets you to the point where it's given a random prompt to Duck AI. You hit the fire button, and then you quit the app. And from that point on, anytime you leave… anytime you quit the app while you're in the Duck AI experience, it will bring you back to the random prompt happening when you come back to the app. So it feels like you're stuck on onboarding and it basically keeps bringing you back to that. I don't know if you have to quit the app for that to happen or if it's like when you leave and come back after a while, if you were in Duck AI at that point.
- **21:05:56 · Fernando Bunn:** Mmhm.
- **21:06:06 · Jonathan Jackson:** I think it happens again. I'm not sure about that piece. But at the very least, when people relaunch the app at that point, they keep getting the Duck AI out.
- **21:06:13 · Fernando Bunn:** Okay. So if we relaunch the IPS2… Okay, let me… Yeah, yeah, yeah, you can see it.
- **21:06:17 · Bartek Waresiak:** Yeah, can you… Yeah, okay. Can you hear me as well?
- **21:06:22 · Fernando Bunn:** Yep.
- **21:06:23 · Bartek Waresiak:** Excellent. So this is like, I ignore this. It's in Polish, but it doesn't really matter. I ignore this.
- **21:06:29 · Fernando Bunn:** Mm-hmm.
- **21:06:29 · Bartek Waresiak:** I ignore the color, whatever, top, and now I want to. Check the AI about the stock market. Okay, so I'm waiting here. This looks good to me, right? I look nice. I'm supposed to clear it. Okay, so I'm clearing this. Excellent. This looks okay. I want to chat. Okay. I accept the terms. Everything looks good. Perfect. So I move away. And I'm in the loop.
- **21:07:14 · Alex Martemyanov:** So. Ah, you're telling there is no… you've got this step at that point?
- **21:07:18 · Bartek Waresiak:** Yeah, exactly.
- **21:07:20 · Alex Martemyanov:** Ah…
- **21:07:20 · Bartek Waresiak:** This is everything here.
- **21:07:22 · Alex Martemyanov:** That is the change. And when was it done?
- **21:07:26 · Bartek Waresiak:** This is the current production.
- **21:07:30 · Alex Martemyanov:** When was it changed? Maybe there is a phishing flag.
- **21:07:35 · Bartek Waresiak:** Like, this is, like, I'm hitting this literally every time I install the app.
- **21:07:40 · Alex Martemyanov:** Hmm, that changes things, and that explains why we started seeing it just now.
- **21:07:46 · Bartek Waresiak:** Oops, sorry, I'm…
- **21:07:47 · Alex Martemyanov:** If the onboarding was changed, so this…
- **21:07:51 · Bartek Waresiak:** Uh, I just real…
- **21:07:52 · Alex Martemyanov:** I think made of those.
- **21:07:52 · JŁ:** My pleasure.
- **21:07:53 · Fernando Bunn:** Just one thing, just to — sorry, just help me. OK.
- **21:07:53 · JŁ:** Can you…
- **21:07:55 · Bartek Waresiak:** Sorry, I have another call I need to jump on, but I believe this is a problem that needs to be surfaced, okay? So…
- **21:08:03 · JŁ:** Bartek, one more thing. Can you record the thing and upload it? Send me so we can have like a.
- **21:08:12 · Bartek Waresiak:** I, I can, like, give me, like, 2 minutes, okay?
- **21:08:12 · Fernando Bunn:** I'll try to do it here in the simulator, yeah.
- **21:08:13 · JŁ:** For experience. Thank you. Okay, sure.
- **21:08:16 · Fernando Bunn:** Just one thing, I don't know if it… like, it could give a hint on what's going on, okay? When I saw the onboarding thing, what it did, I didn't see the terms and conditions. anywhere…
- **21:08:28 · JŁ:** it…
- **21:08:29 · Fernando Bunn:** Is this expected?
- **21:08:31 · JŁ:** There… there were, like, I think, when the chat appeared. I think for the first chat, it didn't appear. I think there was a discussion on Mattermost today on OJ Channel about it. I think Picor asked about it. Let me search quickly.
- **21:08:35 · Fernando Bunn:** Yep. Yep. Okay. No, yeah, I saw the discussion, but…
- **21:08:48 · JŁ:** Shouldn't we already have a privacy information at this stage? And Chuck answered, we have dispensation to allow one prompt here without asking users to agree to the TOS.
- **21:08:59 · Fernando Bunn:** Oh, what prompts? Okay.
- **21:08:59 · JŁ:** So after… Yeah, so Bartek prompted again, and there was a button. So, so did this one work?
- **21:09:05 · Fernando Bunn:** Yeah, okay, so that's fine, okay, so that's fine.
- **21:09:08 · JŁ:** Yeah, but do you have?
- **21:09:09 · Fernando Bunn:** of…
- **21:09:10 · Hassaan Elgarem:** Fernando? Yeah, I have it ready. Let me show you.
- **21:09:11 · JŁ:** that that.
- **21:09:15 · Hassaan Elgarem:** The other thing.
- **21:09:17 · Fernando Bunn:** Audrey's name?
- **21:09:18 · Hassaan Elgarem:** Okay, so I went through the onboarding.
- **21:09:19 · Fernando Bunn:** Okay.
- **21:09:19 · JŁ:** Thank you. The other thing.
- **21:09:26 · Hassaan Elgarem:** And then ask AI, and then… And then I'm stuck in this state. for a while, like it normally takes like 30 to 30 seconds to one minute. And the only thing that I can I can interact with is the fire button. And once I do, actually, it starts pulsing. So it's a weird thing.
- **21:09:50 · Alex Martemyanov:** Okay, just go, just go on deleting it.
- **21:09:51 · Fernando Bunn:** And this is… and… And this is not… you're not using, kind of, uh…
- **21:09:58 · Hassaan Elgarem:** Natural condition or anything? Nothing.
- **21:09:58 · Fernando Bunn:** Uh…
- **21:10:00 · JŁ:** No, no, not using custom chat.
- **21:10:00 · Fernando Bunn:** Okay, it's vanilla, okay.
- **21:10:02 · JŁ:** Yes.
- **21:10:03 · Hassaan Elgarem:** No.
- **21:10:04 · JŁ:** And, yeah, and…
- **21:10:04 · Alex Martemyanov:** Maybe it's…
- **21:10:06 · JŁ:** I think Alex was mentioning that this is by design.
- **21:10:09 · Fernando Bunn:** What?
- **21:10:11 · JŁ:** that…
- **21:10:11 · Alex Martemyanov:** No, it's… basically, I don't see this behavior at my place, and… Even from Kyrgyzstan, it loads fast enough, and I don't think it's, uh… Frequent problem. Because it actually indicates it's not loading.
- **21:10:28 · JŁ:** Yeah, yeah, bye. Right. But am I right that you are saying that it's basically like the the connectivity issue. And.
- **21:10:36 · Alex Martemyanov:** It looks mostly like the connectivity issue.
- **21:10:37 · Fernando Bunn:** So how… I'll do it for those who just… Go to SKI and use a pre-filled answer there.
- **21:10:47 · Hassaan Elgarem:** Yeah. Okay.
- **21:10:51 · Fernando Bunn:** Yeah, it works fine here.
- **21:10:51 · Alex Martemyanov:** Yes, so we just loaded. It's either some routing issue, or DACRIA issue, or… And frankly, I don't know. How would you fix it without redesigning it really?
- **21:11:11 · Fernando Bunn:** So if I close this now…
- **21:11:11 · Alex Martemyanov:** With some lighting bubbling.
- **21:11:14 · Fernando Bunn:** I'm gonna see the look again.
- **21:11:15 · Alex Martemyanov:** Yeah, so that screen was high-five Bartek. When Bartek showed, demoed it, there was no high-five screen.
- **21:11:27 · Fernando Bunn:** Wait, I killed the app, and I went back, and I… I'm not on the KI anymore, so what… what happened with Barton.
- **21:11:28 · Hassaan Elgarem:** the…
- **21:11:34 · Alex Martemyanov:** Yeah, you just killed the app now.
- **21:11:36 · Fernando Bunn:** Yeah.
- **21:11:38 · Alex Martemyanov:** And you go back.
- **21:11:40 · Hassaan Elgarem:** But now it's fast, like now, like I don't have to wait or anything.
- **21:11:45 · Alex Martemyanov:** You're weird.
- **21:11:48 · Hassaan Elgarem:** So it's not really like a connection thing. Like, I think something is off.
- **21:11:55 · JŁ:** Hassan, can you actually trace it in a debugger? like, what is going on in the meantime, like, like, perhaps with a debugger and some, you know, some proxy, whatever thing, like, checking if there are any requests going in and out, okay? Like…
- **21:12:03 · Hassaan Elgarem:** That's right. But, like, this is… Yeah, yeah, yeah, I'll do that. I'll do that.
- **21:12:17 · JŁ:** Thank you so much.
- **21:12:18 · Hassaan Elgarem:** But again, like, if I kill the app and reopen, I'm back in this state again. That's the cherry on top, you know.
- **21:12:31 · JŁ:** So what you mean, like, you killed the app, and…
- **21:12:34 · Hassaan Elgarem:** like what I'm trying to reproduce the the original issue.
- **21:12:37 · JŁ:** Okay.
- **21:12:37 · Fernando Bunn:** But if you see the logs, what is happening there? Is the app actually trying to hit the KI and just latency or? Something else. Like, there's a lot to tell. Anything if you reopen it via Xcode?
- **21:12:52 · Hassaan Elgarem:** I'll try, I'll try, like, I'm… right now, the debugger is not attached, but, like, I'll try… I'll stop sharing. I'll try and let you know if I find something.
- **21:13:03 · Fernando Bunn:** Try to reproduce vertex stuff, but it's not working here.
- **21:13:04 · Alex Martemyanov:** What is this app? What's the app you're sharing? Some… It's not the simulator, is it?
- **21:13:22 · Hassaan Elgarem:** Sorry?
- **21:13:23 · Alex Martemyanov:** Vijay, what is the app you were sharing? It's not a simulator, is it?
- **21:13:27 · Hassaan Elgarem:** Okay. It's the new devices hub. Xcode 27, I think.
- **21:13:35 · Alex Martemyanov:** Is it the same on the regular simulator?
- **21:13:41 · Hassaan Elgarem:** Yeah, it should be, yeah.
- **21:13:41 · Alex Martemyanov:** Or a real device.
- **21:13:43 · Hassaan Elgarem:** It's like, uh…
- **21:13:44 · JŁ:** The same.
- **21:13:45 · Hassaan Elgarem:** Yep.
- **21:14:08 · Fernando Bunn:** Was anyone able to produce vertex stuff through Xcode or just from the App Store?
- **21:14:29 · Hassaan Elgarem:** Fernando, can you repeat your question? Sorry.
- **21:14:33 · Fernando Bunn:** I'm asking if anyone was able to reproduce bar type issue using Xcode or just via App Store. Just to see if the bug release has anything to do with it.
- **21:14:46 · Hassaan Elgarem:** Uh, yeah, like, I'm testing from the Xcode, and I'm able to reproduce the loop, but it's a loop plus the delay, like, I'm getting both issues.
- **21:14:55 · Fernando Bunn:** Oh, okay. Okay, so it's not just a release thing. Okay.
- **21:14:59 · Hassaan Elgarem:** Okay.
- **21:15:06 · Fernando Bunn:** Moving forward… Short story, uh… Skip. Excuse me. Next. Okay. Bago. So I go to Tokyo, explain the stock market. So at this moment, if I… First, close the app and kill it. I'll be on the loop, right? Like, I do not interact with the thing, or it doesn't matter.
- **21:15:43 · Alex Martemyanov:** You interact with it, you burn it. Use the fire button. Yeah, and then you kill it.
- **21:15:47 · Fernando Bunn:** Oh, I need to burn. Okay, then I see this. Should I go to…
- **21:16:02 · Alex Martemyanov:** Yes, and you enter… you activate the address bar, DACAI. And then you killed it again, and you got in the hoop.
- **21:16:11 · Fernando Bunn:** So we'll go to new chat.
- **21:16:13 · Hassaan Elgarem:** Start a new chat and then kill the app.
- **21:16:16 · Fernando Bunn:** Cool, again. Okay.
- **21:16:24 · JŁ:** You got it?
- **21:16:26 · Fernando Bunn:** Yep. Just want to see if I call from Xcode now. Seems bad. So I need to get to the loop. I need to do the same thing again, like fire button, kill the app, and…
- **21:16:45 · Alex Martemyanov:** Uh, yes, you fire button… Then you start Duck AI Chat again, and then you kill it up.
- **21:16:52 · Fernando Bunn:** Done.
- **21:16:53 · Alex Martemyanov:** It's only if you… if you give it 12, the exact AI chat is active.
- **21:16:53 · Fernando Bunn:** Queued up. Oh, I need to kill while he's acting.
- **21:17:00 · Alex Martemyanov:** Yes. So, uh… So what's the resolution we have? We need to rebase the PR to the current release, right? Two seven point. to… 4… 0… and have an expedited review, right? Basically, that's the only option.
- **21:17:28 · Fernando Bunn:** Can I just remove… Is this specific, the KI part and the onboarding behind FeatureFlight?
- **21:17:35 · Alex Martemyanov:** Yes.
- **21:17:37 · Fernando Bunn:** So it's already off.
- **21:17:37 · Alex Martemyanov:** The specific… It's… It's already off.
- **21:17:43 · Fernando Bunn:** But I just reproducing the episode. Okay, so maybe it didn't…
- **21:17:48 · Alex Martemyanov:** Yeah, we cannot disable it.
- **21:17:54 · Fernando Bunn:** Sorry, is it available in production or not?
- **21:17:58 · Alex Martemyanov:** Hmm?
- **21:17:59 · Fernando Bunn:** The onboarding… the Deck AI onboarding step is disabled in production right now, or not? Like, did we disable using a feature flag?
- **21:18:04 · Alex Martemyanov:** Mm-hmm. There is no feature flag.
- **21:18:07 · Fernando Bunn:** Oh.
- **21:18:09 · Alex Martemyanov:** I mean, there was, but it's… Not a thing already. And. So I didn't get it if, uh… Was there… you've got the screen or not?
- **21:18:30 · Fernando Bunn:** I'm going to see if I can reproduce from Axcode catalog. Let's try and compile it for a second. I don't have… Morocco.
+4 -4
View File
@@ -56,8 +56,8 @@
"note_modified_at": "2026-10-06T15:22:23.633Z", "note_modified_at": "2026-10-06T15:22:23.633Z",
"source_url": "https://us01mynotes.zoom.us/gUhtLoqRTfSsTbZsann73A", "source_url": "https://us01mynotes.zoom.us/gUhtLoqRTfSsTbZsann73A",
"folder": "2026-10-06 Jacek Łyp's Zoom Meeting 2026-10-06 20 44(GMT+6 00) gUhtLoqRTfSsTbZsann73A", "folder": "2026-10-06 Jacek Łyp's Zoom Meeting 2026-10-06 20 44(GMT+6 00) gUhtLoqRTfSsTbZsann73A",
"sha256": "3dfaa8d06a1080f47438ca414ccb005e99f5223cc20f9d4ff143e697616c3f06", "sha256": "84d556b621750e2f485ca950b883923701dedb9b03fd4bc601cb27e7ca7999eb",
"cues": 3 "cues": 305
}, },
"280BehgySpW8LynSS473nQ": { "280BehgySpW8LynSS473nQ": {
"title": "Apple Weekly Team Meeting - Zoom ID 2026-10-06 20:01(GMT+6:00)", "title": "Apple Weekly Team Meeting - Zoom ID 2026-10-06 20:01(GMT+6:00)",
@@ -84,8 +84,8 @@
"note_modified_at": "2026-10-06T10:13:40.063Z", "note_modified_at": "2026-10-06T10:13:40.063Z",
"source_url": "https://us01mynotes.zoom.us/NtgNJ-Y-Qxy5sMGMjngukw", "source_url": "https://us01mynotes.zoom.us/NtgNJ-Y-Qxy5sMGMjngukw",
"folder": "2026-10-06 Dominik Kapusta's Personal Meeting Room 2026-10-06 15 12(GMT+6 00) NtgNJ-Y-Qxy5sMGMjngukw", "folder": "2026-10-06 Dominik Kapusta's Personal Meeting Room 2026-10-06 15 12(GMT+6 00) NtgNJ-Y-Qxy5sMGMjngukw",
"sha256": "7366ebe376ba9d8277ef2f2df996e4dd0aad961f3463083e6c9025f72907c213", "sha256": "39d652519c706479f5411755effe0cad959bde0f5e0459ede812f90e41a84560",
"cues": 3 "cues": 294
}, },
"klekRLtLSEuMv0PCCPCjtA": { "klekRLtLSEuMv0PCCPCjtA": {
"title": "All-Company Meeting 2026-10-16 21:31(GMT+6:00)", "title": "All-Company Meeting 2026-10-16 21:31(GMT+6:00)",
+1 -1
View File
@@ -1 +1 @@
{"observed_at":"2026-11-20T21:02:24.649Z","scanned":33,"imported":2,"skipped":31,"withoutTranscript":0,"coverage":"complete","sections":[{"name":"Owned by me","pages":5,"notes":88,"complete":true,"terminal_cursor":true},{"name":"Shared with me","pages":1,"notes":0,"complete":true,"terminal_cursor":true},{"name":"Starred","pages":1,"notes":0,"complete":true,"terminal_cursor":true}],"discovered":88,"gaps":[],"recovered_transcripts":0,"policy":"reread_all_owned_shared_and_starred_notes_since_cutoff","since":"2026-09-30"} {"observed_at":"2026-11-20T21:33:48.204Z","scanned":33,"imported":2,"skipped":31,"withoutTranscript":0,"coverage":"complete","sections":[{"name":"Owned by me","pages":5,"notes":88,"complete":true,"terminal_cursor":true},{"name":"Shared with me","pages":1,"notes":0,"complete":true,"terminal_cursor":true},{"name":"Starred","pages":1,"notes":0,"complete":true,"terminal_cursor":true}],"discovered":88,"gaps":[],"recovered_transcripts":0,"policy":"reread_all_owned_shared_and_starred_notes_since_cutoff","since":"2026-09-30"}