Firefox 70 Desktop looks to be a quiet release compared to FF65-69 based on the cumulative sum of the SUMO Forum questions in the first two weeks
Pontifications
- Based on the plot below, showing the first two weeks of Firefox 70 Desktop, it looks like Firefox 70 in the first four weeks will be a quiet release. (see earlier post: Firefox 65-69 Cumulative Sum Day 1-28 of Firefox Support Questions on support.mozilla.org)
- I could be wrong of course :-) and something could emerge in the next two weeks.
Plot: Firefox 70 first two weeks and Firefox 65-69 first four weeks: cumulative sum of SUMO Forum desktop questions
Code for the above plot
cumulative-sum-unfaceted-january-04november2019-plot-num-questions-release-day-1-28.R :
library(tidyverse)
library(lubridate)
library(directlabels)
library(RColorBrewer)
add_release_day_number <-
function(df_release,
yyyy,
mm,
dd)
{
START_DATE <-
make_datetime(yyyy, mm, dd, 0, 0, 0,
tz = "UTC")
return (df_release %>%
mutate(release_day_number =
(floor(interval(
START_DATE, created
) / days(1))) + 1))
}
create_desktop_df_release_week_num_questions <-
function(df, release, yyyy, mm, dd)
{
# df is CSV with date time, release is "65"
# yyyy, mm, dd are integers e.g. 2019, 1, 29
# remove all questions before january 29, 2019
ymd_str <- sprintf("%d-%d-%d", yyyy, mm, dd)
release_start <- ymd(ymd_str, tz = "UTC")
release_end <- release_start + weeks(4)
release_questions <-
df %>%
filter(created >= release_start & created < release_end)
# add day of release week i.e, 1, 2, 3, 4, 5,6, 7..28
release_questions <-
add_release_day_number(release_questions, yyyy, mm, dd)
release_questions <- release_questions %>%
group_by(release_day_number) %>%
count()
release_questions <- add_column(release_questions, release = release)
return (
release_questions %>%
ungroup() %>%
dplyr::mutate(cumulutive_sum = cumsum(n)))
}
jan_04nov_2019_questions <-
read_csv("https://raw.githubusercontent.com/rtanglao/rt-kits-api2/master/2019BYMONTH/concatenated-2019-01-01-2019-11-04-firefox-creator-answers-desktop-all-locales.csv")
# change created unix time to r time UTC using as_datetime()
jan_04nov_2019_questions <-
jan_04nov_2019_questions %>%
mutate(
created = as_datetime(created, tz = "UTC")
)
ff65_questions <- create_desktop_df_release_week_num_questions(
jan_04nov_2019_questions, "65", 2019, 1, 29)
ff66_questions <- create_desktop_df_release_week_num_questions(
jan_04nov_2019_questions, "66", 2019, 3, 19)
ff67_questions <- create_desktop_df_release_week_num_questions(
jan_04nov_2019_questions, "67", 2019, 5, 21)
ff68_questions <- create_desktop_df_release_week_num_questions(
jan_04nov_2019_questions, "68", 2019, 7, 9)
ff69_questions <- create_desktop_df_release_week_num_questions(
jan_04nov_2019_questions, "69", 2019, 9, 3)
ff70_questions <- create_desktop_df_release_week_num_questions(
jan_04nov_2019_questions, "70", 2019, 10, 22)
jan_04nov_2019_questions_by_release_week <-
bind_rows(ff65_questions, ff66_questions, ff67_questions,
ff68_questions, ff69_questions, ff70_questions)
jan_04nov_2019_plot <-
ggplot(data=jan_04nov_2019_questions_by_release_week,
aes(x=release_day_number, y=cumulutive_sum, group=release,
colour = factor(release)))
x_axis = sprintf("%d", seq(1:28))
jan_04nov_2019_plot = jan_04nov_2019_plot +
geom_line(stat="identity") +
labs(color = 'Release Week 1-4') +
scale_x_discrete(limits = x_axis) +
coord_cartesian(clip="off") +
labs(color = 'DesktopAAQ65-70') +
geom_dl(aes(label = release), method = list(dl.trans(x = x + 0.2), "last.points", cex = 0.8)) +
geom_dl(aes(label = release), method = list(dl.trans(x = x - 0.2), "first.points", cex = 0.8)) +
scale_color_brewer(palette = "Dark2")