Testing for it in Python means that we have to worry about keeping the
`upgrade-zulip-stage-2` backwards-compatible with all versions of
Python which we could ever be upgrading from -- which is all of them.
Factor out the "supported operating systems" check, and share it
between upgrade and install codepaths.
soupsieve is a heavy-weight dependency, and Tornado pulls it in by way
of markdown rendering; since we are only using it for a very simple
process, perform that manually.
Per CSS spec[^1]:
> In quoted <string> url()s, only newlines and the character used to
> quote the string need to be escaped.
[^1]: https://drafts.csswg.org/css-values/#urls
This middleware was highly-specific to a set of URLs, and pulled in a
beautifulsoup dependency for Tornado. Move it closer to where it is
used, minimizing action at a distance, as well as trimming out a
dependency.
9a682fb20a started performing message fetching in a read-only
transaction. However, our use of `get_or_create_huddle` can violate
the read-only promise, and result in a user-facing 500.
In the cases where we're attempting to narrow to a huddle that does
not exist, this is equivalent to a false condition; catch those,
without making the huddle row, and insert a false.
Earlier, we were not verifying that the invoice which got paid is
for the fixed-price plan.
That could result in a bug where another support invoice with
collection_method = "send_invoice" got paid while a fixed-price
plam is already configured. The fixed-price plan would be falsely
activated.
This commit verifies the invoice before activating the fixed-price
plan.
For simiplicty's sake, we can avoid trying to do cache invalidation in
the variety of events that can cause the seat count to change - since
having an up to 1 day delay between users being added and the upload
limit going up is quite reasonable.
Fixes#28621
Till now, this was actually a flat 50GB despite what the /plans/ page
says and was adjusted flexibly when somebody asked for a higher limit.
This actually implements the advertised formula, but changing it to
5GB/user since that's a more reasonable limit.
Keeps the 50GB limit for sponsored Standard Free organizations and also
places it as the floor for the quota for paid orgs, to not lower this
for tiny orgs with less than 5 users.
This might not be the most meaningful change of phrasing, but .is_paid()
sounds like it's a check for whether the customer has already paid their
invoice. is_a_paid_plan() reflects better the meaning that it's whether
it's a plan of a "paid" type.
Most importantly, fixes a bug where a realm with a custom
.upload_quota_gb value (set by changing it in the database via e.g.
manage.py shell) would end up having it lowered while upgrading their
plan via the do_change_realm_plan_type function, which used to just set
it to the value implied by the new plan without caring about whether
that isn't lower than the original limit.
The new approach is cleaner since we don't do db queries by
upload_quota_gb so it's nicer to just generate these dynamically, making
changes to our limit-per-plan rules much easier - skipping the need for
migrations.
The `tidy` parameter is buggy, and ignores all ordering
metaparameters. This is fixed in Puppet 7[^1], but it's helpful to
resolve it now. Specifically, this fixes bugs with tidy running too
early, and deleting the old version of a package before its new
version is installed or symlinked, leaving a race condition if
anything tries to run the binary in this window.
This is mostly not a problem for Supervisor-managed processes, since
the binary is already running, and can continue to run if it is tidied
out from under the running process. For stand-alone tools like wal-g,
which are run frequently by PostgreSQL, this may cause issues if
PostgreSQL tries to call them during a puppet run.
Remove all complicated uses of tidy, and replace them with an `exec`
which does the equivalent. We also generate `file` resources for
binaries, making them easier (and clearer) to specify as dependencies.
[^1]: https://puppet.atlassian.net/browse/PUP-10688
Because loading boto3 is so slow, this saves a significant amount of
time (0.3s or so) in process startup on servers which are not using
the S3 file storage backend.
In commit 32d5b4fe3e, the channel and channels operators were
documented in the API docs. This bumps the API version to match
those changes.
Also, cleans up one Changes note from the same commit noted above.
When users needed to add a new task in the todo widget, they would type
it in the field at the top, but the task would be appended to the list,
showing up at the very bottom, which can seem unintuitive.
Now the `add-task-bar` is at the bottom of the list, so that when a new
task is added, it'll appear right where it was typed. The task field
would then shift lower.
Uptil now, users could add tasks to a todo widget only after creating
it through the `/todo` command in the compose box.
Users can now add an initial list of tasks using the `/todo` command,
with each task on a new line in the compose box, where the 1st `:`
would separate a task from its (optional) description. Example:
`/todo\nTask1:description1\nTask2 without description`.
Fixes part of #20213.
Users can now name task lists by providing the task list title in the
`/todo` command on the same line. Example: `/todo School Work`. If no
title is provided by the user, "Task list" (which is also the
placeholder) is used as default.
The author of a task list can later edit / update the task list title
in the todo widget, just like the question in the poll widget.
Fixes part of #20213.
Uptil now, if the user did not specify a poll question, but included
options when using the `/poll` command, the 1st option wrongly
became the poll question.
Now on leaving the poll question space blank, the next line will not be
confused for it, and the poll widget will have an input for adding a
question.
The schema now ensures extra data is validated according to the widget
type. This also makes it easier to modify the extra data for any widget
or add new widget types in the future.
Shift-clicking a link on most browsers should open the link in a new window,
not navigate inside the current tab, just like ctrl-click opens in a new tab.
This fixes the bug where on hovering / selecting an emoji picker search
results emoji, invariably its canonical name would show / be used, which
is not always the same as the matching alias. For example, searching for
"swe" would show the "sweeping" emoji, but its canonical name is "broom"
which can be a confusing search result for "swe".
Documents the addition of "channel" and "channels" as operators for
search/narrow filters, noting that they are aliases of the "stream"
and "streams" operators.
This change does not yet deprecate the "stream" and "streams"
operators.
Part of stream to channel rename project.
Update the Filter class to use "channels" as the canoncial operator
for public stream searches and web-public message fetch narrows,
but keep using "streams" for user-facing text and URLs.
When searching, "channels" does not create any suggestions until
a colon is added and then it shows suggestions with "streams". And
when a search string with channel as an operator is entered, then
it is replaced by "streams" as well.
Part of stream to channel rename project.
Update the Filter class to use "channel" as the canoncial operator
for streams, but keep using "stream" for user-facing text and URLs.
When searching, "channel" does not create any suggestions until
a colon is added and then it shows suggestions with "stream". And
when a search string with channel as an operator is entered, then
it is replaced by "stream" as well.
Part of stream to channel rename project.
Updates filter_with_new_params so that the terms used to update the
current narrow filter are the canonical version of the terms. This
assures that the adjustments we're making are correct.
Updates filter.test.js to use constants when checking the message
type ("stream" or "private") as this eliminates overlap with the
search/narrow operator "stream" and is useful prep for eventually
updating the frontend code to use "direct" as the message type
for direct mesages.
Part of stream to channel rename project.
Adds backend support for "channels" operator.
This will deprecate/replace the "streams" operator eventually, but
we will keep support of the operator for backwards compatibility
for a while.
Part of renaming stream to channel project.
Adds backend support for "channel" operator.
This will deprecate/replace the "stream" operator eventually, but
we will keep support of the operator for backwards compatibility
for a while.
Part of renaming stream to channel project.
Prep commit for deprecating the "stream" search operator and
replacing it with the "channel" operator, and for depreacting the
"streams" operator and replacing it with the "channels" operator.
roanster007
Alex Vandiver
Prakhar Pratyush
Shubham Padia
Mateusz Mandera
Tim Abbott
Lauryn Menard
N-Shar-ma
Olivier FAURE
Varun Singh
Aman Agrawal
Karl Stolley
Alya Abbott