This updates recommonmark so that relative links to .md files will
correctly be converted to links to html pages by Sphinx/ReadTheDocs.
This was repoprted in https://github.com/rtfd/recommonmark/issues/89
This will allow us to in a future commit use relative links to .md
files so all of the hyperlinks work both on ReadTheDocs and on GitHub.
Earlier the behavior was to raise an exception thereby stopping the
whole sync. Now we log an error message and skip the field. Also
fixes the `query_ldap` command to report missing fields without
error.
Fixes: #11780.
This fixes an issue where invalid emoji name prevents following
emojis from rendering.
This reverts the code change in
8842349629, while still passing the
tests added in that commit (it seems the original commit had
misdiagnosed an ordering bug and thus introduced this issue).
Fixes: #11770.
Most elements, apart from chevrons, now have explicit
font sizes.
In some cases I chose integer values that were close
to the calculated values you would get with all the
ratio calculations.
And then I tweaked how the hashtag/lock icons get
aligned.
The alignment for those icons if off in this version; it'll be fixed
in an upcoming commit.
The fonts there used to be 14.2px due to an arcane
calculation. Now we explicity set a value.
If you expand Private Messages, the font inside now
is explicitly set to 13px. It used to be 12.8px due
to a complex calculation.
The previous commits made the "arrow" class no longer
necessary for these left sidebar elements.
The "arrow" class was always a bit dangerous, as bootstrap
has a class by the same name for a different concept.
We weren't really using consistent styling for "arrow"
across all our chevrons, so even without bootstrap in
play, it was just creating complexity.
This is a fairly big commit, but at the end
it simplifies a lot of things.
It's difficult to fix highly coupled code in
incremental steps because, well, it's highly
coupled code.
The main thing this does is give each type of
chevron in the left sidebar its own class
* all-messages-arrow (NEW)
* stream-sidebar-arrow
* topic-sidebar-arrow
Before this change, the "All messages" chevron
was using stream-sidebar-arrow, which was a
strange name for something that's not actually
in the stream sidebar. Obviously this was
cargo culted.
There was not much JS to change here--we just
fix the click handler for "All messsages".
And then there's a one-line change to the template,
and the rest is re-organizing the CSS.
Using a more specific class avoids confusion related
to the .arrow class, which is not only a popover concept,
but also a Zulip concept in the left sidebar.
The way we build chevrons is super messy and highly
coupled. This comment reflects an audit I did on the
code in its current state.
Subsequent commits will make things a bit easier to
understand.
The topic-sidebar-arrow chevron never shows
here--it was just cargo-culted code from
the topic lists.
It's also a landmine--we don't want to
accidentally build a topic-related sidebar
menu for PMs.
We could arguably just use zero-unreads everywhere,
but we definitely don't want zero-topic-unreads
inside our PM list.
I prefer to just have these two concepts:
zero-pm-unreads
zero-topic-unreads
And it's super easy to share CSS properties for both.
Few folks will be upgrading from versions of Zulip old enough to not
have virtualenv-clone, and those who are won't be able to use it due
to older dependencies having been removed.
The antialiasing decisions we made for the webapp should be constant
over the entire page, not limited to particular subsections or themes.
If we wanted antialiasing, we should do it on the entire page, not
individual random widgets. But it's not clear we actually want to do
it on the entire page. The `-moz-osx-font-smoothing: grayscale`
setting now happens by default in OSX Mojave (40% world market share
right now and growing), so there's no reason to override it. And
without retina displays, generally, subpixel rendering provides better
results than antialiasing (which overrides subpixel rendering).
Thanks to Anders Kaseorg for advice on this issue.
Apparently, while upgrade-zulip-from-git always ensures that zulip
deployment directories are owned by the Zulip user, unpack-zulip (aka
the tarball code path) has them owned by root.
The user ID detection logic in su_to_zulip's helper get_zulip_uid was
intended to support both development environments (where the user ID
might vary) and production environments. For development
environments, the existing code is fine, but given this unpack-zulip
permissions issue, we need to have code to fallback to 'zulip' if the
detection logic detects the "zulip" user has having UID 0.
Vishnu Ks
Samuel Searles-Bryant
Harshit Bansal
overide
Tim Abbott
Anders Kaseorg
Steve Howell