| Commit message (Collapse) | Author | Age | Lines |
|
|
|
| |
To match shift of styles in bb07940.
|
| |
|
| |
|
| |
|
|\ |
|
| | |
|
| | |
|
|/
|
|
| |
Fixes #929. Also move admin hint CSS to base CSS, not just layout.
|
|
|
|
|
| |
Also apply to UK council cobrands.
Fixes #926, more food for #344.
|
|
|
|
| |
For #344 and #935.
|
|
|
|
|
|
| |
Fixes #357.
Also consolidate almost-identical fill_in_details.html template
(for #344).
|
|
|
|
|
| |
Only show "select..." if more than one body, and only mention greyed out
lines if there are any. Fixes #869.
|
|
|
|
|
|
| |
Easier if the default shared styles stick with the sprite for
everything, then you can just change that sprite rather than having to
override CSS as well.
|
|
|
|
| |
This look works well on the default styling also.
|
| |
|
|
|
|
| |
Move UK specific text to fixmystreet.com template.
|
|\ |
|
| |
| |
| |
| |
| | |
This crashes at least one mobile browser; not sure why, but we can do
without it.
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Fixes #932, which is caused by HTML5 validation firing for hidden
elements, by setting #mapForm to novalidate by default, and reverting
this in fixmystreet.js.
Note that the 'novalidate' attribute, which is honoured by the browsers'
HTML5 validation, isn't to be confused with the class (class="validate")
which is picked up by fixmystreet.js, and therefore not problematic for
the no-js case.
|
| |
| |
| |
| | |
For starters, it actually appears on focus.
|
|/
|
|
|
|
| |
Error handling and changes to the base template were added in f0621930,
so an error would cause the form to reshow in the cobrand, but not
display the errors.
|
| |
|
| |
|
|
|
|
|
| |
After the changes in b3c664b, these two lines need to be called with the
cron-wrapper.
|
|
|
|
|
| |
With reference to #922, to see if the more 'businessy' phrase makes any
difference.
|
|
|
|
|
|
| |
This means they access the database the same way all the other scripts
do, preventing any issues due to different setups. Also tidy up state
grouping in graph, add Norway start date.
|
|\ |
|
| |
| |
| |
| |
| |
| |
| | |
This puts back any missing snapshot modules removed in
1708bc30af7176f95f0ec5718289d3a633daf32b, as they are required for
installation on squeeze, which misses or does not have the required
version of these.
|
| |
| |
| |
| |
| | |
Also, s///r isn't present in 5.10, and FixMyStreet should still install
on squeeze.
|
| |
| |
| |
| |
| |
| | |
Try and save other people the potential pain of installing SOAP::Lite;
only UK FixMyStreet and Zurich geocoder use it, so set it as an optional
feature in the cpanfile instead.
|
| |
| |
| |
| |
| |
| |
| |
| | |
Switching to bundler saves confusion about gem paths when running
sass/compass to compile the CSS.
Trusty can use precisely the same packages as precise, with a few tweaks
to generalise the contents.
|
|/
|
|
|
|
|
|
|
|
|
|
| |
The #map_box div wasn't clickable, so we had to reset its z-index from the
broken -1 to 0. IE7 doesn't like setting z-index to 0 in javascript (as it's a
falsy value) so instead we set the z-index of parent and sibling elements (body
and site-header) to 1 to make sure they are visible.
position:relative in base.scss is to make sure that the header is now
shown (similar to @6296cb33)
See https://github.com/mysociety/FixMyStreet-Commercial/issues/639
|
|
|
|
| |
At Noel's request.
|
|
|
|
| |
fixmystreet/com -> fixmystreet.com
|
|
|
|
| |
In case it contains an ampersand or somesuch.
|
|
|
|
| |
If you know to change the URL, anyway, for now.
|
|
|
|
|
|
|
|
| |
This changes the default flow to show no pins and a message about
reporting, with the second tab displaying existing problems (and showing
map pins). Those existing problems are the ones that would currently be
on "problems on the map" followed by those currently on "problems
nearby".
|
|
|
|
| |
This gave no conclusive results, which was interesting.
|
| |
|
|
|
|
|
|
|
| |
see
https://github.com/mysociety/FixMyStreet-Commercial/issues/596
(in separate commit with inline style, to hopefully make easy to revert)
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
works around
https://github.com/mysociety/FixMyStreet-Commercial/issues/614
(This is a chrome bug, but as it doesn't happen with standard
FMS, we investigated further with @stevenday and @zarino.
For reasons I don't understand, the issue is fixed by:
- removing negative z-index from map
- this then makes the site-header not display, so we set
position:relative -- not having done this before might be
the reason for the previous negative index? )
|
|
|
|
|
|
| |
as WCC is (during beta) mixed email (manually updated) and Open311,
override the behaviour to delete contacts not returned by service list,
to ignore ones that look like email addresses.
|
|
|
|
|
|
|
| |
- Tweaks required to get things working in WCC environment
- update Open311 to onsite changes
- updated Open311 parameters after WCC's Bentley and County Highways workshop.
- ... including ce_cpr_id
|
|
|
|
|
| |
Uses LWP::Protocol::PSGI to avoid having to spawn external processes
in order to test these end to end.
|
|
|
|
|
| |
If https://github.com/mysociety/fixmystreet/pull/792 is accepted
then this should be rewritten in terms of that.
|
|
|
|
|
|
|
| |
This parameter was added in 3eb93cea4acd5f9b76efdc0910466363a0b9b445
and is superfluous (as jurisdiction*_id* is the correct parameter,
and is already sent.) That commit mentions only api_key so understand
this wrong parameter is an artefact.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
::Integration::Warwick subclasses ::Integration::Exor
refactor request_class and new_request
Exor service
Warwickshire updates retrieval, with datetimes
stubbing out of Oracle constants, for local testing
We also edit FMS's core PopulateServiceList routine to hide system
fields from FMS:
Bromley/Warwickshire send metadata in their services/FOO.xml
advising that you can pass, e.g. attributes[easting].
FMS by default shows all of these to the user to fill in, however
we don't *want* the user to supply these, rather they are added
by the cobrand.
Bromley had an exception for this (keyed by $body->areas->id).
We write this more generally for Warwickshire too, keying instead
by $body->name (as this is far less likely to be overridden for
installs using global or custom Mapit's)
|
|
|
|
|
|
|
|
| |
See also MooX::ConfigFromFile, but that's underdocumented and
seems overengineered -- may be worth implementing if requirements
become more complex however. (See also Config::Any, which is
well worth doing in future, using YAML only reflects current
usage in FMS though.)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* Get Service Request Updates
This requires a new object ::Service::Request::Update, which of course
is not part of standard spec. So, in order to make the core not too
contaminated by :
* the endpoint should instantiate ::Service::Request::mySociety objects
which know about updates
* have added a learn_additional_types callback from Schema to Endpoint,
so that core doesn't need to know about /open311/service_request_update
* (but ::Spark knows about the exception for updates... meh, but is 1-line)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Subsystems include
* ::Spark encoding conventions for xml/json
* ::Schema using Rx to validate form of inputs and outputs,
including validation for, e.g., dates and CSV as part of Open311
Handles following paths:
* Open311 attributes for Service Definition
http://wiki.open311.org/GeoReport_v2#GET_Service_Definition
* POST service request
* GET Service Requests
* GET Service Request
Objects:
* ::Service
* ::Service::Request
|