History log of /conch/t/ (Results 1 - 25 of 533)
Revision (<<< Hide revision tags) (Show revision tags >>>)Date Author Comments
(<<< Hide modified files)
(Show modified files >>>)
Revision tags: v2.36.0
9f0f491f01-Oct-2019 Karen Etheridge

whitespace

6e038e2211-Oct-2019 Karen Etheridge

separate the test objects by authenticated user

65f6af4911-Oct-2019 Karen Etheridge

use test device objects a bit more consistently

b899144b10-Oct-2019 Karen Etheridge

also add debug logging for build access checks

51cdf41709-Oct-2019 Karen Etheridge

fix: permission checks for unlocated devices in a build

It used to be that we could only do a specific role check for devices that had
a location, but now devices can be location-less an

fix: permission checks for unlocated devices in a build

It used to be that we could only do a specific role check for devices that had
a location, but now devices can be location-less and still allow a user to
have a specific role, via user -> build -> device.

show more ...

13467b4e07-Oct-2019 Karen Etheridge

allow caching of GET /device/:id with If-None-Match and ETag

closes #904.

be79897a04-Oct-2019 Karen Etheridge

allow for caching of GET /schema/* data by checking If-Modified-Since

closes #896.

7ce4085d04-Oct-2019 Karen Etheridge

include a Last-Modified header with GET /version

4f12258307-Oct-2019 Karen Etheridge

merge inconsistent email format checks, in create_user command too

JSON::Validator's email check is a little more loose than Email::Valid, so now
we use both in all places where an email

merge inconsistent email format checks, in create_user command too

JSON::Validator's email check is a little more loose than Email::Valid, so now
we use both in all places where an email address can come into the system

show more ...

23c04ffe02-Oct-2019 Karen Etheridge

database constraints for build columns

This will ensure that build columns remain consistent with each other to
reflect our application logic.

d98f432c04-Oct-2019 Karen Etheridge

make sure documentation rendering issues are visible

a55bd49f04-Oct-2019 Karen Etheridge

use just one variable to store the hardware product

92ee509c02-Oct-2019 Karen Etheridge

avoid repetition

b26125f804-Oct-2019 Karen Etheridge

fix: log the same response that we are returning

This discrepancy occurred in places where $c->status(..) or $c->rendered(..)
was called more than once in a dispatch cycle. The after_dis

fix: log the same response that we are returning

This discrepancy occurred in places where $c->status(..) or $c->rendered(..)
was called more than once in a dispatch cycle. The after_dispatch hook is only
called once, for the first call, but what is returned in the response reflects
the last call.

- refactor the txn_wrapper helper to not actually render a response, but just
use the return value to signal the outcome; the inclusion of the first line of
the exception in the response has moved to the status helper

- also added a check in the after_render hook to ensure this does not happen again.

show more ...

Revision tags: v2.35.0, v3.0.0-a6
298f845824-Sep-2019 Karen Etheridge

look up validation plan by hardware_product, rather than hacky hardcoded heuristic

- add validation_plan_id to hardware_product, using historical validation data to
populate existing row

look up validation plan by hardware_product, rather than hacky hardcoded heuristic

- add validation_plan_id to hardware_product, using historical validation data to
populate existing rows
- this means that we can no longer deactivate validation plans and create new
ones with the same name, and expect them to be used in place of their former
version. we need to either update plans in place, or update references to the
plan (i.e. hardware_product.validation_plan_id).

closes #806.

show more ...

0511e75e02-Oct-2019 Karen Etheridge

include the validation version when logging an issue with it

6e74d4a002-Oct-2019 Karen Etheridge

test for hardware product mismatch in validate_report path too

7cfb3ccf02-Oct-2019 Karen Etheridge

move these tests up to avoid concocting test data twice

e5f1301402-Oct-2019 Karen Etheridge

shorten these calls by using unserialized data

b19c758926-Sep-2019 Karen Etheridge

devices are no longer created upon receipt of the first report

They must first be created, via either
POST /rack/:rack_id/assignment or POST /build/:build_id_or_name/device.

b5d9b81125-Sep-2019 Karen Etheridge

new endpoint: GET /device/:id_or_serial/sku

closes #893.

37e6be6a25-Sep-2019 Karen Etheridge

new endpoint: POST /build/:id_or_name/device

7a7cf2d124-Sep-2019 Karen Etheridge

make hardware_product.sku mandatory

This will be used as the canonical determination of a device's intended
hardware when processing reports.

for #807, #893.

d78645c324-Sep-2019 Karen Etheridge

some more error checking in hardware_product creation, update

7272305527-Sep-2019 Karen Etheridge

return all users with organization data

that is, in GET /organization and GET /organization/:id_or_name.
GET /organization/:id_or_name/users is now redundant and has been removed.

return all users with organization data

that is, in GET /organization and GET /organization/:id_or_name.
GET /organization/:id_or_name/users is now redundant and has been removed.

closes #898.

show more ...

12345678910>>...22