Aidan Harding
My feedback
29 results found
-
1 vote0 comments · Help us improve Gearset » Integrations and connections (Jira, source control, DX etc.) · Admin →Aidan Harding shared this idea ·
-
2 votes1 comment · Help us improve Gearset » Integrations and connections (Jira, source control, DX etc.) · Admin →
An error occurred while saving the comment Aidan Harding shared this idea · -
3 votes0 comments · Help us improve Gearset » Integrations and connections (Jira, source control, DX etc.) · Admin →Aidan Harding shared this idea ·
-
16 votes5 comments · Help us improve Gearset » Integrations and connections (Jira, source control, DX etc.) · Admin →
An error occurred while saving the comment Aidan Harding commentedWe frequently have to add fields to managed objects. Not being able to deploy those fields via CI is a great pain and another hole in the idea that git is the source of truth.
Aidan Harding supported this idea · -
1 voteAidan Harding shared this idea ·
-
2 votesAidan Harding shared this idea ·
-
41 votesAidan Harding supported this idea ·
-
21 votes2 comments · Help us improve Gearset » Integrations and connections (Jira, source control, DX etc.) · Admin →Aidan Harding supported this idea ·
-
29 votesAidan Harding supported this idea ·
-
1 voteAidan Harding shared this idea ·
-
2 votesAidan Harding shared this idea ·
-
1 voteAidan Harding shared this idea ·
-
4 votes
An error occurred while saving the comment Aidan Harding commentedThis would also be useful for org-monitoring.
My no-namespace unlocked packages are part of the code I want to monitor and run PMD on, but the only way I can currently achieve that is to include all packages in the metadata filter. This means I have a baseline amount of static analysis issues which are stuck inside managed packages and so it's harder to notice problems in code that we actually control.
Aidan Harding supported this idea · -
2 votesAidan Harding shared this idea ·
-
18 votes
An error occurred while saving the comment Aidan Harding commentedWe're like the option to disable triggers for the data deployment, then restore them afterwards, too.
Obviously, we never write non-bulkified triggers, but sometimes inherit systems that do. And some trigger may have side-effect (email, callouts) that you don't want to happen when just migrating test data.
The selective approach would be best for this as some triggers may be necessary to keep data integrity, even when inserting test data.
Aidan Harding supported this idea · -
17 votesAidan Harding supported this idea ·
-
2 votesAidan Harding shared this idea ·
-
10 votesAidan Harding supported this idea ·
-
16 votesAidan Harding shared this idea ·
-
5 votesAidan Harding shared this idea ·
I just came across another reason to do this... If there's a structural problem with the metadata (in my case, I'd mistakenly created a VF component in the pages directory), Gearset fails the validation before it's even told GitHub that the check is running. So, from GitHub, there's no sign that the check has failed.