Tests that network seed always affects dna hash #2102

maackle · 2023-03-16T23:56:52Z

Summary

Fixes #2039

TODO:

CHANGELOG(s) updated with appropriate info
Just before pressing the merge button, ensure new entries to CHANGELOG(s) are still under the UNRELEASED heading

Fixes #2038

maackle · 2023-03-24T19:20:06Z

crates/holochain_types/src/app/app_manifest/app_manifest_v1.rs

+            if !matches!(
+                role.provisioning.clone().unwrap_or_default(),
+                CellProvisioning::CreateClone { .. } | CellProvisioning::UseExisting { .. }
+            ) {


This is the fix

…s-affects-dna-hash

crates/holochain/src/conductor/tests/install_app_bundle.rs

ThetaSinner · 2023-03-28T12:54:49Z

crates/holochain_types/src/combinators.rs

@@ -24,3 +24,8 @@ pub fn second_ref<A, B>(tup: &(A, B)) -> &B {
 pub fn swap2<A, B>(tup: (A, B)) -> (B, A) {
    (tup.1, tup.0)
 }
+
+/// Map an iterator into a vec of a new type
+pub fn mapvec<'a, T: 'a, U>(it: impl Iterator<Item = &'a T>, f: impl FnMut(&'a T) -> U) -> Vec<U> {


Should this be map_vec?

It is what it is :)

ThetaSinner · 2023-03-28T12:57:47Z

crates/holochain/src/conductor/tests/install_app_bundle.rs

+        .install_app_bundle(InstallAppPayload {
+            agent_key: agent.clone(),
+            source: AppBundleSource::Bundle(bundle1),
+            installed_app_id: Some("no-seed".into()),


Perhaps seed-a and seed-b or something binary other than yes/no. This reads to me like "there is no seed" but the test is setting one

In the more comprehensive ignored test, I used "no seed", "seed A", and "seed B". But in this case, it really is about seed vs no seed, and in this case there is really no seed.

Co-authored-by: Jost Schulte <jost.schulte@protonmail.com>

* Take `name` out of DnaDefHash Fixes #2038 * Changelog * Write passing test for reported bug #2039 * Add tests for paths -- still passes * Extremely comprehensive test of network seed and location: still passes! * Even more comprehensive test: still passes! * Small fries, wip * Added more cases; still passes * Better debug * Add regression test and fix the problem * Changelog * Update crates/holochain/src/conductor/tests/install_app_bundle.rs Co-authored-by: Jost Schulte <jost.schulte@protonmail.com> --------- Co-authored-by: Jost Schulte <jost.schulte@protonmail.com>

* Fix clone cell idempotency for enable/disable (#2093) * Fix old tokio dep * Failing test and fixes of clone cell enable/disable idempotency * Changelog * Update comment * Differentate between CellMissing and CellDisabled, and other fixes (#2092) * Fix old tokio dep * Possible fix to intermittent disabled apps * Clear all ephemeral functions at once * Make cell_by_id differentiate between reasons for cell missing * Revert app status check and add detailed comment * Fix test * Clippy * Update comment for ListCellIds * Changelog * Take `name` out of DnaDefHash (#2099) * Take `name` out of DnaDefHash Fixes #2038 * Changelog * Update tests * Update crates/holochain_zome_types/CHANGELOG.md Co-authored-by: Jost Schulte <jost.schulte@protonmail.com> --------- Co-authored-by: Jost Schulte <jost.schulte@protonmail.com> * refactor(zome-call): optimize cap grant verification (#2097) (#2116) * refactor(zome-call): cap grant verification * refactor: simplify cap grant validation flow * update changelog * simplify assigned cap access arm * add PR reference to changelog * add comment to test * Apply suggestions from code review * refactor: use high level db query fn * replace hardcoded cap access variant by method --------- Co-authored-by: Michael Dougherty <maackle.d@gmail.com> * fix(conductor-api): reject duplicate clone cells (#1997) * add failing test for duplicate clone cell * reject duplicate dna hash * add duplicate dna hash error * specify error and test again after disabling clone * update changelog * add test for two agents cloning cells in same conductor * check for duplicate cell id in conductor state * OpType -> FlatOp + some reorg (#1909) * OpType -> FlatOp * Reorg * Reorg * Move flat_op to hdi * Tightening up OpType/FlatOp logic (#1910) * Failing test * Fix test * WIP big refactor * Fix all compiler errors * Add FlatOps for storing cap entries * Fix private entry type logic * Test that sort of works * Test passes * Changelog * Removed needless test cases * Fix some warnings * More specific error * Relax test case * Remove dbg!s and backtraces * Apply suggestions from code review Co-authored-by: Jost Schulte <jost.schulte@protonmail.com> --------- Co-authored-by: Jost Schulte <jost.schulte@protonmail.com> --------- Co-authored-by: Jost Schulte <jost.schulte@protonmail.com> * fix typo * fix conductor tests * AppManifest `version` becomes `installed_hash` (#2157) * Make `version` take only a single optional hash * Rename AppManifest's `version` -> `installed_hash` * Changelog * Look up local DNAs before resolving via Location * Include AppManifest in AppInfo (#2156) * Return manifest with AppInfo and rewrite legacy app install * Clippy * update source 'nixpkgs' on branch 'develop' (#2158) [create-pull-request] automated change Co-authored-by: steveeJ <steveeJ@users.noreply.github.com> * update source 'rust-overlay' on branch 'develop' (#2159) [create-pull-request] automated change Co-authored-by: steveeJ <steveeJ@users.noreply.github.com> * fix(release-automation): consider dependency state of all matched crates (#2160) fix(release-automation): consider dependency state of all matched crates (#2149) * fix(release-automation): consider dependency state of all matched crates the change filter could lead to dependency changes not including the changed dependency in the release, if the selected crate itself didn't have any changed. the blocked filter was removed because it adds unnecessary obfuscation. * chore(release-automation): remove duplicate test * test(relase-automation): cover the fixed case * chore: add FIXME comment for a bug in the dependencies_in_workspace_filter function * test(release-automation): extend dependency bump test to cover the transitive case * refactor(release-automation/crate_selection): clarify transitive change detection * update source 'nixpkgs' on branch 'develop' (#2161) [create-pull-request] automated change Co-authored-by: steveeJ <steveeJ@users.noreply.github.com> * Test that manifest includes most recent modifiers * Investigating BadChecksum * Update from_raw_32 to only allow 32 * Use 32 byte hash for a test where it matters * Fix fixturator * Fix one more test --------- Co-authored-by: Holochain Release Automation <100725712+holochain-release-automation2@users.noreply.github.com> Co-authored-by: steveeJ <steveeJ@users.noreply.github.com> Co-authored-by: Stefan Junker <1181362+steveeJ@users.noreply.github.com> --------- Co-authored-by: Holochain Release Automation <100725712+holochain-release-automation2@users.noreply.github.com> Co-authored-by: steveeJ <steveeJ@users.noreply.github.com> Co-authored-by: Stefan Junker <1181362+steveeJ@users.noreply.github.com> * Add backward compat test and fix existing manifest serde test * Rename to `_version` * Fix visibility * Fix bad backport * Tests that network seed always affects dna hash (#2102) * Take `name` out of DnaDefHash Fixes #2038 * Changelog * Write passing test for reported bug #2039 * Add tests for paths -- still passes * Extremely comprehensive test of network seed and location: still passes! * Even more comprehensive test: still passes! * Small fries, wip * Added more cases; still passes * Better debug * Add regression test and fix the problem * Changelog * Update crates/holochain/src/conductor/tests/install_app_bundle.rs Co-authored-by: Jost Schulte <jost.schulte@protonmail.com> --------- Co-authored-by: Jost Schulte <jost.schulte@protonmail.com> * Fix test * Improve HoloHash::from_raw_32 and use it (#2162) * Update from_raw_32 to only allow 32 * Use 32 byte hash for a test where it matters * Fix fixturator * Fix one more test * Fix test --------- Co-authored-by: Jost Schulte <jost.schulte@protonmail.com> Co-authored-by: Holochain Release Automation <100725712+holochain-release-automation2@users.noreply.github.com> Co-authored-by: steveeJ <steveeJ@users.noreply.github.com> Co-authored-by: Stefan Junker <1181362+steveeJ@users.noreply.github.com>

maackle added 4 commits March 16, 2023 15:17

Take name out of DnaDefHash

6f77171

Fixes #2038

Changelog

5492da0

Write passing test for reported bug #2039

2cb32a5

Add tests for paths -- still passes

a57fa3a

maackle changed the base branch from develop to dna-name-not-a-modifier March 16, 2023 23:57

Base automatically changed from dna-name-not-a-modifier to develop March 18, 2023 00:25

Extremely comprehensive test of network seed and location: still passes!

9d2950f

maackle force-pushed the network-seed-always-affects-dna-hash branch from f4a75c3 to 9d2950f Compare March 21, 2023 23:24

maackle added 6 commits March 21, 2023 17:46

Even more comprehensive test: still passes!

591798b

Small fries, wip

2fde8a0

Added more cases; still passes

cf647bb

Better debug

01f0762

Add regression test and fix the problem

6aa083e

Changelog

4fe532a

maackle commented Mar 24, 2023

View reviewed changes

Merge remote-tracking branch 'origin/develop' into network-seed-alway…

49ad82c

…s-affects-dna-hash

maackle mentioned this pull request Mar 24, 2023

[BUG] Network seed is not changing the DNA hash when source is AppBundleSource::Bundle(AppBundle) #2039

Closed

maackle requested review from jost-s, neonphog, thedavidmeister and ThetaSinner March 27, 2023 17:49

neonphog approved these changes Mar 27, 2023

View reviewed changes

jost-s approved these changes Mar 27, 2023

View reviewed changes

crates/holochain/src/conductor/tests/install_app_bundle.rs Outdated Show resolved Hide resolved

ThetaSinner approved these changes Mar 28, 2023

View reviewed changes

maackle and others added 2 commits March 28, 2023 09:23

Merge branch 'develop' into network-seed-always-affects-dna-hash

ea53bc8

Update crates/holochain/src/conductor/tests/install_app_bundle.rs

7a60f53

Co-authored-by: Jost Schulte <jost.schulte@protonmail.com>

maackle enabled auto-merge (squash) March 28, 2023 16:29

maackle merged commit c50ffc9 into develop Mar 28, 2023
6 checks passed

maackle deleted the network-seed-always-affects-dna-hash branch March 28, 2023 16:34

maackle added the ShouldBackport/0.1 label Apr 4, 2023

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Tests that network seed always affects dna hash #2102

Tests that network seed always affects dna hash #2102

maackle commented Mar 16, 2023 •

edited

maackle Mar 24, 2023

ThetaSinner Mar 28, 2023

maackle Mar 28, 2023

ThetaSinner Mar 28, 2023

maackle Mar 28, 2023

Tests that network seed always affects dna hash #2102

Tests that network seed always affects dna hash #2102

Conversation

maackle commented Mar 16, 2023 • edited

Summary

TODO:

maackle Mar 24, 2023

Choose a reason for hiding this comment

ThetaSinner Mar 28, 2023

Choose a reason for hiding this comment

maackle Mar 28, 2023

Choose a reason for hiding this comment

ThetaSinner Mar 28, 2023

Choose a reason for hiding this comment

maackle Mar 28, 2023

Choose a reason for hiding this comment

maackle commented Mar 16, 2023 •

edited