chore(release): publish v2.0.0-rc.1 #3324
Conversation
petermetz
requested review from
RafaelAPB,
takeutak,
izuru0,
jagpreetsinghsasan,
VRamakrishna,
sandeepnRES and
outSH
as code owners
There was a problem hiding this comment.
Hi @petermetz Please resolve my above comment,
and also did you run the last two steps (from release management) regarding updating version in weaver?
To me it seems tools/go-gen-checksum.sh script wasn't run.
It updates the go.mod and go.sum so that published go packages will refer to the latest versions of cacti dependencies instead of previous versions.
@sandeepnRES I ran that script and also the other one [1]. I tried running them again just to make sure I did not forget and they didn't make any changes that I could commit so I'm pretty sure I ran them. My only other guess is that the search and replace I did was wrong. I'll update manually the changes you are asking just so that we make it work and push it over the line and then we can figure out later what went wrong with the scripts or with my search and replace. Also, please feel free to edit these yourself too, the branch should be writable by all of the maintainers. (just make sure to take notes on what you changed so that we can adjust the scripts/release management docs accordingly) [1] ./tools/weaver-update-version.sh 2.0.0-rc.1 .
./tools/go-gen-checksum.sh 2.0.0-rc.1 . |
petermetz
force-pushed
the
release-v2.0.0-rc.1
branch
from
2bf811c
to
c343a68
Compare
That is strange, because without go.mod and go.sum updates, published go modules would break. |
|
Hi @petermetz I ran the same command for script what you shared, and it did change all the For now I've pushed the changes. |
sandeepnRES
force-pushed
the
release-v2.0.0-rc.1
branch
2 times, most recently
from
0fe0f80
to
30d10e8
Compare
Signed-off-by: Peter Somogyvari <peter.somogyvari@accenture.com> Signed-off-by: Sandeep Nishad <sandeep.nishad1@ibm.com>
sandeepnRES
force-pushed
the
release-v2.0.0-rc.1
branch
from
30d10e8
to
9a4e038
Compare
|
Hi @petermetz Two CI tests are failing because of expired certs in fabric testnet of weaver. My PR #3321 is already opened to fix this. |
@sandeepnRES I'm on Linux (Ubuntu 22.04 LTS) indeed so that is most likely the problem because I know that some utilities like ## GO and Docker
VERSION_FILES=("weaver/common/protos-go
weaver/core/network/fabric-interop-cc/libs/utils"
"weaver/core/network/fabric-interop-cc/libs/assetexchange"
"weaver/core/network/fabric-interop-cc/interfaces/asset-mgmt"
"weaver/core/network/fabric-interop-cc/contracts/interop"
"weaver/sdks/fabric/go-sdk"
"weaver/core/drivers/corda-driver"
"weaver/core/drivers/fabric-driver"
"weaver/core/identity-management/iin-agent"
"weaver/core/relay") |
@sandeepnRES Yeah I'm just gonna go ahead and ignore those for now because of what I described above. If this release works (and therefore the process itself works in general) then we can issue another release ( |
Important I know that there are quite a few PRs pending right now, but please do not ask to have those included in this release because that would force me to redo the entire release process from scratch. We can have subsequent v2.0.0-rc.2, v2.0.0-rc.3, v2.0.0-rc.4, etc.. releases as much as we like, but this one is particularly important to be issued ASAP because the currently "latest" version we have out on npm have dependency declaration issues that are urgently needing a fix.
2.0.0-rc.1 (2024-06-14)
Bug Fixes
Features
Performance Improvements
BREAKING CHANGES
ledger versions will no longer have an option for Fabric v1.x
This means that in the core-api package the LedgerType enum has changes
which means that code that depends on that enum value will need to be
updated.
Fabric v1.x has had unmaintained dependencies associated with it such as
the native grpc package that stopped receiving security updates years ago
and therefore it's dangerous to have around.
There are also some issues with Fabric v1.x that make the AIO image flaky
which also makes the relevant tests flaky due to which we couldn't run
the v1.x Fabric tests on the CI for a while now anyway.
In order to reduce the CI resource usage and our own maintenance burden
I suggest that we get rid of the Fabric v1.x support meaning that we can
eliminate the AIO image build and some code complexity from the test ledger
code as well.
In addition some old fixtures can be removed that the tests were using.
Overall a net-positive as deleting code without losing functionality (that
we care about) is always a plus.
Signed-off-by: Peter Somogyvari peter.somogyvari@accenture.com
Pull Request Requirements
upstream/mainbranch and squashed into single commit to help maintainers review it more efficient and to avoid spaghetti git commit graphs that obfuscate which commit did exactly what change, when and, why.-sflag when usinggit commitcommand. You may refer to this link for more information.Character Limit
A Must Read for Beginners
For rebasing and squashing, here's a must read guide for beginners.