Flatbuffers: Release 1.12 tracking

[aardappel]

I was thinking of doing a new release, anyone know of PRs that should land before this happens, or recent PRs that introduced problems that should really be fixed?

Could do this as early as this Thu (12th) if nothing comes up.

This issue to track progress towards 1.12. Please post any blockers here.

@rw @vglavnyy @mustiikhalil @mugisoba @dbaileychess @paulovap @dnfield @somombo @AustinSchuh @maxburke @dpacbach @lu-wang-g

[dpacbach]

I think I have one or two open issues and/or PRs, but I'd consider them low priority/non-blocking, nothing that needs attention for this next release.

[mustiikhalil]

I only have the Object-API issue (WIP), and it's simply an add on to swift, which I think should be in the next release (1.13).

[rw]

I merged one small Rust PR just now. Otherwise, Rust, Python, and Go don't have any major PRs waiting to be merged, so I'm fine releasing a version. 👍

[vglavnyy]

I did small checking, nothing critical.
There are some nitpicking that can be easily fixed before release.

idl_parser.cpp 1332: assert(key) instead of FLATBUFFERS_ASSERT(key)

This is mistyping.

V509 [CWE-703] The 'new' operator is used in the noexcept 'EquipmentUnion' function. It should be located inside the try..catch block, as it could potentially generate an exception. monster_generated.h 700

The copy constructor of EquipmentUnion marked as noexcept but it can throw an exception if new or WeaponT throw. Generated WeaponT does not have noexcept qualifier.
This can be dangerous for applications that depend on the generated code.

Low priority:

V519 [CWE-563] The 'get_nested_method_name' variable is assigned values twice successively. Perhaps this is a mistake. Check lines: 850, 851. idl_gen_csharp.cpp 851

Probably mistyping.

V2005 C-style explicit type casting is utilized. Consider using: static_cast/const_cast/reinterpret_cast. idl_gen_python.cpp 97
V2005 C-style explicit type casting is utilized. Consider using: static_cast/const_cast/reinterpret_cast. idl_gen_python.cpp 1003

10003: one_instance[0] = char(tolower(one_instance[0]));

The idl_gec_cpp has ToUpper, idl_gen_swift has LowerCase. Need to move these functions to util.cpp and replace all calls to ::upper and ::lower. This can be done after the release.

[paulovap]

@aardappel only one I am waiting for is https://github.com/google/flatbuffers/pull/5735

[aardappel]

@vglavnyy can you make a PR with the minimal fixes you think should go in before the release?
@paulovap let me look at that again on Thu, though if its just a speedup can also go in after release.

Generally if you're waiting for my response and haven't seen any in over a week, feel free to ping me, it's likely it escaped my attention :)

[AustinSchuh]

Nothing from over here in bazel land.

[vglavnyy]

@aardappel

1. PR with assert: #5799.
2. For noexcept I created an issue #5800. This code exists for about 3 years, I think we can fix it after release.
3. The third issue #5801 connect with old-style-cast has low priority, I will make PR after release.

[maxburke]

@rw / @aardappel What about flexbuffers for Rust? #5669

[tsingson]

for Go, please add a little go.mod file to support go module.

go.mod file is only 2 line , should be stored in root path of github.com/google/flatbuffers/ repo :

module github.com/google/flatbuffers

go 1.13

thanks.

[somombo]

for Go, please add a little go.mod file to support go module.

go.mod file is only 2 line , should be stored in root path of github.com/google/flatbuffers/ repo :

module github.com/google/flatbuffers

go 1.13

thanks.

if all that's done is adding just the go.mod file rather than a variety of things that were done in PR https://github.com/google/flatbuffers/pull/5739, then the repo's tests (and some other things) will break

[aardappel]

@maxburke we'd have to ask @CasperN .. but I think this can also land after release if its not ready.

[aardappel]

Ok, looks like we're mostly ready for a release.. I will start working down this list shortly https://github.com/google/flatbuffers/wiki/FlatBuffers-release-process and update here if there's any issues :)

[aardappel]

@mustiikhalil see Swift related FIXME's in https://github.com/google/flatbuffers/wiki/FlatBuffers-release-process.. these should be updated from generate_code.sh..

[aardappel]

Ok, release is live: https://github.com/google/flatbuffers/releases/tag/v1.12.0

Packages for package managers of various languages not updated yet, that happens next

[aardappel]

Packages for most languages uploaded, except for Java and C#

@rw, can you update Rust (version has already been bumped) and @mustiikhalil Swift?

[rw]

@aardappel Just a minor bump on the Rust port, right?

[aardappel]

@rw I already bumped it 6.1 -> 6.2, so you can just release it.

[aardappel]

Announce tweet: https://twitter.com/wvo/status/1238589009602830336

[aardappel]

Java gRPC also released (which had to wait on the Java Maven release to finish).

[ryandesign]

This issue to track progress towards 1.12.

1.12 was released months ago. Should this issue be closed?