building apache-superset-core fails due to incompatibility of sqlglot

[PickleRiiiick]

Bug description

I am trying to implement an upgrade after getting this bug:
#37977

The docker containers are running on a linux debian 12.

I am following these steps:
https://superset.apache.org/admin-docs/installation/upgrading-superset#docker-compose

When executing docker compose up I get the following error:

superset-init-1       |       Built apache-superset-core @ file:///app/superset-core
superset-init-1       | Prepared 1 package in 899ms
superset-init-1       | Uninstalled 1 package in 2ms
superset-init-1       | warning: Failed to hardlink files; falling back to full copy. This may lead to degraded performance.
superset-init-1       |          If the cache and target directories are on different filesystems, hardlinking may not be supported.
superset-init-1       |          If this is intentional, set `export UV_LINK_MODE=copy` or use `--link-mode=copy` to suppress this warning.
superset-init-1       | Installed 1 package in 1ms
superset-init-1       |  - apache-superset-core==0.0.1rc3 (from file:///app/superset-core)
superset-init-1       |  + apache-superset-core==0.0.1rc4 (from file:///app/superset-core)
superset-init-1       | Reinstalling the app in editable mode
superset-init-1       |   × No solution found when resolving dependencies:
superset-init-1       |   ╰─▶ Because only apache-superset-core==0.0.1rc4 is available and
superset-init-1       |       apache-superset-core==0.0.1rc4 depends on sqlglot>=28.10.0,<29, we
superset-init-1       |       can conclude that all versions of apache-superset-core depend on
superset-init-1       |       sqlglot>=28.10.0,<29.
superset-init-1       |       And because apache-superset==0.0.0.dev0 depends on
superset-init-1       |       apache-superset-core and sqlglot>=27.15.2,<28, we can conclude that
superset-init-1       |       apache-superset==0.0.0.dev0 cannot be used.
superset-init-1       |       And because only apache-superset==0.0.0.dev0 is available and you
superset-init-1       |       require apache-superset, we can conclude that your requirements are
superset-init-1       |       unsatisfiable.
superset-init-1 exited with code 1

Screenshots/recordings

No response

Superset version

master / latest-dev

Python version

3.9

Node version

16

Browser

Firefox

Additional context

No response

Checklist

• I have searched Superset docs and Slack and didn't find a solution to my problem.
• I have searched the GitHub issue tracker and didn't find a similar bug report.
• I have checked Superset's logs for errors and if I found a relevant Python stacktrace, I included it here as text in the "additional context" section.

[sadpandajoe]

@dosu do we have instructions for this? Are we missing a package install?

[dosubot]

This looks like a dependency version misalignment issue with the new apache-superset-core package that was introduced for backend extensions. The sqlglot upgrade to 28.10.0 appears to have been applied to apache-superset-core but not synchronized with the main apache-superset package constraints.

I don't see specific documented instructions for this scenario - it appears to be a version constraint synchronization issue between the two packages that needs to be fixed in the repository.

As a workaround, try:

1. docker compose down -v
2. git pull origin master (ensure you have the latest code)
3. docker compose build --no-cache
4. docker compose up

If that doesn't work, this may need a fix to align the sqlglot version constraints between the main pyproject.toml and superset-core/pyproject.toml. Both should specify the same sqlglot version range.

@maintainers: This looks like a version constraint sync issue - should both packages specify sqlglot>=28.10.0,<29 or sqlglot>=27.15.2,<28?

To reply, just mention @dosu.

---

^{How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other}

   

[fowlerp397-tech]

Appreciate the detailed report, very helpful

…

On Tue, Mar 17, 2026, 3:01 AM Mayank Aggarwal ***@***.***> wrote: *Mayankaggarwal8055* left a comment (apache/superset#38310) <#38310 (comment)> Hi, I was able to reproduce this issue locally and spent some time digging into it. It looks like the problem comes from how columns_by_name is constructed in get_sqla_query(). Right now, verbose_name can end up overriding column_name in the mapping, which leads to incorrect column resolution when filters are applied. I’ve made a small change to ensure that column_name always stays as the primary key, and verbose_name is only used as an alias when there’s no conflict. This seems to fix the filtering behavior. I’ve opened a PR with the fix: #38689 <#38689> — Reply to this email directly, view it on GitHub <#38310 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/B72HASMKI733JTTQ5XIKOBD4REA35AVCNFSM6AAAAACWDHZGHWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHM2DANZTGA2TSNJZGA> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>