Author Image

CEO & Co-founder of Visivo

Visivo v2.0.1 and v2.0.2: Nested Layouts and a Browser Onboarding Wizard

The first 2.0 patches: nested dashboard layouts, a new visivo init that opens a browser onboarding wizard, in-card scrolling for wide tables, plus Snowflake env-var config and quoted-identifier fixes.

Visivo 2.0 nested dashboard layouts and browser onboarding wizard

The first 2.0 patches are here. v2.0.2 introduces nested dashboard layouts, a new visivo init that opens a browser onboarding wizard, and in-card scrolling for wide tables. The earlier v2.0.1 aligned the docs and error messages to the 2.0 Insights terminology and added Snowflake environment-variable configuration plus a file-API fix. Together they are the steady, practical follow-through on the 2.0 launch.

A major version is never the end of the story, it is the start of one. After 2.0 redefined Visivo around Insights and the semantic layer, the next job is to smooth the path that real users walk: a friendlier first run, more expressive layouts, tables that behave, and connection config that matches how teams actually deploy. v2.0.1 and v2.0.2 do exactly that.

The first 2.0 patches

This recap clusters two releases, because they belong together as the immediate post-launch follow-through.

  • v2.0.1 was the alignment-and-config patch: docs and error messages brought in line with 2.0 Insights terminology, Snowflake configuration via environment variables, and a file-API fallback fix.
  • v2.0.2 is the feature patch and the one this week belongs to: nested dashboard layouts (v1), a new visivo init that opens a browser onboarding wizard, in-card scrolling for wide tables, and preserved case for Snowflake quoted identifiers.

The headline features live in v2.0.2. v2.0.1 is the prior patch that cleared the runway for them. Here is what each delivers.

Docs and errors aligned to 2.0 Insights terms (v2.0.1)

The first thing v2.0.1 did was unglamorous and important: it brought the documentation and the error messages into line with the 2.0 vocabulary.

A major version that renames the core concepts has a hidden cost. The product speaks the new language, but the docs and the errors may still mutter the old one. When the Insights and semantic-layer model became the way you build charts in 2.0, every doc page and every error string needed to reflect that, so that what you read in the docs matches what you write in your project, and an error tells you about the concepts you are actually using.

v2.0.1 closed that gap. The docs and the error messages now use 2.0 Insights terminology consistently. This is the kind of fix that pays off invisibly: a new user reading the docs and hitting an error sees one coherent vocabulary, not a confusing mix of old and new. Coherent language is a feature, especially right after a major version.

Snowflake env-var config and a file-API fix (v2.0.1)

v2.0.1 also improved how Visivo connects to Snowflake, by supporting environment-variable interpolation in more of the connection config.

You could already keep credentials out of your YAML. v2.0.1 extended that to the account, warehouse, role, and timezone fields, so each of those can now be supplied via an environment variable using the standard ${env.X} reference. In practice that looks like:

sources:
  - name: warehouse
    type: snowflake
    account: ${env.SNOWFLAKE_ACCOUNT}
    warehouse: ${env.SNOWFLAKE_WAREHOUSE}
    role: ${env.SNOWFLAKE_ROLE}
    timezone: ${env.SNOWFLAKE_TIMEZONE}

This matters because connection details legitimately differ across environments. Your local machine, CI, staging, and production each point at a different account, warehouse, or role. Hard-coding those values forces you to maintain divergent config files, which is exactly the brittleness that BI-as-code exists to kill. With environment-variable config, one committed project definition works everywhere, and the per-environment specifics live in the environment, not in the repo.

v2.0.1 also shipped a file-API fallback fix, tightening how Visivo reads files in the cases where the primary path was not available. Quiet, but it is the sort of robustness fix that keeps the local loop reliable.

Nested dashboard layouts, v1 (v2.0.2)

The marquee feature of v2.0.2 is nested dashboard layouts, version one. You can now place rows inside items, which lets you compose dashboards with genuine structure instead of a single flat grid.

Flat layouts are fine until they are not. The moment you want a tall chart beside a stack of small KPI cards, or a main visualization with a cluster of supporting numbers tucked alongside it, a flat grid fights you. Nested layouts solve this directly: an item can contain its own rows, so you can build composed regions within a dashboard. A KPI column next to a chart, a header band over a detail grid, related panels grouped into a single visual unit.

This is v1 of the capability, which means the foundation is in place and more will build on it. But even the first version meaningfully expands what a dashboard can express. It is the difference between arranging tiles and actually designing a page, and it directly serves the goal of building BI for the analysts who use it rather than forcing every layout into the same rigid template.

A browser onboarding wizard with visivo init (v2.0.2)

v2.0.2 also reworks the very first thing a new user does. The visivo init command now opens a browser-based onboarding wizard.

First runs are where tools win or lose people. A bare command-line scaffold that drops a config file and leaves you to read the docs is a high bar for someone just trying Visivo for the first time. The new visivo init lowers that bar dramatically: run it, and a guided wizard opens in your browser to walk you through getting set up, connecting a source, and producing your first project, with a visual flow instead of a blank file and a manual.

This is a real change in the on-ramp. It pairs a code-first product with a welcoming, guided first experience, so the power of BI-as-code does not come at the cost of an intimidating start. If you are trying Visivo for the first time, the new wizard is the front door. Our get started guide walks through the same flow.

Wide tables scroll in-card and Snowflake case is preserved (v2.0.2)

Two further refinements round out v2.0.2.

Wide tables now scroll in-card. A table with many columns no longer pushes past the boundaries of its card. Instead, the table scrolls horizontally within the card it lives in, so a wide dataset stays contained and the surrounding dashboard layout holds. This continues the table-behavior work from the late 1.x line: tables should respect the space they are given, not break the composition around them.

Snowflake quoted-identifier case is preserved. Snowflake treats unquoted identifiers as uppercase but preserves the exact case of quoted ones. If a column is defined as "MixedCase", that casing is significant, and silently folding it produces "column not found" errors that are maddening to debug. v2.0.2 preserves the case of quoted Snowflake identifiers, so your queries resolve against the names that actually exist. This is the kind of correctness fix that you only notice when it stops biting you.

What it unlocks

v2.0.1 and v2.0.2 are a confident first chapter of the 2.0 line. The pattern is exactly what you want after a major release: align the language (v2.0.1 docs and errors), shore up the connections (Snowflake env vars, quoted-identifier case), and then push the experience forward (nested layouts and a real onboarding wizard).

Nested layouts unlock dashboards with structure, and they are explicitly v1, so there is a foundation to extend. The browser onboarding wizard lowers the barrier to a first project, which matters most precisely when a major version is bringing new people in. And the steady stream of correctness fixes is the trust-building work that turns a fresh major version into a dependable one.

If you are on 2.0, update for the nested layouts and the new visivo init flow. If you have not tried 2.0 yet, run visivo init and let the wizard show you around, or follow the get started guide. You can also browse the examples gallery to see what composed dashboards look like.

Previously in Visivo

In the previous installment, Visivo v1.0.82: Metric and Dimension Publishing, End to End, we closed out the 1.x line by completing the semantic-layer publishing loop right before 2.0. These first 2.0 patches pick up where that left off, building on the new foundation with nested layouts, a guided onboarding wizard, and a round of correctness fixes.

Install command copied