Textual Any Renderer: Remove msg header screen

[amaury1093]

Summary

The proposal is to remove the line in red in the current Any value renderer.

{ "text": "This transaction has 1 Message" },
{ "text": "Message (1/1): Object: /cosmos.bank.v1beta1.MsgSend" },
{ "text": "Message (1/1): /cosmos.bank.v1beta1.MsgSend" },
- { "text": "MsgSend object" },
{ "text": "From address: cosmos1ulav3hsenupswqfkw2y3sup5kgtqwnvqa8eyhs" },
{ "text": "To address: cosmos1abc" },
... // indentation omitted

Problem Definition

Context:

• Initially, when writing the initial draft version of the spec, we omitted the MsgSend line, as it's redundant
• Then, when writing the Any value renderer, we added this line, for the sake of consistency feat: value renderer for google.protobuf.Any #13671 (see below)
• And we're revisiting in this issue this decision, ref refactor(textual): Update Spec to have shorter Any header screen #14117 (comment)

Proposal

Pros of removing the line:

• shorter rendered output (but needs a small "hack" in Any value renderer)

Pros of keeping the line:

• Consistency, when thinking about value renderers.

For example, for nested fields, we use:

My field: [[MsgSend object
> From address: cosmos1ulav3hsenupswqfkw2y3sup5kgtqwnvqa8eyhs]]

Which has a clear spec: <field name>: <value-rendered msg>, where the part between brackets [[ ]] denotes the message value renderer output (the brackets themselves are not shown on screen, just here to clarify the purpose).

The Any spec is:

<Type URL>
<value-renderer msg>

which gives the initial output with the message header screen.

So it's a tradeoff between shorter screens vs consistency.

[amaury1093]

In Jim's PR, I approved to go with the additional header screen, because it felt more correct code-wise.

However, I'm changing my mind, and I think Textual is about human readability before anything, and shorter is better in this case. Morever, I suspect that the above consistency argument only stands because of our chosen implementation.

[JimLarson]

I'm softening on my resistance to this - the extra header of the Any really does make the message header unnecessary.