On 2024/03/28 11:27:11 +0100, Stefan Sperling <stsp@stsp.name> wrote:
> During a conversation with Theo at g2k19 I somehow became convinced
> that Got should avoid calling unveil before launching an editor.
> 
> The code I had written initially didn't make this assumption. So I
> already knew at the time that running an editor would work just fine
> with unveil already set up. But I was cautioned about future changes
> to how unveil would operate.
> Either the design of unveil wasn't quite finalized yet at that point,
> or I misunderstood what Theo was trying to get across (no surprise,
> that happens quite often ;)
> 
> Either way, this removes all the lies about unveil from got/got.c.
> The issue persists in cvg/cvg.c but that can be fixed later (cvg has
> no regression tests yet).
> 
> Tests are passing and some manual testing suggests interactive use
> of the editor still works as expected.

Yeah, there's no way this would work as-is.  gotd would be broken since
we unveil the helpers.

> ok?

ok op@