On 11.02.23 20:24, Andres Freund wrote: > > I think on a green field it'd be clearly better to do something like the > above. What does give me pause is that it seems quite likely to break > existing queries, and to a lesser degree, might break applications relying on > inferred column names > > Can anybody think of a good way out of that? It's not like that problem is > going to go away at some point...
I think we should just do it and not care about what breaks. There has never been any guarantee about these.
I'm going to toss a -1 into the ring but if this does go through a strong request that it be disabled via a GUC. The ugliness of that option is why we shouldn't do this.
Defacto reality is still a reality we are on the hook for.
I too find the legacy design choice to be annoying but not so much that changing it seems like a good idea.