Re: [PATCH] Incremental sort (was: PoC: Partial sort)
| От | James Coleman |
|---|---|
| Тема | Re: [PATCH] Incremental sort (was: PoC: Partial sort) |
| Дата | |
| Msg-id | CAAaqYe-14hsP437g6yoQswfizfexRBR4GM+QduKdjyPRfcUczw@mail.gmail.com обсуждение исходный текст |
| Ответ на | Re: [PATCH] Incremental sort (was: PoC: Partial sort) (Tom Lane <tgl@sss.pgh.pa.us>) |
| Список | pgsql-hackers |
On Fri, Mar 13, 2020 at 4:22 PM Tom Lane <tgl@sss.pgh.pa.us> wrote: > > Alvaro Herrera <alvherre@2ndquadrant.com> writes: > > Also, I wonder if it would be better to modify our policies so that we > > update typedefs.list more frequently. Some people include additions > > with their commits, but it's far from SOP. > > Perhaps. My own workflow includes pulling down a fresh typedefs.list > from the buildfarm (which is trivial to automate) and then adding any > typedefs invented by the patch I'm working on. The latter part of it > makes it hard to see how the in-tree list would be very helpful; and > if we started expecting patches to include typedef updates, I'm afraid > we'd get lots of patch collisions in that file. > > I don't have any big objection to updating the in-tree list more often, > but personally I wouldn't use it, unless we can find a better workflow. How does the buildfarm automate generating the typedefs list? Would it be relatively easy to incorporate that into a tool that someone could use locally with pgindent? Thanks, James
В списке pgsql-hackers по дате отправления: