Thread: pgsql: Mention "pg_ctl stop" in pgindent README instructions.
Mention "pg_ctl stop" in pgindent README instructions. Branch ------ master Details ------- http://git.postgresql.org/pg/commitdiff/bb8f0c4b4863d697264cc6f295f50627b15a6608 Modified Files -------------- src/tools/pgindent/README | 1 + 1 files changed, 1 insertions(+), 0 deletions(-)
On 06/09/2011 08:51 PM, Bruce Momjian wrote: > Mention "pg_ctl stop" in pgindent README instructions. > This change make no sense at all. Why would you be running 'pg_ctl stop' when there is nothing to suggest that there's a postgres running. And even if there is there's not the slightest reason for stopping it unless it's in the same location you're installing to. Certainly having this line in the README with no comment about when you might want it is going to be highly confusing at best. I think you need to rethink this. cheers andrew
Andrew Dunstan wrote: > > > > > On 06/09/2011 08:51 PM, Bruce Momjian wrote: > > Mention "pg_ctl stop" in pgindent README instructions. > > > > This change make no sense at all. Why would you be running 'pg_ctl stop' > when there is nothing to suggest that there's a postgres running. And > even if there is there's not the slightest reason for stopping it unless > it's in the same location you're installing to. Certainly having this > line in the README with no comment about when you might want it is going > to be highly confusing at best. I think you need to rethink this. Well, I have a Postgres running, and it the install goes into the same directory, so my regression tests failed, so I documented the command to shut down the postmaster. -- Bruce Momjian <bruce@momjian.us> http://momjian.us EnterpriseDB http://enterprisedb.com + It's impossible for everything to be true. +
On 06/09/2011 11:28 PM, Bruce Momjian wrote: > Andrew Dunstan wrote: >> >> >> >> On 06/09/2011 08:51 PM, Bruce Momjian wrote: >>> Mention "pg_ctl stop" in pgindent README instructions. >>> >> This change make no sense at all. Why would you be running 'pg_ctl stop' >> when there is nothing to suggest that there's a postgres running. And >> even if there is there's not the slightest reason for stopping it unless >> it's in the same location you're installing to. Certainly having this >> line in the README with no comment about when you might want it is going >> to be highly confusing at best. I think you need to rethink this. > Well, I have a Postgres running, and it the install goes into the same > directory, so my regression tests failed, so I documented the command to > shut down the postmaster. Then the answer is "don't do that." Surely it's far saner to run configure --prefix=/path/to/temploc and then go from there. Disturbing a running postmaster for this purpose is crazy. BTW, for years I've used a gadget to set up every piece of development work with its own install location and port so that this sort of problem doesn't happen. It's fairly trivial but I'll gladly share it with you. I rather imagined every serious postgres developer would have such a thing. cheers andrew
Andrew Dunstan wrote:
>
>
> On 06/09/2011 11:28 PM, Bruce Momjian wrote:
> > Andrew Dunstan wrote:
> >>
> >>
> >>
> >> On 06/09/2011 08:51 PM, Bruce Momjian wrote:
> >>> Mention "pg_ctl stop" in pgindent README instructions.
> >>>
> >> This change make no sense at all. Why would you be running 'pg_ctl stop'
> >> when there is nothing to suggest that there's a postgres running. And
> >> even if there is there's not the slightest reason for stopping it unless
> >> it's in the same location you're installing to. Certainly having this
> >> line in the README with no comment about when you might want it is going
> >> to be highly confusing at best. I think you need to rethink this.
> > Well, I have a Postgres running, and it the install goes into the same
> > directory, so my regression tests failed, so I documented the command to
> > shut down the postmaster.
>
> Then the answer is "don't do that." Surely it's far saner to run
> configure --prefix=/path/to/temploc and then go from there. Disturbing a
> running postmaster for this purpose is crazy. BTW, for years I've used a
> gadget to set up every piece of development work with its own install
> location and port so that this sort of problem doesn't happen. It's
> fairly trivial but I'll gladly share it with you. I rather imagined
> every serious postgres developer would have such a thing.
Well, I use that README most and I wanted to be reminded to shut down
the postmaster. I don't use Postgres for any production work so
installing in my default location makes sense for me.
We mention in our docs shutting down the postmaster before installing:
http://www.postgresql.org/docs/9.0/static/install-procedure.html
Installing the Files
Note: If you are upgrading an existing system and are going to
install the new files over the old ones, be sure to back up your data
and shut down the old server before proceeding, as explained in Section
15.4 above.
--
Bruce Momjian <bruce@momjian.us> http://momjian.us
EnterpriseDB http://enterprisedb.com
+ It's impossible for everything to be true. +
Bruce Momjian wrote: > > Then the answer is "don't do that." Surely it's far saner to run > > configure --prefix=/path/to/temploc and then go from there. Disturbing a > > running postmaster for this purpose is crazy. BTW, for years I've used a > > gadget to set up every piece of development work with its own install > > location and port so that this sort of problem doesn't happen. It's > > fairly trivial but I'll gladly share it with you. I rather imagined > > every serious postgres developer would have such a thing. > > Well, I use that README most and I wanted to be reminded to shut down > the postmaster. I don't use Postgres for any production work so > installing in my default location makes sense for me. > > We mention in our docs shutting down the postmaster before installing: > > http://www.postgresql.org/docs/9.0/static/install-procedure.html > > Installing the Files > > Note: If you are upgrading an existing system and are going to > install the new files over the old ones, be sure to back up your data > and shut down the old server before proceeding, as explained in Section > 15.4 above. I am not sure why you are so concerned about this one line in a developer README, but if you would prefer, I can add text to remind people to shut down the postmaster before the install, rather than mentioning 'pg_ctl stop'. I need something to remind me so I don't have this problem in the future. -- Bruce Momjian <bruce@momjian.us> http://momjian.us EnterpriseDB http://enterprisedb.com + It's impossible for everything to be true. +