Обсуждение: filenames in pg_basebackup
Hi, I notice that pg_basebackup has lots of messages where file names are not in quotes. Is this okay? There's also this one which is probably against our conventions: #: pg_receivexlog.c:181 #, c-format msgid "%s: segment file '%s' is incorrect size %d, skipping\n" I also wonder about "is" vs. "has" in the above message. There's also a --help line that's too long. Apparently, this needs a thorough revision ... -- Álvaro Herrera <alvherre@alvh.no-ip.org>
Alvaro Herrera <alvherre@alvh.no-ip.org> writes:
> I notice that pg_basebackup has lots of messages where file names are
> not in quotes. Is this okay?
They should be in double quotes, per our message style guidelines.
> msgid "%s: segment file '%s' is incorrect size %d, skipping\n"
> I also wonder about "is" vs. "has" in the above message.
Yeah, "has" would be more correct.
> Apparently, this needs a thorough revision ...
Evidently. Peter tends to see to that sort of thing while he works on
translations, but I'm sure he wouldn't mind some help.
regards, tom lane
Excerpts from Tom Lane's message of mié jul 25 16:25:36 -0400 2012: > Alvaro Herrera <alvherre@alvh.no-ip.org> writes: > > Apparently, this needs a thorough revision ... > > Evidently. Peter tends to see to that sort of thing while he works on > translations, but I'm sure he wouldn't mind some help. Here's an attempt at that. One thing I'm not clear about is the "WAL file" vs "transaction log file" terminology. We use both in various error messages. Do we want to consistently use one? It seems to me that we're using the very verbose "transaction log" phrase just to avoid exposing users to the "WAL" acronym, but that's probably a lost cause. We also have the issue of calling those files "files" or "segments". I understand that internally we don't want to confuse them, but I don't see that the distinction makes any sense to users. -- Álvaro Herrera <alvherre@commandprompt.com> The PostgreSQL Company - Command Prompt, Inc. PostgreSQL Replication, Consulting, Custom Development, 24x7 support
Вложения
Alvaro Herrera <alvherre@commandprompt.com> writes:
> One thing I'm not clear about is the "WAL file" vs "transaction log
> file" terminology. We use both in various error messages. Do we want
> to consistently use one? It seems to me that we're using the very
> verbose "transaction log" phrase just to avoid exposing users to the
> "WAL" acronym, but that's probably a lost cause. We also have the issue
> of calling those files "files" or "segments". I understand that
> internally we don't want to confuse them, but I don't see that the
> distinction makes any sense to users.
Yeah, we did talk about that a bit when Heikki was working on the
64-bit-XLogPointer changes. AFAIR, there was general consensus that
it'd be better to standardize on one terminology, but little agreement
on which wording to use :-(. I'd say review the thread and pick one.
He who does the work gets to make the decision.
regards, tom lane
Excerpts from Alvaro Herrera's message of mié jul 25 18:23:46 -0400 2012: > Excerpts from Tom Lane's message of mié jul 25 16:25:36 -0400 2012: > > Alvaro Herrera <alvherre@alvh.no-ip.org> writes: > > > > Apparently, this needs a thorough revision ... > > > > Evidently. Peter tends to see to that sort of thing while he works on > > translations, but I'm sure he wouldn't mind some help. > > Here's an attempt at that. One more thing. There are new switches for the interval of status messages in pg_basebackup and pg_receivexlog. They are --statusint=INTERVAL. I find this rather ugly, and we tend to avoid abbreviations. I propose --status-interval=INTERVAL instead. -- Álvaro Herrera <alvherre@commandprompt.com> The PostgreSQL Company - Command Prompt, Inc. PostgreSQL Replication, Consulting, Custom Development, 24x7 support
On Thu, Jul 26, 2012 at 12:58 PM, Alvaro Herrera <alvherre@commandprompt.com> wrote: > > Excerpts from Alvaro Herrera's message of mié jul 25 18:23:46 -0400 2012: >> Excerpts from Tom Lane's message of mié jul 25 16:25:36 -0400 2012: >> > Alvaro Herrera <alvherre@alvh.no-ip.org> writes: >> >> > > Apparently, this needs a thorough revision ... >> > >> > Evidently. Peter tends to see to that sort of thing while he works on >> > translations, but I'm sure he wouldn't mind some help. >> >> Here's an attempt at that. > > One more thing. There are new switches for the interval of status > messages in pg_basebackup and pg_receivexlog. They are > --statusint=INTERVAL. I find this rather ugly, and we tend to avoid > abbreviations. I propose --status-interval=INTERVAL instead. +1 And --noloop should be --no-loop in pg_receivexlog? Regards, -- Fujii Masao
On Thu, Jul 26, 2012 at 10:15 AM, Fujii Masao <masao.fujii@gmail.com> wrote: > On Thu, Jul 26, 2012 at 12:58 PM, Alvaro Herrera > <alvherre@commandprompt.com> wrote: >> >> Excerpts from Alvaro Herrera's message of mié jul 25 18:23:46 -0400 2012: >>> Excerpts from Tom Lane's message of mié jul 25 16:25:36 -0400 2012: >>> > Alvaro Herrera <alvherre@alvh.no-ip.org> writes: >>> >>> > > Apparently, this needs a thorough revision ... >>> > >>> > Evidently. Peter tends to see to that sort of thing while he works on >>> > translations, but I'm sure he wouldn't mind some help. >>> >>> Here's an attempt at that. >> >> One more thing. There are new switches for the interval of status >> messages in pg_basebackup and pg_receivexlog. They are >> --statusint=INTERVAL. I find this rather ugly, and we tend to avoid >> abbreviations. I propose --status-interval=INTERVAL instead. > > +1 > > And --noloop should be --no-loop in pg_receivexlog? Or maybe --no-retry. -- Robert Haas EnterpriseDB: http://www.enterprisedb.com The Enterprise PostgreSQL Company