Thread: [doc] modifying unit from characters to bytes
Hi, The manual describes the size of pg_stat_activity.query as below: | By default the query text is truncated at 1024 characters; When considering multibyte characters, it seems more accurate to change the unit from "characters" to "bytes". I also searched other "[0-9] characters" in the manual. I may overlook something, but apparently it seems ok because of their contexts which are limited to ASCII character or other reasons. Regards, -- Atsushi Torikoshi NTT DATA CORPORATION
Attachment
On 2020/07/08 10:54, torikoshia wrote: > Hi, > > The manual describes the size of pg_stat_activity.query > as below: > > | By default the query text is truncated at 1024 characters; > > When considering multibyte characters, it seems more > accurate to change the unit from "characters" to "bytes". Agreed. Barring any objection, I will commit this patch. For record, this change derived from the discussion about other patch [1]. Regards, [1] https://postgr.es/m/cd0e961fd42e5708fdea70f7420bf214@oss.nttdata.com -- Fujii Masao Advanced Computing Technology Center Research and Development Headquarters NTT DATA CORPORATION
> On 8 Jul 2020, at 04:25, Fujii Masao <masao.fujii@oss.nttdata.com> wrote: > > On 2020/07/08 10:54, torikoshia wrote: >> Hi, >> The manual describes the size of pg_stat_activity.query >> as below: >> | By default the query text is truncated at 1024 characters; >> When considering multibyte characters, it seems more >> accurate to change the unit from "characters" to "bytes". > > Agreed. Barring any objection, I will commit this patch. +1 to commit this patch, following the link to track_activity_query_size it's even specified to be bytes there. IIRC the NULL terminator is also included in the 1024 bytes which prevents it from being 1024 characters even for non-multibyte. cheers ./daniel
On 2020/07/08 16:17, Daniel Gustafsson wrote: >> On 8 Jul 2020, at 04:25, Fujii Masao <masao.fujii@oss.nttdata.com> wrote: >> >> On 2020/07/08 10:54, torikoshia wrote: >>> Hi, >>> The manual describes the size of pg_stat_activity.query >>> as below: >>> | By default the query text is truncated at 1024 characters; >>> When considering multibyte characters, it seems more >>> accurate to change the unit from "characters" to "bytes". >> >> Agreed. Barring any objection, I will commit this patch. > > +1 to commit this patch, following the link to track_activity_query_size it's > even specified to be bytes there. IIRC the NULL terminator is also included in > the 1024 bytes which prevents it from being 1024 characters even for > non-multibyte. Yes, so we should document "truncated at 1023 bytes" for accuracy, instead? This might be more confusing for users, though.... Regards, -- Fujii Masao Advanced Computing Technology Center Research and Development Headquarters NTT DATA CORPORATION
> On 8 Jul 2020, at 10:05, Fujii Masao <masao.fujii@oss.nttdata.com> wrote: > > On 2020/07/08 16:17, Daniel Gustafsson wrote: >>> On 8 Jul 2020, at 04:25, Fujii Masao <masao.fujii@oss.nttdata.com> wrote: >>> >>> On 2020/07/08 10:54, torikoshia wrote: >>>> Hi, >>>> The manual describes the size of pg_stat_activity.query >>>> as below: >>>> | By default the query text is truncated at 1024 characters; >>>> When considering multibyte characters, it seems more >>>> accurate to change the unit from "characters" to "bytes". >>> >>> Agreed. Barring any objection, I will commit this patch. >> +1 to commit this patch, following the link to track_activity_query_size it's >> even specified to be bytes there. IIRC the NULL terminator is also included in >> the 1024 bytes which prevents it from being 1024 characters even for >> non-multibyte. > > Yes, so we should document "truncated at 1023 bytes" for accuracy, instead? > This might be more confusing for users, though.... I think that's overcomplicating things, since we do (will) specify bytes and not characters. cheers ./daniel
On 2020/07/08 17:12, Daniel Gustafsson wrote: >> On 8 Jul 2020, at 10:05, Fujii Masao <masao.fujii@oss.nttdata.com> wrote: >> >> On 2020/07/08 16:17, Daniel Gustafsson wrote: >>>> On 8 Jul 2020, at 04:25, Fujii Masao <masao.fujii@oss.nttdata.com> wrote: >>>> >>>> On 2020/07/08 10:54, torikoshia wrote: >>>>> Hi, >>>>> The manual describes the size of pg_stat_activity.query >>>>> as below: >>>>> | By default the query text is truncated at 1024 characters; >>>>> When considering multibyte characters, it seems more >>>>> accurate to change the unit from "characters" to "bytes". >>>> >>>> Agreed. Barring any objection, I will commit this patch. >>> +1 to commit this patch, following the link to track_activity_query_size it's >>> even specified to be bytes there. IIRC the NULL terminator is also included in >>> the 1024 bytes which prevents it from being 1024 characters even for >>> non-multibyte. >> >> Yes, so we should document "truncated at 1023 bytes" for accuracy, instead? >> This might be more confusing for users, though.... > > I think that's overcomplicating things, since we do (will) specify bytes and > not characters. Agreed. So I pushed the proposed patch. Thanks! Regards, -- Fujii Masao Advanced Computing Technology Center Research and Development Headquarters NTT DATA CORPORATION
On 2020-07-09 13:47, Fujii Masao wrote: > On 2020/07/08 17:12, Daniel Gustafsson wrote: >>> On 8 Jul 2020, at 10:05, Fujii Masao <masao.fujii@oss.nttdata.com> >>> wrote: >>> >>> On 2020/07/08 16:17, Daniel Gustafsson wrote: >>>>> On 8 Jul 2020, at 04:25, Fujii Masao <masao.fujii@oss.nttdata.com> >>>>> wrote: >>>>> >>>>> On 2020/07/08 10:54, torikoshia wrote: >>>>>> Hi, >>>>>> The manual describes the size of pg_stat_activity.query >>>>>> as below: >>>>>> | By default the query text is truncated at 1024 characters; >>>>>> When considering multibyte characters, it seems more >>>>>> accurate to change the unit from "characters" to "bytes". >>>>> >>>>> Agreed. Barring any objection, I will commit this patch. >>>> +1 to commit this patch, following the link to >>>> track_activity_query_size it's >>>> even specified to be bytes there. IIRC the NULL terminator is also >>>> included in >>>> the 1024 bytes which prevents it from being 1024 characters even for >>>> non-multibyte. >>> >>> Yes, so we should document "truncated at 1023 bytes" for accuracy, >>> instead? >>> This might be more confusing for users, though.... >> >> I think that's overcomplicating things, since we do (will) specify >> bytes and >> not characters. > > Agreed. So I pushed the proposed patch. Thanks! Thanks for applying! Regards, -- Atsushi Torikoshi NTT DATA CORPORATION