Thread: char array overhead
I'm angling toward using a very wide char(1) array. Is the one-byte overhead for char(n<126) applied to each element or to the array?
On Mar 31, 2014, at 8:08 AM, Rob Sargent <robjsargent@gmail.com> wrote: > I'm angling toward using a very wide char(1) array. Is the one-byte overhead for char(n<126) applied to each element orto the array? Each element, it's a variable length type. There's probably a better way of storing your data, but if you end up really needing a one-byte long character type, thereis "char" (with the quotes). Cheers, Steve
On 03/31/2014 09:48 AM, Steve Atkins wrote:
Thank you! First for the confirmation on the overhead (how disappointing) and secondly for the "char". That had not crossed my radar.On Mar 31, 2014, at 8:08 AM, Rob Sargent <robjsargent@gmail.com> wrote:I'm angling toward using a very wide char(1) array. Is the one-byte overhead for char(n<126) applied to each element or to the array?Each element, it's a variable length type. There's probably a better way of storing your data, but if you end up really needing a one-byte long character type, there is "char" (with the quotes). Cheers, Steve
Jsyk, I'm toying with a rather large number of small valued datapoints per sample. I'm tying text and smallint as well.
Thanks again,
rjs
Rob Sargent wrote: > Jsyk, I'm toying with a rather large number of small valued > datapoints per sample. I'm tying text and smallint as well. You could try "char"[] ... -- Álvaro Herrera http://www.2ndQuadrant.com/ PostgreSQL Development, 24x7 Support, Training & Services