Thread: Example is too wide
Could someone pick a different example for the pg_stats values in UG section 11.2, especially one that is significantly narrower than 396 columns? If possible, format verbatim environments to be at most 80 characters wide so they look nicely in print output. It's possible to scale the fonts down automatically so they fit, but that obviously only works within reason. -- Peter Eisentraut peter_e@gmx.net http://funkturm.homeip.net/~peter
Peter Eisentraut <peter_e@gmx.net> writes:
> Could someone pick a different example for the pg_stats values in UG
> section 11.2, especially one that is significantly narrower than 396
> columns?
> If possible, format verbatim environments to be at most 80 characters wide
> so they look nicely in print output.
It's going to be difficult to produce any 80-col-wide examples of any
of the interesting columns in pg_stats. How do you feel about
artificially line-breaking the printout?
regards, tom lane
Tom Lane writes: > It's going to be difficult to produce any 80-col-wide examples of any > of the interesting columns in pg_stats. How do you feel about > artificially line-breaking the printout? I tried the example query on the tenk1 table and got a pretty good output. Not only is it only 120 columns, which makes it fit really well into the HTML browser window (mine anyway), it also contains more diverse rows which can serve the example, e.g., output for unique columns, and a few interesting n_distinct values. I'm tempted to use that one instead. -- Peter Eisentraut peter_e@gmx.net
Peter Eisentraut <peter_e@gmx.net> writes:
> Tom Lane writes:
>> It's going to be difficult to produce any 80-col-wide examples of any
>> of the interesting columns in pg_stats. How do you feel about
>> artificially line-breaking the printout?
> I tried the example query on the tenk1 table and got a pretty good output.
> Not only is it only 120 columns, which makes it fit really well into the
> HTML browser window (mine anyway), it also contains more diverse rows
> which can serve the example, e.g., output for unique columns, and a few
> interesting n_distinct values. I'm tempted to use that one instead.
Well, the thing is that tenk1 is completely artificial data, so the
stats it produces are a tad on the meaningless/uninteresting side.
AFAICT the "road" table is the only realistic sample data we have in
the distribution.
However, I won't argue the point too hard. If the tenk1 example works
from a formatting point of view, go with it.
regards, tom lane