On 10/14/21 1:47 PM, Aleksander Alekseev wrote: > Hi Gilles, > >> Any though and interest in this feature? > > Personally, I wouldn't call this feature particularly useful. `SELECT > *` is intended for people who are working with DBMS directly e.g. via > psql and want to see ALL columns.
I disagree strongly with this. It is really annoying when working interactively with psql on a table that has a PostGIS geometry column, or any other large blobby type column.
I have not looked at the patch, but +1 for the feature.
Cannot be better to redefine some strategies for output for some types.
I can agree so sometimes in some environments proposed features can be nice, but it can be a strong footgun too.
Maybe some strange data can be filtered in psql and it can be better solution. I agree, so usually print long geometry in psql is useless.
Pavel this doesn't concern only output but input too, think about the INSERT or COPY without a column list. We can add such filter in psql but how about other clients? They all have to implement their own filtering method. I think the HIDDEN attribute provide a common and basic way to implement that in all client application.
I like the idea - being able to hide computed columns such as tsvectors from CRUD queries by default seems like it would be very nice for example.
