Issue Description

What are you doing?

I have a model that contained precomputed fields including a column that is an array of int (tags_ids). Here is its definition :

'use strict';

// This sub table is useful to save computation times for search
module.exports = (sequelize, DataTypes) => {
    let Exercise_Metrics = sequelize.define('Exercise_Metrics', {
        // 

When adding a new data store database in Metabase it tries to connect first with SSL (if the driver supports SSL), and then without, in that order. If either connection succeeds, the database details are accepted as valid. Yes, you can add driver specific JDBC options to use SSL, but there is no good feed

Issue type:

[x] question
[ ] bug report
[ ] feature request
[x] documentation issue

Database system/driver:

[ ] cordova
[x] mongodb
[ ] mssql
[ ] mysql / mariadb
[ ] oracle
[ ] postgres
[ ] cockroachdb
[ ] sqlite
[ ] sqljs
[ ] react-native
[ ] expo

TypeORM version:

[x] latest
[ ] @next
[ ] 0.x.x (or put your version here)

When a data source gets deleted, it's queries are still accessible (detailed explanation can be found in #4336) but they error out due to data_source being None.

We should provide a helpful error message there (perhaps by raising QueryDetachedFromDataSourceError).

Environment

• PostgreSQL version: All
• PostgREST+ version: 6.0.2 (713b214)

Current state

Currently, Comment on Table

COMMENT ON SCHEMA todos IS
  'Todos object description';

generate a summary for all operations on this resource

 /todos:
    get:
      tags:
        - todos
      summary: Todos object description
      parameters:

But sum

Now you can view common image files inside DBeaver when editing table data. You can open/save the binary data from/to filesystem. What would be very useful is the ability to open a CLOB value in external editor. For example storing pdf files into database would become convenient because the user could open the value in a pdf reader to check the contents of the value. Opening in external editor wou

Environment

Knex version: 0.20.10
Database + version: Postgres 12

Bug

1. Explain what kind of behaviour you are getting and how you think it should do

knex.raw("?", [
  knex.insert({ id: "b", fields: { a: undefined } }).into("test"),
])

2. Error message

Throw NullPointException when INSERT column and values mismatch.
It is better to throw suitable SQLException for JDBC-driver and proxy.

The tasks are:

• Throw suitable SQLException for JDBC-driver.
• Report suitable error packet with MySQL proxy.
• Report suitable error packet with PostgreSQL proxy.

• Add option to enable analyze on queries only
• Add option to export as JSON rather than text (for import into https://tatiyants.com/pev/#/plans/new)
• Add button (or similar) to easily copy the SQL query or explain results

pgcli depends on humanize which is https://github.com/jmoiron/humanize.

The current version has a slightly odd behavior of rounding down unexpectedly:

In [11]: humanize.naturaltime(datetime.timedelta(seconds=120))                                                                         
Out[11]: '2 minutes ago'

In [12]: humanize.naturaltime(datetime.timedelta(seconds=119.9))      

We are suddenly running out of background workers and were told on Slack to increase timescaledb.max_background_workers from 8 to 16.

The documentation makes it seem like the timescaledb-tune tool will automatically set the key to a correct value. However the tool [always sets the value to 8](https://github.com/timescale/timesca

The [API documentation for pg.Client][] does not document client.release. The method is documented elsewhere in the [overview of pooling][] and the [API documentation for pg.Pool][] (see brianc/node-postgres#1302 (comment)), but, due to my inattentive blindness, I had missed that the examples used client.release, not client.end. It would be good if

If I were to deploy cube.js using AWS serverless architecture, is athena required?

The docs https://cube.dev/docs/deployment#serverless do not mention whether Athena is optional or required. But when reading it, I assume it is required because there are athena keys in the serverless.yml config. I'm evaluating the idea of using Postgres RDS as the sole datasource for cubejs.

Currently (on 0.9.4) the content of bytea columns is shown as a UTF-8 string in the table view.
It would be nice if these columns could be shown using hex encoding instead.

The encoding can be done in postgres using:

SELECT encode(my_checksum, 'hex') FROM my_table;

for example，i want the Name field search with name like 、> 、<、 is null、 ... ,not only name = ?

type Dapp struct {
	UUID          string `xorm:"varchar(36) 'uuid'"`
	Name          string
	Subtitle      string
	Category      string
	Permission    string
	Synopsis      string
	Score         float64
	Logo          string
	Banner        string
	Snapshot      string
	Video     

In the tutorial events there is this example:

const User = bookshelf.model('User', {
  tableName: 'users',

  initialize() {
    this.on('saving', (model) => {
      return User.forge({email: model.get('email')}).fetch().then((user) => {
        if (user) 
            throw new Error('That email address already exists')
    

See https://forum.newsblur.com/t/how-to-set-default-full-article-retrieval-in-android-app/7111/4. The most recent comment says that it wasn't discoverable. Let's make it discoverable by adding a preference to turn it off (defaulted to on of course).

Given that we support CURRENT_TIMESTAMP, the lack of CURRENT_DATE feels like an omission. While you can hack it with date(now), I keep finding myself wishing there was a direct way to reference this. Given that now means CURRENT_TIMESTAMP, I think there's precedent for today meaning CURRENT_DATE. This is supported by all backends we support (and is in the ISO standard I believe).

if I understood it corretly from README.MD, we can install like this:

$ git clone https://github.com/donnemartin/dev-setup.git && cd dev-setup
$ ./.dots bootstrap osxprep brew osx

and later when we need datastores, we run

$ cd ~/dev-setup
$ ./.dots datastores

I understand that bootstrap copies the dot files to the home directory, such as .bash_profile and .exports.
but

1、我最近在使用Enterprisedb 10.4，然后参照下面的文章：https://www.enterprisedb.com/docs/en/10.0/Ora_Reference_Guide_v10/Database_Compatibility_for_Oracle_Developers_Reference_Guide.1.035.html

在pg admin里面输入了下面这些代码
CREATE DATABASE LINK chicago
CONNECT TO sue IDENTIFIED BY 'mypassword'
USING oci '//127.0.0.1/acctg';

但是却报了下面这个错，很疑惑，语法都是按照文档上来的，为什么会有这个问题呢
ERROR: syntax error at or near "TO"
SQL state: 4

What went wrong?

I'm getting deprecation warnings with openSSL encryption.

[2019/09/05 08:38:52][info] Using Encryptor::OpenSSL to encrypt the archive.
[2019/09/05 08:40:22][warn] Pipeline STDERR Messages:
[2019/09/05 08:40:22][warn] (Note: may be interleaved if multiple commands returned error messages)
[2019/09/05 08:40:22][warn]
[2019/09/05 08:40:22][warn] *** WARNING : depre

The steps described in our README does not work for me: https://github.com/citusdata/citus/blob/master/CONTRIBUTING.md#following-our-coding-conventions

The main problem is that after cloning the tools repo we need to run make install from top level instead of uncrustify because many of the variables are defined at the top level.

In the documentation for connect() function, parameter server_settings has the following description:

An optional dict of server runtime parameters. Refer to PostgreSQL documentation for a list of supported options.

The link in the description returns erro

https://emotion.sh/

After this change in Material UI: mui-org/material-ui#6115

^{Now with extra diff highlights, praise Github}

When using the template_directory setting to put your Twig files in. There are some small inconsistencies:

1. Template values are stored with a beginning forward slash / when template_directory is set, i.e. /index.twig vs index.twig. Not a big problem