DEV Community: Nick F

Postgesql - create a user for a database

Nick F — Wed, 22 Nov 2023 22:23:31 +0000

How to setup a new user and give that user permission over a database? I found this a hard question to answer, so I'm writing down my soltuion.

This worked for me on Postresql 15.

You need a superuser account, to create the database.

create database <database name>;
create user <user name> with encrypted password '<password>';
alter database <database name> owner to <user name>;

At that point, the user you created can login and create/edit/delete tables as you would expect within that one database.

Using "git diff" and "git apply" to share small changes.

Nick F — Fri, 09 Sep 2022 17:45:57 +0000

Sometimes in a Pull Request Review you want to tell someone about a little change to make. You make your change locally, maybe a few lines different, and note that it works locally. Your git diff output will show your change. How can you share this little snippit? Just copy the output of git diff and sent that to your team mate. Here is a small change I've made locally:

diff --git a/conf/local.rb b/conf/local.rb
index 57672691..28fd024c 100644
--- a/conf/local.rb
+++ b/conf/local.rb
@@ -1,6 +1,6 @@
-@whitePagesCacheTtl = "60s"
-
-@executionEnvironment = "local"
+@whitePagesCacheTtl = "120s"
+@considerEnv = "false"
+@executionEnvironment = "external"

 @lightstep_collector_host = "HO123"

Let's say I copied all that. On a mac, we have the cool pbcopy command to capture stdout and put it on the clipboard. So git diff | pbcopy will copy it for us. Now you can put it in a pull request comment or send it to someone on chat.

Ok so a team mate sent you that. You can read it as a human and see it changes a few lines. You could edit the file to make it match, but git can do it for you. Pipe the output to git apply. Again on mac we have the nice pbpaste command to send our clipboard to stdout. So one option is to copy the text from the pull request comment and pipe it to git apply.

$ pbpaste | git apply

Another option is to put the contents into a file and point git apply to that. Say you paste the diff into new.patch, you can apply the changes with:

$ git apply new.patch && rm new.patch

I delete the patch since we are done with it and don't want to commit it to git.

This is something I do often. I used git for a long time without knowing about apply or that the output of diff can be used as input. It has helped me a few times.

Failed deploy to dokku

Nick F — Mon, 12 Oct 2020 02:58:16 +0000

I've got this Scala Play app that I develop on my laptop and deploy to a Dokku on a small Linode VPS.

These days, when I deploy, I get these failures:

$ git push dokku master
Counting objects: 72, done.
Delta compression using up to 4 threads.
Compressing objects: 100% (70/70), done.
Writing objects: 100% (72/72), 10.72 KiB | 731.00 KiB/s, done.
Total 72 (delta 45), reused 0 (delta 0)
-----> Cleaning up...
-----> Building yt from herokuish...
-----> Adding BUILD_ENV to build environment...
-----> Warning: Multiple default buildpacks reported the ability to handle this app. The first buildpack in the list below will be used.
       Detected buildpacks: multi scala
-----> Multipack app detected
=====> Downloading Buildpack: https://github.com/heroku/heroku-buildpack-scala
=====> Detected Framework: Play 2.x - Scala
-----> Installing JDK 1.8... done
-----> Running: sbt compile stage
       [info] Loading global plugins from /tmp/scala_buildpack_build_dir/.sbt_home/plugins
       [info] Loading settings for project scala_buildpack_build_dir-build from plugins.sbt ...
       [info] Loading project definition from /tmp/scala_buildpack_build_dir/project
       [info] Loading settings for project root from build.sbt ...
       [info] Set current project to play-ytdl (in build file:/tmp/scala_buildpack_build_dir/)
       [info] Executing in batch mode. For better performance use sbt's shell
       [info] Compiling 21 Scala sources and 1 Java source to /tmp/scala_buildpack_build_dir/target/scala-2.13/classes ...
remote: .sbt_home/bin/sbt-extras: line 241:   480 Killed                  "$@" < /dev/null

       !     ERROR: Failed to run sbt!
       We're sorry this build is failing. If you can't find the issue in application
       code, please submit a ticket so we can help: https://help.heroku.com
       You can also try reverting to the previous version of the buildpack by running:
       $ heroku buildpacks:set https://github.com/heroku/heroku-buildpack-scala#previous-version

       Thanks,
       Heroku

To MYDOMAIN.com:yt
 ! [remote rejected] master -> master (pre-receive hook declined)
error: failed to push some refs to 'dokku@MYDOMAIN.com:yt'

The .sbt_home/bin/sbt-extras: line 241: 480 Killed is the important part. Searching the web about it, this is not a Dokku issue. I can reproduce by checking out my source code on the VPS directly and trying to sbt stage to build the project myself. I get the same Killed message eventually.

The problem is I don't have enough RAM to build the project :-O

My options are to get more RAM, or compile the project locally and deploy an artifact. I'll look into the second option, but I might get lazy and buy a larger VPS. It will depend on my free time.

I make this post in case someone else is putting this error message into google and needs a hint on what is happening.