[-] tko@tkohhh.social 2 points 3 days ago

It was not a prediction of victory... it was a 71% chance of winning the electoral college. Said another way, they gave her a 29% chance of losing.

https://projects.fivethirtyeight.com/2016-election-forecast/

[-] tko@tkohhh.social 2 points 4 days ago

ahhhh yes, that makes perfect sense... thank you for pointing that out! Especially since I'm not good enough with vi to know how to bulk delete the first character in specific lines, I had to manually arrow and delete.

[-] tko@tkohhh.social 2 points 4 days ago* (last edited 4 days ago)

I successfully migrated postgres 15 to 16. I followed the general idea of the guide you posted, but I found it a little easier to do a slightly different process. Here's what I did:

  1. docker-compose down for the lemmy instance

~~2. edit the docker-compose.yml file and comment out all of the services except postgres. In addition, add a new volume to the postgres service that looks something like this: - ./volumes/miscfiles:/miscfiles~~

  1. edit the docker-compose.yml file and add a new volume to the postgres service that looks something like this: - ./volumes/miscfiles:/miscfiles
  2. docker-compose up -d postgres (this starts just the postgres service from the docker compose file)
  3. docker exec -it [container name] pg_dumpall -U [username] -f /miscfiles/pgdumpall20240628 (I think this will work, but it's not exactly what I did... rather, I ran docker exec -it [container name] bash, and then ran pgdumpall -U [username] -f /miscfiles/pgdumpall20240628. The end result is a dumpall file saved in the ./volumes/miscfiles directory on the host machine)
  4. docker-compose down
  5. mv ./volumes/postgres ./volumes/postgresBAK20240628 (move your existing postgres data to a new directory for backup purposes)
  6. mkdir ./volumes/postgres (re-create an empty postgres data folder. make sure the owner and permissions match the postgresBAK20240628 directory)
  7. edit the docker-compose.yml and update the postgres image tag to the new version
  8. docker-compose up -d postgres (you'll now have a brand new postgres container running with the new version)
  9. docker-exec -it [container name] psql -U [username] -f /miscfiles/pgdumpall20240628 (again, I think this will work, but I bashed in and ran the command from within the container. This also allows you to watch the file execute all of the commands... I don't know if it will do that if you run it from the host.)
  10. docker-compose down

~~12. edit the docker-compose.yml and un-comment all of the other services that you commented out in step 2~~

  1. docker-compose up -d

Hopefully that helps anyone that might need it!

edited to reflect the comment below

[-] tko@tkohhh.social 6 points 4 days ago

I agree... I was simply clarifying that Nate Silver did NOT predict that Hillary would win (nor is he predicting that Trump will win this election), which is a common misunderstanding about probability. For these types of models to be meaningful to the public, there needs to be literacy on what is meant by the percentages given. Really, I'm just reinforcing rodneylives' point from another angle!

[-] tko@tkohhh.social 11 points 4 days ago

I don't remember him predicting that she would win. His model (https://projects.fivethirtyeight.com/2016-election-forecast/) gave her a 71% chance of winning. 71% is a long way from 100%, and the result of that election definitely fit within the model.

That said, you are absolutely correct... we need to keep shining a light on the realities of each of these candidates, because in the light of day Biden is a much better choice than Trump.

[-] tko@tkohhh.social 1 points 6 days ago

Thank you, that's super helpful!

11
0.19.4 Prerequisites (tkohhh.social)

The release notes for 0.19.4 mention that this version requires an upgrade to both pictrs and postgres, but there isn't any information about what versions of those are supported. Is there someplace else I should be looking for this information? Or should I just YOLO with the latest version of each?

[-] tko@tkohhh.social 47 points 5 months ago
[-] tko@tkohhh.social 32 points 6 months ago

Clarification question: Why would you expect somebody to DM you?

8
submitted 7 months ago by tko@tkohhh.social to c/unraid@reddthat.com
18
submitted 7 months ago by tko@tkohhh.social to c/unraid@reddthat.com
16
submitted 10 months ago by tko@tkohhh.social to c/unraid@reddthat.com
[-] tko@tkohhh.social 21 points 10 months ago

I think the logic is that you, by virtue of the fact that you made the comment (or post), agree with the contents. If you comment or post but don't agree with what you wrote, you could manually remove the upvote. Alternately, you could just not make the comment to begin with.

16
Report Function? (tkohhh.social)
submitted 10 months ago by tko@tkohhh.social to c/lemmy_support@lemmy.ml

I'm a little confused as to how the "report" function works. I'm an instance admin, but if I report a post on another instance, that report shows up in MY report queue. If I mark that report as "resolved" to get rid of the notification count, does it resolve it on the home instance?

3
submitted 10 months ago by tko@tkohhh.social to c/food@beehaw.org

I'm curious what everyone thinks about the posts that have shown up here that are simply a link to purchase an item on Amazon?

These posts are coming from users that don't have any other engagement other than these posts. This does not seem like a good-faith effort to engage with the community, but rather a cheap attempt to generate clicks and sales.

I would like to see these posts banned, but I'm curious how the rest of the community feels.

[-] tko@tkohhh.social 20 points 11 months ago

I'm not familiar with lateral thinking, but this reads like a joke... Is it intended to be funny, or are you actually expected to come up with all of those answers?

[-] tko@tkohhh.social 34 points 11 months ago

This is a question for Randall Munroe. I wish I knew how to summon him.

[-] tko@tkohhh.social 70 points 11 months ago

Selling is an art... and this is why I'm not in sales. I have absolutely no interest in trying to convince somebody to buy something.

1
submitted 11 months ago by tko@tkohhh.social to c/usnews@beehaw.org
7
submitted 11 months ago by tko@tkohhh.social to c/lemmy_support@lemmy.ml

I'm hoping one of the developers can help point me in the right direction.

I'm trying to create a theme, and I started by downloading the litely.scss and _variables.litely.scss files from the github repo. I renamed the files and made some changes, and then used https://bootstrap.build to generate a css file. I dropped that file (along with my scss files) into the extra_themes directory on my server.

Sure enough, my new theme was available in the list of available themes. However, when I applied the theme, I found that the layout was broken (see screenshot).

So, my question is: how can I generate the css file in the proper manner so as not to break the layout? Is there a specific Bootstrap tool that we should be using to generate the css? Are there additional input files that need to be included in order to generate the correct css?

Thanks for any help you can provide!

1
submitted 11 months ago by tko@tkohhh.social to c/unraid@reddthat.com
3
submitted 1 year ago* (last edited 1 year ago) by tko@tkohhh.social to c/unraid@reddthat.com

My favorite new feature of 6.12 is the "Exclusive Shares" concept. Here's a rundown:

Background

Unraid user shares are a FUSE, which allows data on multiple drives to be presented as a single file system. This idea is at the heart of Unraid's Array, as well as the concept of "Cache Pools" (now known as Named Pools). Any time you see a path that starts with /mnt/user in Unraid, that's a FUSE (prior to 6.12, that is).

FUSE is great for giving us this transparent way to view our files without having to worry about which physical drive those files reside on. However, this comes at a cost... and that cost is reduced performance for applications running on an SSD Named Pool.

This performance penalty wasn't always noticeable, but it would sometimes rear it's ugly head in unexpected ways (exampes: graylog and duplicati). There was a workaround, assuming your appdata user share was stored entirely on one Named Pool: you could update your docker bind mounts to /mnt/[poolname]/appdata instead of /mnt/user/appdata. This bypassed the FUSE.

Exclusive Shares

With Unraid 6.12, Limetech introduced "Exclusive Shares" as part of the Share Storage Conceptual Change. This gives us a built-in way to bypass FUSE on an entire user share.

In order for a share to be designated an Exclusive Share, the following must be true:

  • "Primary Storage" must be a Named Pool
  • "Secondary Storage" must be set to none
  • Files for that share must exist entirely on the Primary Storage device

Setup

In order to use Exclusive Shares, you first have to enable them. Go to Settings > Global Share Settings and change the Permit Exclusive Shares setting to Yes. You'll have to stop your array in order to make this change.

Next, make sure that your appdata share is stored entirely on the Named Pool. Go to Shares and click Compute in the Size column for your appdata share. This will tell you how much data for this share is saved on each drive. If the only drive reported is your Named Pool, you're all set. If you've got more than one drive, you'll need to disable docker in settings, and then run the mover.

Once your sure that your entire appdata share is saved on your Named Pool, you need to change your appdata share settings. On the Shares tab, click on appdata to bring up the settings. Change the Secondary Storage option to None.

If you did it correctly, after you Apply the changes you will see the Exclusive Access field on the appdata share change from No to Yes.

Finally, if any of your docker container bind mounts use mnt/[poolname]/appdata/..., you can change those to mnt/user/appdata/....

I hope this helps anybody who might have been frustrated with the appdata FUSE performance in the past! If you have any questions, let me know!-

view more: next ›

tko

joined 1 year ago