camber-weaver

joined 11 months ago
[–] camber-weaver@alien.top 1 points 9 months ago

This is pretty much exactly what I do. The only real hiccups I can think of:

  1. If you use Org Roam the different machines will have their own Sqlite databases and thus you may need to do an org-roam-db-sync when hopping.

  2. org-agenda may need to be refreshed when hopping machines.

I have bound keys for both of those things to make the manual portion as non-disruptive as possible.

[–] camber-weaver@alien.top 1 points 9 months ago

Yes, this. I've had this happen a number of times and the solution was just to bump the specified elpaca version. I've had no negative consequences from doing so.

[–] camber-weaver@alien.top 1 points 9 months ago

Thank you for all of this info. I'm planning to slowly roll out three of these servers with Proxmox as a cluster. I'm just upgrading all of the available power outlets and spec'ing UPSes now.

[–] camber-weaver@alien.top 1 points 10 months ago

As a minimal effort first attempts, I just installed XanMod kernel. If that makes a noticeable impact I will loop back here to report.

[–] camber-weaver@alien.top 1 points 10 months ago (1 children)

I've noticed I have a lot of latency right after startup and the thing which seems to help is a manual garbage collect. That's bit odd because the last thing I do in my config is set a 5 second timer which is fed a lambda to run a gc after the timer expires. Regardless, I still have better results if I run the gc manually.

I do load of metric ass ton of packages (and modes) and I recursively parse my Org/Org Roam directory for org-agenda items so I'm doing a lot of questionable things.

As others have said, sounds like we might both benefit by spending some quality time with the profiler.

 

Hey hey.

I use the heck out of tuedachu/ytdl. One thing I would like to improve about it is that when you queue a download with ytdl-download it appears to be a synchronous call and it freezes emacs until youtube-dl/yt-dlp is able to start the actual download.

I've yet to dive into the code to see what synchronous function is being called -- I'm very aware of the "single-threaded" nature of Emacs -- but I use other libraries that are likely shelling and they do not hang like this and so I would like to see if there's a way to I can bring whatever technique they are using over to a PR against tuedachu/ytdl.

Appreciate any advice from the crew here.

[–] camber-weaver@alien.top 1 points 10 months ago (1 children)

This is pretty interesting...

It appear as though your workflow is pretty similar to mine when producing technical documentation. I need to dig into this a bit. Hope you don't mind me asking a question after only having done a quick fly-by...

What is the ob type 'compile'? I've never seen that one. Something you've created or something that has just been under my radar?

[–] camber-weaver@alien.top 1 points 10 months ago

Out of curiosity I installed the .debs. They appear to have been empty as a 'dpkg -L' only shows two directories under /usr/local.

I see recent commits but it appears as though the builds are broken at the least.

[–] camber-weaver@alien.top 1 points 10 months ago (2 children)

Via another reddit thread on the topic, I found this YouTube playlist which has been super helpful. Just re-posting it here for posterity:

https://www.youtube.com/playlist?list=PLbNp7AIu8VYyzsocLUdCScF4zvou8rzS9

Thanks, y'all.

I don't think I need so many drives so now its just a matter of verifying that if I scoop one or two low-RAM-specs off of eBay for cheap I will be able to re-populate them with RAM.

 

I'm just pushing into rackmount hardware and I've been considering picking up an r730. From what I'm reading it seems *possible* that there are a couple of variants. Maybe related to available drive bays? Any chance someone could point me to some docs or some resource on variants?

Note: I don't consider differing CPU and RAM spec within a model as variants.

thx!

 

My "server" -- and by that I mean a re-purposed desktop with a boatload of SSDs -- just died. I'm not sure if I'm going to be able to resurrect it and I thought in parallel I would look into finally going rackmount. I've a 15U 21.5" deep StarTech rack and I'm curious what folks here would recommend for that depth for running Proxmox? It'd be nice if it were power-efficient as possible. Not super noisy might be a consideration too but I'm running cable to locate the rack in a quiet corner of the basement so perhaps not a huge concern.

In a few months I might pick up a handful of additionals and do a Proxmox cluster.

Appreciate any advice y'all might have.