Manufacturing took over simple work so that we could do the interesting stuff. Computers should be the same.
(This is why I’ve just spent some time bashing custom HTTP requests so that the ST:Ascendancy PBF can have an automatic status dump, which is (players+1) custom images dropped into a post, rather than my having to paste them into said post.)
Anyone got a recommendation for a USB switch? Essentially looking for a KVM without the “V”. There are an awful lot on Amazon.
To switch what devices? If all you want to do is switch input devices, any should work fine. (but some devices don’t deal with it well.)
If you’re wanting to switch disks or something, good luck…
If the devices are on the same network, you could use a “virtual KVM”. I’ve used similar things in the past, but haven’t in a while. And currently it’s not an option for me between my personal computer and my work laptop, as my work laptop has a direct VPN connection that blocks LAN traffic (which is just fine, anyway, since I connect my work laptop to my guest network anyway).
Searching your favorite search engine should reveal some “virtual KVM” options… I’m somewhat wary of them these days due to the likelihood of them just being outright keyloggers…
Just a wired K and M. Looking at some of the Amazon reviews it seems that some of the switches do dubious things with the power rails and I see people also asking questions about keyloggers…
I previously owned this but I lost it in a move at some point and no longer need the V ports anyhow. I do like the idea of being able to secret the box under my desk and route a button to somewhere convenient.
This is something I used to do, but one of these devices is a work laptop and I’m not even sure at this point whether the security stuff will even let me plug in an un-recognised KVM…
Basically middleware that sits between your music collection and some sort of audio output device. Particularly useful because it has a very open interface and it’s easy to write software which will talk to it. (For example, I’ve put together a shim to link to melodice and a playlist manager for parties.) There are also mobile clients, so when I have people over they can queue music out of my collection with their phones and whatnot.
I was aware of mpd, but I parsed that domain name as “music police department”, and was afraid to click, lest they tell me I’m too uncool to be allowed music.
If anyone was too uncool to be allowed music, it would be a music police department.
I have a new phone.
I spent quite a few hours yesterday moving stuff.
And now I am waiting on 2 paper letters to be able to access some important stuff from the new phone. Yeah!
I am carrying around tons of messenger data I noticed. I skip moving the Wire history as it is only a single chat I use wire for these days. I should probably consider asking my friend to switch over to something else. But my Signal has about 10k messages in it and my threema backup is 1GB (including photos)
I am not sure I should actually keep that many messages. They are fleeting and transient it is not like I go back more than a few weeks to look up stuff. Signal has a mode where it automatically forgets older messages. Maybe I should use that. Do any of you do this? How do you deal with all the „old“ information?
So far I’m letting Signal build up, but I don’t use it very heavily (only one group chat and that’s fairly sporadic). I’m an archivist packrat by nature but as long as the stuff doesn’t pile up in physical space I don’t tend to mind it.
I used to keep old emails… like I have old letters… but when my server crashed in 2016 and took R‘s emails with it (mine survived as I started backing up with my own mails and the thing went away halfway through the panicking)… I started looking at older stuff and decided I was never going back to those things anyway. I still have a few from early on that I moved from account to account. They are like historical artifacts from my „elm“ days. But except for photos and my „writing“ … old data… ah well now that I use git for some things it will obviously accumulate. But digitalization should not only help us keep data it should also make it easy to get rid of it with precision like „please delete anything older than…“
I was a bit sad when my Hotmail account wiped itself (because of a year of not logging in) as that was a vast chunk of my adolescence. But also that was a vast chunk of my adolescence.
I’ve been approaching the limit of my Gmail account for some time now, I expect there’s an awful lot of stuff I could get rid of. But sometimes I find it handy to dig up an order confirmation from a decade ago.
My parents mainly use Skype which doesn’t seem to keep that long a record of conversations which is a bit of a shame, but I also don’t ever really scroll back that far.
Not to be too techie in general chat… several people I know have mentioned the term “boss code”, which happens when the boss was a coder back in the day but isn’t really any more…
I’m feeling empathy for those bosses right now, while I cope with a god-awful object-obfuscated nightmare of a codebase that has emerged, dripping from layer upon layer of oily abstractions, out of something that was originally understandable to humans. The part which sometimes makes me feel like a “back in the day” coder is that apparently some people prefer this.
Not A Coder, here - and believe me I’ve been trying to teach myself first Java, then Python for the last 20 odd years! The abstraction continues with cloud storage. We all know that at it’s base it’s just “other people’s servers” so why do they have to make up arbitrary abstractions like “block storage” “object storage” and “compute units”? Is it just to make us old fashioned sysadmins sad and uncomfortable?
And confuse anyone trying to measure costs. Though to be fair I understand that they’re trying to get away from “a server does x and y” into “a unit does x, and you don’t know or care which servers your x,000 parallel units are running on”.
They are but Lego bricks.
Sure, you could write a piece of software that handles it’s own database storage, encoding bits directly into a binary blob file; and it could include its own memory manager and a custom-written TCP-IP stack.
But the development practices of today are based upon herding together stacks of libraries and utility shims to assemble applications, services and platforms without any one particular task or subsystem taking away from the goal at-hand.
If you want to make a microblogging website, your effort should be on making a microblogging website, not on the backend storage systems.
Block and object storage solve very different problems. Block storage is just a filesystem, stored on someone else’s disks; it’s suitable for general purpose use, and for loads with lots of updates. There are some serious constraints on size/performance (block stores bigger than few tens of terrabytes don’t work well), replication across large distances if you care about consistency, and latency.
Object storage is for blobs of bits that are read only or nearly read only (write means replace, in most systems; some support versioning). That makes them well suited for things like image stores, video streams, and lots of other things. Because the client asks for objects by id, it’s easy to do replication, multilevel caching, and to scale to zettabytes, if required. They usually also have metadata attached to them, which allows indexing, searching, etc.
That’s just the problem. I’m Asperger’s enough to care what the infrastructure I am building on looks like. Which makes me infinitely unsuited for the modern, multi-layer-abstracted world of modern corporate IT. 
Agreed. I’m the same way. But I also have ADHD and if I don’t leverage other people’s work, I get bogged down rewriting make(1) because I don’t like the quirkiness of Makefile syntax.