I worked at a bank at the time. We were moving to a new system and running recons against the old system to check the behaviour was the same. I had to run a manual recon of the old system vs the new 4 times per day. There was a lot of focus from management and users on the new system.
The week leading up to Christmas, I was the one person not on holiday yet, and also the most junior person on the project. I found that week so stressful, as I had to run these recons and quickly decide whether each break was real or not before reporting to the users. Despite having worked on that system, I had very little confidence and didn't have the same intuitive mental model of the system my colleagues had. I had to dig into each break case-by-case, but they seemed much more able to understand what was going on via a few simple queries.
Anyway, I get through the week and left for the holidays on Thursday evening. I'm just grateful that I've gotten through it. Then, around 3pm on the Friday, I see a missed call from the tech lead. I log in, and everything's on fire. I join the incident call, and it turns out that we hadn't processed a single trade in the new system that whole week. I discover that it was thanks to a config change I'd made several weeks before, that had just made it to production. No-one (neither the users, nor I) had realised! But we missed several hundred million pounds worth of payments in that week as a result.
It was so jarring, having been relieved that I made it to the holiday, then joining the incident call and struggling to work out what to do. I completely dissociated and my mind was blank. I remember being on the call and really passively and calmly walking around my room. I kept thinking "I need to do x, I need to do y" but my mind couldn't focus and I was just staring at the screen. At some point I just lay in bed with my laptop while on the call.
There had been a total failure of process: my change had been approved by two people, the nonprod environment was configured differently in a way that didn't expose the bug, the recon failures looked very similar to the false positives, and there were so many false positives that it was impossible to dig into all of them. Meanwhile, we didn't have basic queries monitoring that trades were flowing in, and the users weren't paying much attention either, until they realised that it was broken.
Still, I made a lot of mistakes. I should have just escalated that there were breaks instead of trying to figure it out myself. I shouldn't have been afraid to call the tech lead and bring them out of their holiday. And I shouldn't have been afraid of the confrontation with the users.
Anyway, that experience really messed me up mentally for a long time. I lost so much confidence and became so much more scared of production (not in a healthy way). It really was not the right environment for me.
The Software Freedom Conservancy, among other things, acts as a fiscal host for its projects. You can donate to individual projects, or make a general donation to conservancy and they will use it in whatever way best advances software freedom.
Likewise, the FSF has a "Working together for free software fund" which has a few projects listed. I can't see a way to donate to the fund as a whole, but if you don't mind some (most?) of the money going to things that aren't directly developing software, you could make a standard donation to the FSF.
I never stop being amazed by running snac on the server, with toot and tuba as clients, and seeing everything work seamlessly with everything else. None of this software ever expected to interact with the others - and it all works!
I should have the same feeling about email, the web, XMPP etc, but I'm just so used to those that it isn't awe-inspiring to me anymore.
Personally, I'm pretty sentimental about my bike (also a beater, but single speed) and want to keep it going almost no matter what. I think only the frame has remained the same throughout, and at times, I've definitely spent more than the bike is worth to fix it up.
I know that with cars, there's a definite cliff edge where it becomes prohibitively expensive to maintain compared to a new one. But I feel like that's not really true with bikes, perhaps because there are just fewer parts to replace, and it's less invasive to strip out any given part?
I suppose it depends on a few things. Do you have any sentimental attachments to it, and is there anything you would change about the bike if you were buying a new one? Do you think you can get a better one for $200?
Someone else suggested osmium tags-filter on the downloaded PBF files (which are ~150 MB), and that's working well at the moment. I'll keep this in mind as I'm presuming that importing into a database will be more efficient in case I ever increase the size of the map I'm working with.
Thank you for the tips! I should have been more precise in my question. The downloaded maps are ~150MB, in PBF format (although I would have been happy to use any other standard format if needed). I went with osmium tags-filter in the end, and it seems to be working well.
It's an alternative to Lemmy with some different features. Since it uses the same protocol under the hood, its instances federate with Lemmy. There's more info on the differences here.
I get you. I can never think of anything that would be interesting to post or ask in the more discussion-oriented communities, let alone choose a specific one to post in. I definitely find comments easier, as well as posting to more niche communities. I feel the scope is usually better defined there.
Would you say it's about not knowing if your post would be accepted in the community, or just finding the best place for it? If it's the latter, AskLemmy could be good for general questions, or failing that, any of the casual chat communities such as !chat@beehaw.org.
As long as your post meets the rules of the community/instance, I feel it's better to post somewhere than not at all - people can always crosspost it elsewhere if they like.
You can trust the software in your distro's repositories (if you run a distro with well-maintained repositories). This is because, generally only well-known software gets packaged, the packager should be familiar with both the project and the code, and everything is rebuilt on the distro's own infrastructure, to ensure that a given binary actually corresponds to the source.
It might still be possible for things to slip through, but it's certainly much safer than random programs from online.
I independently thought of the same idea. While I'm daydreaming, I had some extra features that would be useful to me in a dream world:
It would be good to be able to apply this to posts (that are not mine) as well, or even to a link (i.e. all comments that would show up under the crosspost aggregation feature)
One problem I have with GitHub is that the subscription list perpetually grows and is never pruned.
It would be nice if I could make such subscriptions, for instance, automatically expire n days after the last interaction
Or, if there is a list of subscriptions somewhere, if I could manually "prune all whose last interaction is more than n days"
I'm not sure what the best UI would be, whether everything should go in notifications, or whether there should be a dedicated view for these subscriptions
And, should that view show the whole thread underneath the top-level post you subscribed to?
Now, I'm not asking companies to open-source their entire codebase. That's unrealistic when an app is tied to a larger platform. What I am asking for: publish a basic GitHub repo with the hardware specs and connection protocols. Let the community build their own apps on top of it.
I agree with this. I think the most important thing is not necessarily the original company releasing their proprietary code (although that would be nice), but it being easy (and legal!) for hackers to reverse engineer and/or build on top of the platform.
The irony is that, since most such products will have some GPL'd code in there somewhere, most products already basically have such a requirement, thanks to the section requiring complete corresponding source including installation instructions. Hopefully, the Vizio case will establish the precedent that users, as well as copyright holders, can take action against such companies.
This might be a silly question, but in what ways did it get worse? Is it the size of the keyboard changing, the predictions not being as good anymore or something else?
With my knowledge of tech companies, I'm not exactly surprised, but I'm not an iPhone user and struggling to understand how a keyboard of all things could get worse.
I worked at a bank at the time. We were moving to a new system and running recons against the old system to check the behaviour was the same. I had to run a manual recon of the old system vs the new 4 times per day. There was a lot of focus from management and users on the new system.
The week leading up to Christmas, I was the one person not on holiday yet, and also the most junior person on the project. I found that week so stressful, as I had to run these recons and quickly decide whether each break was real or not before reporting to the users. Despite having worked on that system, I had very little confidence and didn't have the same intuitive mental model of the system my colleagues had. I had to dig into each break case-by-case, but they seemed much more able to understand what was going on via a few simple queries.
Anyway, I get through the week and left for the holidays on Thursday evening. I'm just grateful that I've gotten through it. Then, around 3pm on the Friday, I see a missed call from the tech lead. I log in, and everything's on fire. I join the incident call, and it turns out that we hadn't processed a single trade in the new system that whole week. I discover that it was thanks to a config change I'd made several weeks before, that had just made it to production. No-one (neither the users, nor I) had realised! But we missed several hundred million pounds worth of payments in that week as a result.
It was so jarring, having been relieved that I made it to the holiday, then joining the incident call and struggling to work out what to do. I completely dissociated and my mind was blank. I remember being on the call and really passively and calmly walking around my room. I kept thinking "I need to do x, I need to do y" but my mind couldn't focus and I was just staring at the screen. At some point I just lay in bed with my laptop while on the call.
There had been a total failure of process: my change had been approved by two people, the nonprod environment was configured differently in a way that didn't expose the bug, the recon failures looked very similar to the false positives, and there were so many false positives that it was impossible to dig into all of them. Meanwhile, we didn't have basic queries monitoring that trades were flowing in, and the users weren't paying much attention either, until they realised that it was broken.
Still, I made a lot of mistakes. I should have just escalated that there were breaks instead of trying to figure it out myself. I shouldn't have been afraid to call the tech lead and bring them out of their holiday. And I shouldn't have been afraid of the confrontation with the users.
Anyway, that experience really messed me up mentally for a long time. I lost so much confidence and became so much more scared of production (not in a healthy way). It really was not the right environment for me.