|
There was a consensus that we spend too much time in meetings, so a new decree came down: no more meetings! On Wednesday afternoons. All other times fine. 10% set aside for actual independent work is plenty
|
#
?
Nov 1, 2022 01:49
|
|
|
|
| # ? May 10, 2024 01:32 |
|
|
Macichne Leainig posted:They want to do daily sync-ups on top of our daily standup which is now in the afternoon and includes more people than ever Think I mentioned before, many years ago I was working at EA when they were launch EA.COM. They had standups at 6p and 10am. It took me awhile to realize what they were really saying, "Enough work should be happening after 6pm to necessitate a change of status for a 10AM standup."
|
|
#
?
Nov 1, 2022 01:55
|
|
|
If I have a meeting at 6 and another meeting at 10 then I have (at best) zero business minutes between them. I would record and replay my 6pm update at 10am. I would not work at EA very long. edit: also I would not go to the 6pm meeting, maybe nobody notices
|
|
#
?
Nov 1, 2022 02:16
|
|
|
Yeah, the reputation that the games industry has exists for a reason.
|
|
#
?
Nov 1, 2022 02:18
|
|
|
Yep this was over 20 years ago when we were young and in our twenties ourselves. Can't imagine that flying today the way the industry has evolved.
|
|
#
?
Nov 1, 2022 02:30
|
|
|
Hughlander posted:Think I mentioned before, many years ago I was working at EA when they were launch EA.COM. They had standups at 6p and 10am. It took me awhile to realize what they were really saying, "Enough work should be happening after 6pm to necessitate a change of status for a 10AM standup." Place I worked, we were behind on schedule and the guy running the team (new manager) had the brain genius idea to schedule two standups a day instead of one to fix this. At least it was like 9am and 1pm, but still, he expected this to improve productivity, and, dear reader, somehow it did not.
|
|
#
?
Nov 1, 2022 14:37
|
|
|
Scammers are already jumping on the Twitter bandwagon to make some money https://www.reddit.com/r/Scams/comments/yj6qsl/a_scam_targeting_bluechecks_on_twitter/Trapick posted:There was a consensus that we spend too much time in meetings, so a new decree came down: no more meetings! On Wednesday afternoons. All other times fine. 10% set aside for actual independent work is plenty 1/5 is 20%? Wondering if I missed some kinda joke 
|
|
#
?
Nov 1, 2022 16:10
|
|
|
Where are all of the good engineering managers? I had a 6 month stint under an amazing manager and the team was kicking rear end, I was the happiest I've been at work my entire career, and we were building great marketable poo poo. The manager left, stepping back into a founding engineer role somewhere else, and the new manager is just like every other one I've had. Slow to act, slow to take the lead, slow to collaborate with other managers and instead trying to figure everything out on their own. I just want a good manager, which god do I need to sacrifice my goats to? mitztronic posted:1/5 is 20%? Wondering if I missed some kinda joke Only Wednesday afternoon, so half that.
|
|
#
?
Nov 1, 2022 16:14
|
|
|
Judge Schnoopy posted:Where are all of the good engineering managers? I had a 6 month stint under an amazing manager and the team was kicking rear end, I was the happiest I've been at work my entire career, and we were building great marketable poo poo. Early in my career, a week after I started a job, the manager I had interviewed with quit, along with half his team. They all went to work together somewhere else. I'm pretty sure they've been following each other to different companies most of their careers. As essentially a mercenary for most of my early career, I couldn't quite understand that mentality. But I'm starting to get it. Good managers are really hard to come by. After a string of terrible managers, my last two have been perfect. Both spent most of their time getting out of the team's way, managing up, and gently encouraging us to do things that would help our careers. I would probably follow my current manager to the next company if they ever leave.
|
|
#
?
Nov 1, 2022 16:25
|
|
|
Judge Schnoopy posted:Where are all of the good engineering managers? I had a 6 month stint under an amazing manager and the team was kicking rear end, I was the happiest I've been at work my entire career, and we were building great marketable poo poo. One of mine for sure. I flagged him as future director material when he was a lead. If I stick around long enough I'm sure I'll be working for him some day.
|
|
#
?
Nov 1, 2022 17:47
|
|
|
I have a manager like that, I'll follow him until the ends of the earth. I have to hold him accountable sometimes but I guess it works because he introduces me as his second-in-command
|
|
#
?
Nov 1, 2022 17:49
|
|
|
I had a manager who sat next to me and relied on me for keeping a bunch of poo poo straight and I felt valued as basically a second-in-command too. We would even hang out sometimes outside of work, including when he invited me to his wedding. Then he moved to an office and made me sit next to my least-favorite coworker and never figured out why I was so mad at him ever since lol
|
|
#
?
Nov 1, 2022 18:34
|
|
|
CPColin posted:Then he moved to an office and made me sit next to my least-favorite coworker and never figured out why I was so mad at him ever since lol You're an adult, did you ever consider talking to him about it? 
|
|
#
?
Nov 1, 2022 18:35
|
|
|
Of course not
|
|
#
?
Nov 1, 2022 19:09
|
|
|
sim posted:Early in my career, a week after I started a job, the manager I had interviewed with quit, along with half his team. They all went to work together somewhere else. I'm pretty sure they've been following each other to different companies most of their careers. As essentially a mercenary for most of my early career, I couldn't quite understand that mentality. But I'm starting to get it. Good managers are really hard to come by. After a string of terrible managers, my last two have been perfect. Both spent most of their time getting out of the team's way, managing up, and gently encouraging us to do things that would help our careers. I would probably follow my current manager to the next company if they ever leave. I am 100% mercenary in my work (and I am open to my managers about it), but following good managers to new roles makes perfect sense to me. Working with a good manager is waaaaay more comfortable than working with even an okay one, and that is one of the criteria I optimize for. The company can go gently caress itself for all I care, I want to work on interesting stuff for good money and with chill workplace environment
|
|
#
?
Nov 1, 2022 21:11
|
|
|
A coworker was gushing about Github Copilot today and lo and behold there is now a class-action lawsuit over it.
|
|
#
?
Nov 4, 2022 00:52
|
|
|
eXXon posted:A coworker was gushing about Github Copilot today and lo and behold there is now a class-action lawsuit over it. I love copilot but they're also not wrong
|
|
#
?
Nov 4, 2022 05:21
|
|
|
Will copilot be better or worse than just copy and pasting the first Stack Overflow result like junior devs do today?
|
|
#
?
Nov 4, 2022 12:53
|
|
|
worse so much worse (at least for C++)
|
|
#
?
Nov 4, 2022 13:08
|
|
|
smackfu posted:Will copilot be better or worse than just copy and pasting the first Stack Overflow result like junior devs do today? Junior?
|
|
#
?
Nov 4, 2022 13:38
|
|
|
smackfu posted:Will copilot be better or worse than just copy and pasting the first Stack Overflow result like junior devs do today? It's really good for boilerplate I find. Like if I'm writing unit tests once I get the first couple ones down and it figures out what I'm doing it'll practically write my next ones for me. But yeah it's not a replacement for knowing how to code, the code it suggests is often a bit wrong just like using StackOverflow if you're just blindly accepting what it gives you your poo poo is gonna be broken.
|
|
#
?
Nov 4, 2022 15:58
|
|
|
Copilot will happily create SQL injection attacks. Upvoted answers on StackOverflow almost always wont.
|
|
#
?
Nov 4, 2022 17:15
|
|
|
If someone commits and merges bad code, I think you should blame the dev who did it instead of the tools that they used
|
|
#
?
Nov 4, 2022 18:24
|
|
|
Nice username/post combo but the whole point of AI is to create a liability shield
|
|
#
?
Nov 4, 2022 18:44
|
|
|
Hi, it's me, your liability shield
|
|
#
?
Nov 4, 2022 19:15
|
|
|
This would be a great April Fool's post in QCS: all moderator and admin actions are now machine learning-based and irreversible
|
|
#
?
Nov 4, 2022 19:16
|
|
|
Cup Runneth Over posted:Nice username/post combo but the whole point of AI is to create a liability shield Discussion at work is the point is to be sued out of existence. This lawsuit supports that.
|
|
#
?
Nov 4, 2022 19:30
|
|
|
Love Stole the Day posted:This would be a great April Fool's post in QCS: all moderator and admin actions are now machine learning-based and irreversible Kind of like how that Amazon AI that got racist from its input data and flagged all black people, a Something Awful admin AI fed our shitposts will just ban all of us. Except that the Something Awful AI would have good cause to do it.
|
|
#
?
Nov 4, 2022 19:38
|
|
|
In the new LinkedIn sharing API if you want to submit a text post with a url in it and you want it to display OpenGraph data you have to submit all the data yourself. Meaning you have to fetch the opengraph data from the url, download the image, upload it to LinkedIn, hit another LinkedIn endpoint a few times until it says the image is done processing, and then submit that image ID along with the meta title and description you pulled out. For reference if you submit a tweet to the Twitter API with a url in it it just displays the opengraph data like you'd expect. But if you submit a LinkedIn post with a Twitter URL in the text the LinkedIn API sends back a 400 response. After this job I'm never doing anything related to social media again.
|
|
#
?
Nov 5, 2022 05:28
|
|
|
eXXon posted:A coworker was gushing about Github Copilot today and lo and behold there is now a class-action lawsuit over it. It was cool when it was free but now I'm not going to pay a monthly fee just to be able to write out a sentence in a comment to see what some other person I'll never meet wrote
|
|
#
?
Nov 9, 2022 18:58
|
|
|
Quote of the day:quote:We did all the work, but nobody knew what the requirements were.
|
|
#
?
Nov 10, 2022 21:15
|
|
|
Was just in a teams meeting and some dude was sharing his screen to show a spreadsheet, and the browser window behind him had search results for 'mouse jiggler' from amazon.com
|
|
#
?
Nov 14, 2022 10:26
|
|
|
beuges posted:Was just in a teams meeting and some dude was sharing his screen to show a spreadsheet, and the browser window behind him had search results for 'mouse jiggler' from amazon.com Snitches get stitches
|
|
#
?
Nov 14, 2022 19:03
|
|
|
bvj191jgl7bBsqF5m posted:Snitches get stitches
|
|
#
?
Nov 14, 2022 19:07
|
|
|
I want to do some event logging in our systems. We monitor some external services and import select data, and I'm often asked "why didn't x piece of data from y service get imported?" and right now I'm never able to provide a good answer to that question. What I want is to create an audit log, so I can say "we evaluated that piece of data at 12:33pm and it didn't meet any of the conditions for importing so we skipped it." I would also like to be able to build a UI for this so that other people in the org can find those answers, so just console.logging and then searching the log files isn't gonna work. What I don't know is what storage backend I should be looking at. We have a MySQL database but I don't know if this belongs in there. Do I want something like Redshift, DynamoDB, ElasticSearch, or are there other options I should be looking at? It's also definitely an option to delete all data after a day or two since we're mainly trying to figure out something that just happened, so in that scenario maybe just chucking it in MySQL is fine? It's going to be a TON of writes though and I don't think I want to clog our main DB with that. Another possibility is logging to a service like https://axiom.co and using their JSON API to power the UI. That might be easier but maybe less performant than using a database? prom candy fucked around with this message at 22:39 on Nov 14, 2022 |
|
#
?
Nov 14, 2022 22:19
|
|
|
Elasticsearch, Logstash, Kibana is one set of tools that matches what you are wanting to do pretty directly. I dont know how it compares to alternatives at this point but looking into that should get you started
|
|
#
?
Nov 15, 2022 16:50
|
|
|
Steve French posted:Elasticsearch, Logstash, Kibana is one set of tools that matches what you are wanting to do pretty directly. I dont know how it compares to alternatives at this point but looking into that should get you started This is something Ive also been wanting to investigate more we currently log to graylog (free tier for now) and wanted to start pulling metrics into Prometheus, but Ive been wanting to look at alternatives, potentially something more integrated like ELK, or, if it makes more sense to split things out into more specialised apps Im happy to do that too. There seem to be a whole lot of log/metric/apm platforms out there is it worth paying for graylog when the time comes ($7500/year) and doing metrics separately in Prometheus as we currently are doing/planning? Or is it worthwhile switching to ELK or something else? Were not a big company and dont have anyone specifically trained or with a lot of experience in any of these platforms either so wed also not want something super powerful but which needs a lot of expertise to configure properly
|
|
#
?
Nov 15, 2022 17:07
|
|
|
Steve French posted:Elasticsearch, Logstash, Kibana is one set of tools that matches what you are wanting to do pretty directly. I dont know how it compares to alternatives at this point but looking into that should get you started Perfect, thank you!
|
|
#
?
Nov 15, 2022 20:18
|
|
|
beuges posted:This is something I’ve also been wanting to investigate more… we currently log to graylog (free tier for now) and wanted to start pulling metrics into Prometheus, but I’ve been wanting to look at alternatives, potentially something more integrated like ELK, or, if it makes more sense to split things out into more specialised apps I’m happy to do that too. I don't know exactly what your usage is but that price tag seems high - datadog does log aggregation and APM monitoring and charges per host rather than a bulk rate, it might be more aligned with what you need. Alternatively, yeah, ELK is free (or rather, costs hosting) but obviously there's self-maintenance overhead involved. You mentioned Prometheus - if this is going into kubernetes, the logstash bit of ELK is replaceable with fluentd for a more native approach.
|
|
#
?
Nov 15, 2022 22:48
|
|
|
|
| # ? May 10, 2024 01:32 |
|
|
ChickenWing posted:I don't know exactly what your usage is but that price tag seems high - datadog does log aggregation and APM monitoring and charges per host rather than a bulk rate, it might be more aligned with what you need. Alternatively, yeah, ELK is free (or rather, costs hosting) but obviously there's self-maintenance overhead involved. I think its in the ballpark for 24/7/365 dedicated compute instances SaaS product to do log collection, APM and tracing. If you put a basic instance in Elastic Cloud calculator its $5k. https://cloud.elastic.co/pricing Datadogs pricing model is complex, usage often based on abstract concepts and to me seems pretty clearly intended to sell you part of the product on a per-host monthly cost to get their foot in the door then constantly surprise and delight with incremental costs for features and/or usage surprises. https://docs.datadoghq.com/account_management/billing/pricing/ So its: $40/month/host for the APM that sorta similar but not as good as Kibana PLUS $0.10/Mo/GB of ingested logs PLUS 1.70/1M Log Lines/Mo for them to store it. Aside from that I would 100% encourage people to implement a log aggregation, distributed tracing and APM tool in their work. Its life changing how much it can reduce cycle time to understand how complex systems are working (or not) at all stages of the SDLC. E: Just to be clear I prefer ELK and my experience is using these tools to help find and convince dev teams of whats going wrong in large distributed systems and how to un-gently caress it. I have not had to provision, configure or maintain the observability platforms professionally but I have used ELK a few times to ingest logs related to gaming poo poo I was doing to understand internal state of the game better  Docs were good, it was easy to spin up a cluster locally using docker.
FormatAmerica fucked around with this message at 02:20 on Nov 16, 2022 |
|
#
?
Nov 16, 2022 01:22
|
|
