|
Re: point to point file transfer This is your once-a-decade reminder that "magic-wormhole" and it's single binary equivalent* "wormhole-william" exists *magic-wormhole is tragically written in python Edit: alternate which I've used in the past is schollz/croc which is my favored one because the binary is named with 4 chars Hadlock fucked around with this message at 23:10 on Jan 4, 2024 |
#
?
Jan 4, 2024 22:58
|
|
|
|
| # ? Jun 7, 2024 07:34 |
|
|
acksplode posted:It's still way better than email I hear coworkers say this about non-chatroom things they prefer doing through Slack and it makes zero sense to me.
|
|
#
?
Jan 5, 2024 00:12
|
|
|
say what you will about Zulip the service (I never used it), but the team managed to sell it to Dropbox before launch, got Dropbox to open source it, then spun up a new company to operate and monetize the open source project. gotta respect the hustle
|
|
#
?
Jan 5, 2024 00:46
|
|
|
I wish I could use Teams. We're using Webex and it's such a piece of poo poo.
|
|
#
?
Jan 5, 2024 01:41
|
|
|
Wooooow I remember using that... In 2014 Kudos to your Cisco sales guy
|
|
#
?
Jan 5, 2024 09:14
|
|
|
thotsky posted:We're using Webex goondolences
|
|
#
?
Jan 5, 2024 14:29
|
|
|
Huzzah. My corporate health insurance's prescription plan now has a deductible.
|
|
#
?
Jan 5, 2024 21:37
|
|
|
thotsky posted:I wish I could use Teams. We're using Webex and it's such a piece of poo poo. Oh my god
|
|
#
?
Jan 8, 2024 03:15
|
|
|
Harriet Carker posted:I still like Slack. What do you all dislike about it? The recent technical decisions in their product redesigns point at an obvious play to take on Notion in the structured content space, except that product doesn't actually exist so you're just left with a bunch of weird baggage
|
|
#
?
Jan 8, 2024 17:46
|
|
|
Hadlock posted:We had "Zulip" at one company and there were at least three different mutinies trying to switch over to slack. Including a formal vote where slack won "but not by enough". Eventually the product team got slack and over time engineering all managed to get slack licenses and we all migrated over leaving the VP of engineering and his group of cronies (which he never hesitated to mention, were all in his wedding) over there Hadlock posted:It's like 11 years old at this point. Some people use it i loving loved zulip when i was going to recurse center lol, it's the perfect application for piling a ton of people talking about wildly different things into it. RC still uses it and i pop in there to talk with the other alumni from time to time. but yeah the search is a nightmare it was also the first company i ever applied for a job at, back before they got acquihired and the product was open sourced. they brought me up to boston to interview and i was like "yep i just do javascript and frontend things, you have a web app, i'd love to make a cool web app that's cutting edge" and then failed the interview because they asked me questions about sql and shell scripting, because the whole team was a bunch of MIT people who's previous startup was live kernel patching, so that's all they knew how to interview about. i'm not saying that i would have managed to make a UI that was competitive with slack, because it's hard to get people to wrap their heads around threads, but also in retrospect i think the thread system was way better than what slack/discord have now. at least there's no loving sidebars in zulip, and you can have a single screen that shows you all of your messages without being overwhelming. take boat posted:say what you will about Zulip the service (I never used it), but the team managed to sell it to Dropbox before launch, got Dropbox to open source it, then spun up a new company to operate and monetize the open source project. gotta respect the hustle absolutely, for as bad as they were at interviewing and building a successful company they were all genuinely really nice people. i'm glad they managed to exit and open source and continue supporting the app so the handful of long term users could stick with. honestly the best end for a startup imaginable
|
|
#
?
Jan 10, 2024 02:17
|
|
|
I'm a moderately old, moderately underpaid embedded SW/HW guy. I've got an offer all but in hand from a 10 person start up with an innovative idea for a type of sensor. They're a startup so they're still struggling on market and implementation, but it could do some good stuff in some relatively specialized markets. They've got over a year of runway, and expect to have more once some stuff underway closes, this is mostly from SBIR work. This is Boston, so its not a classic startup with VC and a board, its got advisors, partners, and contracts. The place I'm at now is getting a bit dull. We were bought a while ago, and the mother ship is mainly mech E, and my group is fitting in more and more awkwardly. I've had to explain Git a couple of times, and people are really concerned about getting the right solidworks models of our PCBs. I really like the idea of working on product development and strategically growing features, but for a variety of reasons that's been pushed aside in favor of contract work, and it gets very MVP, how much exactly will this cost. The startup seems interesting, but there's a chance I'll be there for a bit and be 50 and unemployed. I could make a final push to turn my current job into the job I want, but that is very much an uphill fight and I'd need someone else in a position to do the more customer focused parts of product development which I don't want to do. Or I could keep searching, but I really hate job searching. Any thoughts? E: Salary at the new place would be comparable, but I could probably get an upside if something good happens, we haven't gotten that far yet.
|
|
#
?
Jan 16, 2024 02:10
|
|
|
Are you in the Boston area? Would you have to relocate? The tech job market there is OK and probably not as bad a place to be an older programmer as some of the bigger markets, but the city is stupidly expensive and the lifestyle might be worse than what you have now.
|
|
#
?
Jan 16, 2024 02:34
|
|
|
StumblyWumbly posted:I'm a moderately old, moderately underpaid embedded SW/HW guy. I've got an offer all but in hand from a 10 person start up with an innovative idea for a type of sensor. They're a startup so they're still struggling on market and implementation, but it could do some good stuff in some relatively specialized markets. They've got over a year of runway, and expect to have more once some stuff underway closes, this is mostly from SBIR work. This is Boston, so its not a classic startup with VC and a board, its got advisors, partners, and contracts. i think that if you can wangle your way into the mit orbit sbir world that would be a really good choice if you're already in the area. idk if it seems like it's worth moving for reasons upthread
|
|
#
?
Jan 16, 2024 02:46
|
|
|
Yeah, I'm in Boston already. I definitely see a lot of jobs I could do but I don't have a great network and getting call backs is not easy. I'd also rather not work in medical or defense, so that ends up limiting me. E: This is not an MIT startup, nor Harvard, but the IP does come from a solid school
|
|
#
?
Jan 16, 2024 02:47
|
|
|
sbirs are not all defense and the ones that are defense are not all weapons systems, depending on how loosely you define "defense". i've honestly frequently kind of wished to work at an sbir shop, it seems really fun and fresh
|
|
#
?
Jan 16, 2024 02:59
|
|
|
Yeah, I'm not that anti-defense, I'm fine working on soldier support stuff, but I'd rather not go to Raytheon and work on missile guidance systems. My company used to do SBIRs before we were bought. It can be neat stuff, coming up with different ways to handle random situations like "How to clean uniforms without water." It's often a big consulting distraction, but occasionally you can really win a lottery ticket. It does sound like this company has SBIRs that are well aligned with their intended product.
|
|
#
?
Jan 16, 2024 03:39
|
|
|
Phobeste posted:sbirs are not all defense and the ones that are defense are not all weapons systems, depending on how loosely you define "defense". i've honestly frequently kind of wished to work at an sbir shop, it seems really fun and fresh it was kind of a grind when i did it. always new projects around the corner, but we were also booked to 3-5 projects at once.
|
|
#
?
Jan 16, 2024 03:39
|
|
|
the pay was also like 60% of what you can make in games, which itself isnt notably high for tech jobs
|
|
#
?
Jan 16, 2024 03:40
|
|
|
StumblyWumbly posted:I'm a moderately old, moderately underpaid embedded SW/HW guy. I've got an offer all but in hand from a 10 person start up with an innovative idea for a type of sensor. They're a startup so they're still struggling on market and implementation, but it could do some good stuff in some relatively specialized markets. They've got over a year of runway, and expect to have more once I'm not extremely bullish on startups right now. Money is ludicrously expensive compared to what it was two years ago. It's going to get slightly better but not a lot Is this the founders first start up or have they founded a company previously that also got funding? Investors won't touch first time founders hardly at all right now TL;DR triple check your level of confidence on runway extension
|
|
#
?
Jan 16, 2024 10:28
|
|
|
StumblyWumbly posted:Yeah, I'm in Boston already. I definitely see a lot of jobs I could do but I don't have a great network and getting call backs is not easy. Working at a startup that goes under isn't the worst way to bootstrap a network. Make some friends along the way, and soon you've got folks all over who are happy to refer you.
|
|
#
?
Jan 16, 2024 11:24
|
|
|
StumblyWumbly posted:Any thoughts? For a phase I SBIR, you're looking at maybe $150K funding for a 9 month program. Proposal preparation effort is a sunk cost whether you win or not, so any winning proposals need to soak up the expense of preparing the losing ones. Any presentations for status reports and final submission are additional expenses coming out of that funding. The remaining funds may fund one engineer for about 6 months or so to do a feasibility study or a rough prototype. These programs are a loss-leader for the organization, since you need to win and do well on a phase I program to have a chance at winning a phase II. For a phase II SBIR, you're looking at maybe $750K for a two year program. Better than a phase I, but the bar for deliverables is much higher. Do well on this, though, and you could win one of the lucrative phase III programs where you can get millions in funding and a guaranteed government market for the product. Phase III programs are quite rare, though. Before considering doing work for a SBIR-funded company, be sure to ask: 1. How many phase I/II SBIR programs will I be expected to work on simultaneously? 2. Will I be working on writing SBIR proposals? If so, in what capacity? 3. What is your submitted-to-won ratio for phase I proposals? 4. What is your success rate in turning won phase I programs into won phase II programs? 5. Have you ever won a phase III program? 6. Will the principal investigator (PI) for a program (usually a PhD) be a project manager or a researcher? Note that if you are involved in the SBIR proposal writing process your Christmas and New Years holidays will be ruined every single year by the proposal crunch period. If a company has a pile of phase I programs and few-to-no phase II programs, that is a warning sign that you'll be overworked and constantly context switching. PhD PIs without practical industry experience will often work you into the ground because academics often have a poor sense of work-life balance.
|
|
#
?
Jan 16, 2024 19:08
|
|
|
Ralith posted:Working at a startup that goes under isn't the worst way to bootstrap a network. Make some friends along the way, and soon you've got folks all over who are happy to refer you.
|
|
#
?
Jan 16, 2024 20:26
|
|
|
As I've gotten older, I've increasingly come to value work-life-balance more than the precise nature of whatever project I happen to be working on.
|
|
#
?
Jan 16, 2024 20:49
|
|
|
The new head of engineering (6 days in) did a Q&A session today. He demonstrated a reasonable understanding of the what's wrong but appears to have no plan what to do about it. He seems open to the larger changes we need, but doesn't seem to be interested in driving any sort of change himself. How screwed is this situation? We're ~100 engineers probably looking to IPO late 2024 to 2025. If it makes a differences the head of sales was also just exited from the organization.
|
|
#
?
Jan 17, 2024 00:33
|
|
|
Sign posted:The new head of engineering (6 days in) did a Q&A session today. He demonstrated a reasonable understanding of the what's wrong but appears to have no plan what to do about it. He seems open to the larger changes we need, but doesn't seem to be interested in driving any sort of change himself. How screwed is this situation? We're ~100 engineers probably looking to IPO late 2024 to 2025. If it makes a differences the head of sales was also just exited from the organization. he may need to work on resolving problems outside the engineering group before directly resolving problems within.
|
|
#
?
Jan 17, 2024 00:42
|
|
|
It may be a bit much to expect a new director to have concrete action plans after only 6 days. That's still in the "trying to figure out what's actually going on and how far my leash extends" phase. Don't make promises you can't be sure you'll keep, in other words.
|
|
#
?
Jan 17, 2024 00:50
|
|
|
It's going to take at least until the end of Q1 to have any kind of political capital to do what he/she needs to do If management wanted results in 6 days they would have hired a management consultant
|
|
#
?
Jan 17, 2024 01:00
|
|
|
I uhh got sidetracked seeing how far I could push github copilot. Sort of my litmus test since LLMs came out was to be able to render a 3d wireframe cube and then move a live camera around it. Periodically they update the LLM behind various services so it's worth checking back every 3-6 months and see how far it's gotten.Today I accomplished this task. https://pastebin.com/tQDscVd6 This paste has instructions on how create a (rust) cargo project and what to add to the cargo.toml (only the crate "minifb") and the source code. I did not write any of this code, I just gave it the prompt in an empty main.rs file quote:using minifb draw a 3d wireframe cube made from pimitive white lines as seen from a viewport or camera. the background should be blue. rotate the camera slowly around the centerpoint of the cube, 45 degrees above the cube. the lines should be 1 pixel wide. do not use anti-aliasing.  I kept adding features like moving the camera in 4 dimensions plus a pan feature, and you can move the cube in 6 DOF + roll/yaw/pitch. It is about 400 lines at which point copilot stopped writing all the code for me and started suggesting smaller changes. But yeah other than setting the camera position from 0 to 180, 100% of this was written by copilot. caveats: it's writing to a framebuffer so there's a lot of flicker, and there's no error handling for when objects move out of the field of view
|
|
#
?
Jan 17, 2024 07:46
|
|
|
Sign posted:The new head of engineering (6 days in) did a Q&A session today. He demonstrated a reasonable understanding of the what's wrong but appears to have no plan what to do about it. He seems open to the larger changes we need, but doesn't seem to be interested in driving any sort of change himself. How screwed is this situation? We're ~100 engineers probably looking to IPO late 2024 to 2025. If it makes a differences the head of sales was also just exited from the organization. I agree with what others have said, there's no indication you're screwed. Give them enough time to validate all the information they're getting, from what sounds like a good chunk of people, before making changes. I'd personally be more worried if after only 6 days they were acting as though they knew exactly what everyone needed to do.
|
|
#
?
Jan 17, 2024 12:36
|
|
|
Hadlock posted:I uhh got sidetracked seeing how far I could push github copilot. Sort of my litmus test since LLMs came out was to be able to render a 3d wireframe cube and then move a live camera around it. Periodically they update the LLM behind various services so it's worth checking back every 3-6 months and see how far it's gotten.Today I accomplished this task. You miss primitive in your prompt, see if fixing that improves things at all
|
|
#
?
Jan 17, 2024 16:42
|
|
|
Sign posted:The new head of engineering (6 days in) did a Q&A session today. He demonstrated a reasonable understanding of the what's wrong but appears to have no plan what to do about it. He seems open to the larger changes we need, but doesn't seem to be interested in driving any sort of change himself. How screwed is this situation? We're ~100 engineers probably looking to IPO late 2024 to 2025. If it makes a differences the head of sales was also just exited from the organization. Figure I owe you all an update. Yesterday he decided no service can be deployed to production until they have their monitoring dashboard approved by him. So there are now ~100 engineers stuck waiting for him to approve dashboards about topics he doesn't understand. So far there has been no information/feedback about what he is expecting on those dashboards. And the best part is we've got several services with messed up shared owner models that now have like 5 teams trying to coordinate on how to monitor them to make a singular dashboard. The VP of customer success is also on their way out.
|
|
#
?
Feb 2, 2024 18:39
|
|
|
that was quick, at least
|
|
#
?
Feb 2, 2024 18:51
|
|
|
That seems like a pretty effective way to understand the whole stack in short order Is he working with the head of tech operations on this or is he personally doing this Two VP level departures (who presumably have fat stacks of options) in short order at a company planning to IPO in 300-600 days isn't confidence inspiring
|
|
#
?
Feb 2, 2024 19:46
|
|
|
Sign posted:Figure I owe you all an update. Yesterday he decided no service can be deployed to production until they have their monitoring dashboard approved by him. So there are now ~100 engineers stuck waiting for him to approve dashboards about topics he doesn't understand. So far there has been no information/feedback about what he is expecting on those dashboards. devils advocate, this shouldn't be difficult and if it is, its worth figuring out why. let the dude cook.
|
|
#
?
Feb 2, 2024 20:04
|
|
|
If you're talking uptime, performance, message processing delays etc sure fine but even those could have product implications that someone brand new probably doesn't grok yet. Going through a runbook to understand things like that takes time.
|
|
#
?
Feb 2, 2024 20:10
|
|
|
Sign posted:Figure I owe you all an update. Yesterday he decided no service can be deployed to production until they have their monitoring dashboard approved by him. So there are now ~100 engineers stuck waiting for him to approve dashboards about topics he doesn't understand. So far there has been no information/feedback about what he is expecting on those dashboards. This sounds like the company has been running way, way too hot for a long time. If a team of 100 is deadlocked and completely blocked in 2 weeks time, then you're already running way too fast and loose for enterprise software. If he wants to micromanage deploys, then sure, he's not going to last long. If he's just trying to slow down the company while y'all get some DORA metrics in place to know that your poo poo is actually working, sounds like a good way to do that.
|
|
#
?
Feb 2, 2024 20:40
|
|
|
Hadlock posted:That seems like a pretty effective way to understand the whole stack in short order AFAIK he's doing it himself, but we don't have a singular head of operations those 3 teams EMs all report to one of his reports along with a big chunk of other stuff. And I'd call it a third VP counting the guy he replaced. It is extra silly for my team since we don't own any services and can't do anything else because of this. And the company wide on site is all of next week.
|
|
#
?
Feb 2, 2024 20:45
|
|
|
Bruegels Fuckbooks posted:devils advocate, this shouldn't be difficult and if it is, its worth figuring out why. let the dude cook. If you currently don't have monitoring dashboards or they're worthless then yeah that's something you need to fix, but it's not something you can bang out in an hour and blocking all deploys company-wide until it's done is a terrible idea.
|
|
#
?
Feb 2, 2024 21:20
|
|
|
Sign posted:It is extra silly for my team since we don't own any services and can't do anything else because of this. And the company wide on site is all of next week. How much runway do you guys have And how up to date is your resume
|
|
#
?
Feb 2, 2024 22:27
|
|
|
|
| # ? Jun 7, 2024 07:34 |
|
|
Hadlock posted:How much runway do you guys have Good bit of runway year+. Resume is reasonably up to date. I get lots of random inbound stuff for jobs I'm not interested in despite not looking which considering what it seems like the hiring environment is at this point.
|
|
#
?
Feb 3, 2024 00:32
|
|