|
Boiled Water posted:lol this sounds awful i'll take "what is athena for 1000 alex"
|
#
?
Jul 4, 2018 13:06
|
|
|
|
| # ? Jan 22, 2025 11:46 |
|
|
Blinkz0rz posted:i'll take "what is athena for 1000 alex" now that looks neat i wonder how much more expensive it is than just a sql server or is it on top of an already existing sql server? 
|
|
#
?
Jul 4, 2018 13:29
|
|
|
Blinkz0rz posted:i'll take "what is athena for 1000 alex" Is this AWS's answer to bigquery? Or is that redshift? I'm not familiar with AWS these days. I would definitely ask why they felt the need to create their own querying engine on top of S3, maybe it predates a lot of these new services and it'll be a bunch of legacy maintenance crap, or maybe there really is some niche need it fills, which could be kind of cool. Either way asking "why not this" makes you look good so long as you aren't a dick about it. Boiled Water posted:or is it on top of an already existing sql server? The data's being served from S3 so its probably for huge analytics loads vs more transactional stuff like relational databases.
|
|
#
?
Jul 4, 2018 13:35
|
|
|
god help you if the reason it exists is 'our athena queries were running longer than 30 minutes and timing out'
|
|
#
?
Jul 4, 2018 13:48
|
|
|
as far as i understood it, it was the realtime backend for some kind of advertising stats tool. he mentioned it was "cool for a startup to be able to throw 1000 cores at a query for two seconds"
|
|
#
?
Jul 4, 2018 13:51
|
|
|
TheFluff posted:some kind of advertising stats tool Whoops nevermind adtech is boring
|
|
#
?
Jul 4, 2018 13:53
|
|
|
Boiled Water posted:now that looks neat it's very very expensive and not at all a 1-1 match for a rdbms. it's designed to query structured s3 objects and is both as slow and as brittle as you'd expect the only time i've used it was ad hoc querying cloudtrail logs which already has a schema definition available
|
|
#
?
Jul 4, 2018 14:06
|
|
|
ADINSX posted:Whoops nevermind adtech is boring it is really easy and pays well though so its a pretty good job if you have good hobbies
|
|
#
?
Jul 4, 2018 14:08
|
|
|
Blinkz0rz posted:it's very very expensive and not at all a 1-1 match for a rdbms. it's designed to query structured s3 objects and is both as slow and as brittle as you'd expect that sounds horrible
|
|
#
?
Jul 4, 2018 15:09
|
|
|
Boiled Water posted:that sounds horrible it is not having to learn a new, proprietary query language is nice tho
|
|
#
?
Jul 4, 2018 15:22
|
|
|
$5 per TB of data scanned
|
|
#
?
Jul 4, 2018 20:48
|
|
|
Edit: nvm
qhat fucked around with this message at 21:06 on Jul 4, 2018 |
|
#
?
Jul 4, 2018 20:53
|
|
|
Athena is good if you have a few hundo tb of data in s3 buckets and you don’t want to reinvent the wheel on querying that data
|
|
#
?
Jul 4, 2018 22:37
|
|
|
we've used athena a lot at work and (compared to our in-house unmaintained hive/presto thing) it is super needs-suiting for us!
|
|
#
?
Jul 4, 2018 23:52
|
|
|
athena is great. it's just presto that you pay for per byte instead of per second. at a prior job we cut our spend by 70% by switching from presto on emr to athena queries
|
|
#
?
Jul 5, 2018 02:42
|
|
|
Just did one of those "programming" puzzles that was a mix of figuring out poo poo was in base64, sql injections, fixing bugs, etc to get the next answer. One of the steps was to send a hex value via PATCH to a url. The value it wanted was '0xBAD'
|
|
#
?
Jul 5, 2018 13:00
|
|
|
TimWinter posted:Becoming a dad makes me think so much less of my own capacity to be responsible or make decisions when I was single, in a relationship, or DINKing it.  for all of this and the illegality of the question they'd have to ask is sort of  . It's like a signalling thing (at least it was for me in my mid-20s, now probably not)?
Schadenboner fucked around with this message at 13:08 on Jul 5, 2018 |
|
#
?
Jul 5, 2018 13:05
|
|
|
Kilometres Davis posted:Just did one of those "programming" puzzles that was a mix of figuring out poo poo was in base64, sql injections, fixing bugs, etc to get the next answer. One of the steps was to send a hex value via PATCH to a url. dsyp
|
|
#
?
Jul 5, 2018 13:10
|
|
|
Kilometres Davis posted:Just did one of those "programming" puzzles that was a mix of figuring out poo poo was in base64, sql injections, fixing bugs, etc to get the next answer. One of the steps was to send a hex value via PATCH to a url. Speaking of subtle signalling, is this the sort of thing a company has three weeks to blow on because they are doing well, or because there is no business to keep them occupied? Hmm. edit: I say as my google search opens up a giant maw and presents me a multi-hour coding test to see if I'm good enough for google.
|
|
#
?
Jul 5, 2018 16:01
|
|
|
TimWinter posted:Speaking of subtle signalling, is this the sort of thing a company has three weeks to blow on because they are doing well, or because there is no business to keep them occupied? Hmm. have you been on the other end of the interviewing cycle? it's an unending stream of idiots who have no idea what they're doing with resumes that you would plausibly want to hire
|
|
#
?
Jul 6, 2018 00:35
|
|
|
Yeah there's a lot of bad to wade through to find decent people
|
|
#
?
Jul 6, 2018 00:55
|
|
|
yeah ok, except if that was true, all of us itt would pass every interview we ever take with flying colors, because we're decent at our jobs right? yet i bet you have all bombed an interview, and that day you were, or at least appeared to be, one of the "idiots who have no idea what they're doing with resumes that you would plausibly want to hire". on the other hand, many common interview questions can be studied for and faked. so exactly what signal do we suppose we're getting here, and how do we separate it from the noise?
|
|
#
?
Jul 6, 2018 02:24
|
|
|
DELETE CASCADE posted:yeah ok, except if that was true, all of us itt would pass every interview we ever take with flying colors, because we're decent at our jobs right? yet i bet you have all bombed an interview, and that day you were, or at least appeared to be, one of the "idiots who have no idea what they're doing with resumes that you would plausibly want to hire". on the other hand, many common interview questions can be studied for and faked. so exactly what signal do we suppose we're getting here, and how do we separate it from the noise? it's cheaper in the long run to be selective and pass on qualified candidates than it is to take a chance and hire someone that's terrible.
|
|
#
?
Jul 6, 2018 03:41
|
|
|
DELETE CASCADE posted:because we're decent at our jobs right?
|
|
#
?
Jul 6, 2018 04:02
|
|
|
Blinkz0rz posted:it's cheaper in the long run to be selective and pass on qualified candidates than it is to take a chance and hire someone that's terrible.  a bad hire should be easy to gently manage out or isolate where they can't do any real damage until you can turn them into a somewhat okay hire. they cost you a little bit of money and a little bit of productivity (for wasted time onboarding or ongoing training) but that's about it the cost you pay for passing on qualified candidates is unfilled roles and increased time spent hunting unicorns. the productivity cost is massive even if the money-out-the-door cost seems minimal
|
|
#
?
Jul 6, 2018 04:24
|
|
|
i've increasingly become convinced that if your org can't prevent a bad hire from being actively harmful then it's probably also doing a terrible job of maximizing the value of your good hires.
|
|
#
?
Jul 6, 2018 05:06
|
|
|
the talent deficit posted:
bad hire includes weirdos that cause sexual harassment lawsuits that costs way more real dollars than whatever costs come from unfulfilled roles.
|
|
#
?
Jul 6, 2018 05:11
|
|
|
a lot of the stuff you need done doesnt require super productive people to do it reasonably well
|
|
#
?
Jul 6, 2018 05:22
|
|
|
I'm way more productive in my current job than my last one and I'm pretty sure I'm not that different. There's something to organizational efficiency. A lot of something.
|
|
#
?
Jul 6, 2018 09:34
|
|
|
the talent deficit posted:
agree to disagree then my old team hired on 2 objectively bad candidates because they were having trouble with recruiting and felt like they just needed butts in seats. it took around 6 months to finally manage them both out. in those 6 months the team lost 2 other members to unrelated issues and the net effect of replacing 2 good employees with 2 bad ones was a massive drop in morale for the rest of the team to the point that i genuinely believed that the other 3 would leave and the team would fold. being selective about your candidates isn't unicorn hunting it's about making sure that the people you do hire can actually do the work with the team. if that means that you reject candidates you're on the fence about, so be it.
|
|
#
?
Jul 6, 2018 10:50
|
|
|
I wish contract to hire didn't have a stigma around it; we've gotten some really good people out of our contract to hires (they didn't mind the stigma, and/or appreciated the 60 days to evaluate us). I don't know if I'd ever try one though, but ideally it would allow companies to be a little more lax with hiring knowing that worst case they can get rid of them in a month or two at most.
|
|
#
?
Jul 6, 2018 13:29
|
|
|
ADINSX posted:I wish contract to hire didn't have a stigma around it; we've gotten some really good people out of our contract to hires (they didn't mind the stigma, and/or appreciated the 60 days to evaluate us). I don't know if I'd ever try one though, but ideally it would allow companies to be a little more lax with hiring knowing that worst case they can get rid of them in a month or two at most. It's because contract-to-hire is never ever time-limited on the company side (and this is done at no expense to the company), 60 becomes 90 becomes "next quarter" becomes "next FY". You can't expect worker trust when a large segment of the industry's entire business model revolves around denying workers any stability or leverage in the employment relationship.  E: Maybe an hours contract, with a "poo poo or get off the pot" clause whereby the rate goes sharply up after the 60 day hour-equivalent is reached/breached would work? Schadenboner fucked around with this message at 13:36 on Jul 6, 2018 |
|
#
?
Jul 6, 2018 13:34
|
|
|
Schadenboner posted:E: Maybe an hours contract, with a "poo poo or get off the pot" clause whereby the rate goes sharply up after the 60 day hour-equivalent is reached/breached would work? This is the correct way to do it but it also assumes that you have the leverage to make that demand
|
|
#
?
Jul 6, 2018 13:47
|
|
|
Rex-Goliath posted:This is the correct way to do it but it also assumes that you have the leverage to make that demand "Leverage"? Sounds like communism to me, bucko. 
|
|
#
?
Jul 6, 2018 13:48
|
|
|
ADINSX posted:I wish contract to hire didn't have a stigma around it; we've gotten some really good people out of our contract to hires (they didn't mind the stigma, and/or appreciated the 60 days to evaluate us). I don't know if I'd ever try one though, but ideally it would allow companies to be a little more lax with hiring knowing that worst case they can get rid of them in a month or two at most. My current job was this and so was the one before it. Schadenboner posted:It's because contract-to-hire is never ever time-limited on the company side (and this is done at no expense to the company), 60 becomes 90 becomes "next quarter" becomes "next FY". You can't expect worker trust when a large segment of the industry's entire business model revolves around denying workers any stability or leverage in the employment relationship. And this is exactly what happened to me both times. Same with literally everyone else I've met that has contracted. IMO only take a contract if you really need the work.
|
|
#
?
Jul 6, 2018 13:53
|
|
|
Schadenboner posted:It's because contract-to-hire is never ever time-limited on the company side (and this is done at no expense to the company), 60 becomes 90 becomes "next quarter" becomes "next FY". You can't expect worker trust when a large segment of the industry's entire business model revolves around denying workers any stability or leverage in the employment relationship. Oh I guess I never realized that. I figured it was a 60-90 day mutual evaluation period and at the end its either hire or no hire. Like I said, I've never considered it mostly because of stigma, I want the company to hire me as a full employee and I'll still gently caress off after a few months and just have a gap in my resume if the job is lovely.
|
|
#
?
Jul 6, 2018 13:59
|
|
|
ADINSX posted:ideally it would allow companies to be a little more lax with hiring knowing that worst case they can get rid of them in a month or two at most. we do ‘probationary’ periods for new starters - gives new hires six months to their first proper evaluation and if they’re not measuring up contract says we can let them go. ok it’s a bit longer than ‘a month or two at most’ but it still gives you a written-down ‘out’ that you can’t really lawyer against.
|
|
#
?
Jul 6, 2018 14:00
|
|
|
Schadenboner posted:It's because contract-to-hire is never ever time-limited on the company side (and this is done at no expense to the company), 60 becomes 90 becomes "next quarter" becomes "next FY". You can't expect worker trust when a large segment of the industry's entire business model revolves around denying workers any stability or leverage in the employment relationship. my advice is to take the CtH position if it appeals to you (you need a job right now and cant be too choosy, its an opportunity at a company/field you really want to work for/in, youre changing career paths, etc) but to put a strict limit on conversion to full time - if after 60 days (or whatever) theyre still hemming and deflecting about bringing you on full time, restart your job hunt
|
|
#
?
Jul 6, 2018 14:42
|
|
|
Janitor Prime posted:bad hire includes weirdos that cause sexual harassment lawsuits that costs way more real dollars than whatever costs come from unfulfilled roles. okay cool tell me how arduous technical interviews and rejecting candidates who used iteration instead of recursion to solve some problem (or vice versa if you're a c++ or python shop) prevents future sexual harassment cases
|
|
#
?
Jul 6, 2018 14:46
|
|
|
|
| # ? Jan 22, 2025 11:46 |
|
|
Blinkz0rz posted:agree to disagree then you had two open spots and you lost two additional employees in six months and you think your hiring practices were to blame?
|
|
#
?
Jul 6, 2018 14:47
|
|