|
TodPunk posted:I can't speak for the whole US, but for the western half, Rails is mostly a startup language. If you are not willing to have most of your pay be in stock options, you probably don't want to go that route. Though the contracting bits are accurate, I don't think they're as prevalent here. Most of the contracts I've seen on the likes of Dice for Washington and Utah have been maintenance/integration pieces. Getting picked up quickly though, that's very true if you're in the right area. In Utah valley (around Salt Lake City), there are a LOT of startups, so you can hop quite a bit. Thanks for this. I've been considering pushing for visa positions in the US and it's good to know what sort of places want these skills.
|
#
?
Apr 22, 2014 13:09
|
|
|
|
| # ? Apr 24, 2024 15:40 |
|
|
TodPunk posted:I can't speak for the whole US, but for the western half, Rails is mostly a startup language. If you are not willing to have most of your pay be in stock options, you probably don't want to go that route. Uh yeah I'm pretty sure this is false.
|
|
#
?
Apr 22, 2014 16:18
|
|
|
shrughes posted:Uh yeah I'm pretty sure this is false.
|
|
#
?
Apr 22, 2014 17:30
|
|
|
shrughes posted:Uh yeah I'm pretty sure this is false. I thought it was just a generalization, never presented as some sort of "fact." I'm assuming you mean the piece about taking stock options, as opposed to being a startup language (It's not even a decade old, how do you see it as an established production business tech?) Obviously your anecdotes will be different from mine (care to share?), and how you would generalize the startup world is going to be different as well. Hell, I've worked for startups most of my career and I only took stock options once (hahaha they're worthless now), so even in my own experience it's not "true." That doesn't mean I didn't turn down a lot of jobs paying peanuts + stock. Still, we make generalizations to communicate this poo poo as usefully as possible. Java is verbose. Obviously not true, but if you look at most Java code, you're going to run into a hell of a lot of verbosity, especially on the job. Can you make a lot of money as a musician? Sure, but you'd be silly giving that advice to someone asking, because it's not the norm.
|
|
#
?
Apr 23, 2014 02:49
|
|
|
TodPunk posted:I thought it was just a generalization, never presented as some sort of "fact." I'm assuming you mean the piece about taking stock options, as opposed to being a startup language (It's not even a decade old, how do you see it as an established production business tech?) I mean the piece about it being a startup language. It's not. You'll find it quite prevalent outside of startups.
|
|
#
?
Apr 23, 2014 05:27
|
|
|
shrughes posted:I mean the piece about it being a startup language. It's not. You'll find it quite prevalent outside of startups. If you don't have code left over from the 1990's or before, you're clearly a startup 
|
|
#
?
Apr 23, 2014 06:06
|
|
|
TodPunk posted:(It's not even a decade old, how do you see it as an established production business tech?) An established business tech is a tech used by an established business, doesn't really matter how old it is. When you said "a startup tech" I was a little surprised because in the UK it's popular with startups, but there's multi-million pound business using it who've long since passed on that title.
|
|
#
?
Apr 23, 2014 08:19
|
|
|
Ruby is the language. Rails is not a language. Ruby has existed since 1995 Rails has existed since July 2004 (9 years, ~9 months) No one is going to get fired for suggesting Rails as their web framework. So I would consider that pretty stable.
|
|
#
?
Apr 24, 2014 02:11
|
|
|
Strong Sauce posted:Ruby is the language. Linux is not an OS.  To be fair, while you are totally correct. Ruby deserves the close association with Rails, because that's arguably the framework the catapulted it into the mainstream. Definitely not a startup-only thing, but after a certain size it won't be good enough to avoid a polyglot architecture or JVM layer (i.e. JRuby).
|
|
#
?
Apr 24, 2014 03:08
|
|
|
Doctor w-rw-rw- posted:Definitely not a startup-only thing, but after a certain size it won't be good enough to avoid a polyglot architecture or JVM layer (i.e. JRuby). What do you consider "a certain size"? GOV.UK gets around 8 million hits a week with tens of millions more on the way, and that's entirely MRI Ruby. https://www.gov.uk/performance/site-activity If you're talking about hitting Rails directly, yeah that's not going to scale very well, but put most anything behind a good caching layer and you're fine.
|
|
#
?
Apr 24, 2014 11:38
|
|
|
Arachnamus posted:GOV.UK gets around 8 million hits a week with tens of millions more on the way, and that's entirely MRI Ruby. https://www.gov.uk/performance/site-activity Looks like they also use Python to me: https://github.com/alphagov
|
|
#
?
Apr 24, 2014 17:06
|
|
|
Doctor w-rw-rw- posted:GNU/Linux is the OS. There was more to the post, where I also had "Java has existed since 1995, Spring has existed since 2003" Since Spring was the first framework for Java that came to my mind and matched my narrative but decided to delete it because I honestly don't know if the comparison was fair other than the fact that Spring is only one year older than Rails and would probably count as being for businesses. I realize now removing it made me sound like the Ruby/Rails thing was a huge deal. But IMO if you're going to make a statement about it, you should at least get that part correct. Rails gets a bit of a bad rap because Twitter moved over to the JVM. I think that's a bit unfair given how much traffic Twitter gets over almost anyone else. GitHub runs on Rails and it works just fine. It seems the switch away from Rails happened because of an internal change in direction. Given Twitter's resources they could have contributed more to Rails and/or Ruby. Ruby 2.x is starting to see a lot more speed gains, but imagine if Twitter had people helping out with Ruby. Maybe they did and decided it wasn't worth the trouble. I've heard Ruby is insanely hard to parse correctly. Wish I had that blog post where some guy tried to make a ruby parse then said, "Nope!" after going through and finding all the one-offs you had to account for in parsing.
|
|
#
?
Apr 24, 2014 17:17
|
|
|
Strong Sauce posted:There was more to the post, where I also had "Java has existed since 1995, Spring has existed since 2003" Since Spring was the first framework for Java that came to my mind and matched my narrative but decided to delete it because I honestly don't know if the comparison was fair other than the fact that Spring is only one year older than Rails and would probably count as being for businesses. I realize now removing it made me sound like the Ruby/Rails thing was a huge deal. But IMO if you're going to make a statement about it, you should at least get that part correct. Re: Twitter They did, it is, and they hired some really good Java people (i.e. Marius Eriksen). Twitter decided to use the best tool for them - I don't really think it's any company's responsibility to improve a platform if it isn't the product they're selling. There's goodwill and technical benefits to be gained from doing so, but not an obligation.
|
|
#
?
Apr 24, 2014 18:31
|
|
|
Doctor w-rw-rw- posted:Re: Twitter Isn't that how we end up with issues like Heartbleed though?
|
|
#
?
Apr 24, 2014 19:07
|
|
|
bonds0097 posted:Isn't that how we end up with issues like Heartbleed though?
|
|
#
?
Apr 24, 2014 19:29
|
|
|
Doctor w-rw-rw- posted:Yes, and? It wasn't any company's responsibility to keep OpenSSL secure, it was their responsibility to keep their own products secure. The industry in general failed at that because we failed to recognize the connection between collectively paying attention and money to OpenSSL and securing ourselves. I think that if we treat the open-source world as a communal toolbox that we take from without any responsibility to improve/maintain/secure, the likes of Heartbleed are inevitable. This is ultimately a detriment to both the industry and consumers. Sure, as a consumer you initially benefited from a cheaper license cost for IBM Sametime Server Community Edition since they just popped in OpenSSL and didn't worry about it, but in the aftermath of heartbleed, was that actually a net benefit versus a higher initial cost to help pay for some development time spent securing OpenSSL (or whatever)? I guess I'm just not sure that a 'gently caress you, I got mine' mentality is a good direction for the industry.
|
|
#
?
Apr 24, 2014 19:57
|
|
|
Doctor w-rw-rw- posted:Github - also polyglot. Used Erlang to serve up repos. I'm not sure what you mean by serve up repos. Github uses Erlang, but (from last knowledge) used it to load balance and handle ruby processes. Regardless Rails is still handling the bulk of their front facing traffic. I think you're misconstruing my post a little bit. I didn't say Twitter _had_ to help out ruby. What I meant was, it seems like the people who made the decisions preferred to use the JVM rather than Ruby/Rails. So rather than go the route of helping ruby get faster, they put more resources into converting to using JVM languages.
|
|
#
?
Apr 24, 2014 21:23
|
|
|
Doctor w-rw-rw- posted:Looks like they also use Python to me: https://github.com/alphagov Performance Platform do their own thing because they want to, not because Ruby wouldn't do the job. Bear in mind @alphagov encompases all of the UK's Government Digital Service, not just the GOV.UK website. Almost all content served by GOV.UK (the 8 million weekly hits I mentioned) comes from https://github.com/alphagov/whitehall and https://github.com/alphagov/frontend which are both Rails apps. The point is not that companies use 100% Ruby wall to wall; there are situations where it's not the best tool for the job or you fancy trying something else. Getting a system running at large scale is more about the architecture of the system as a whole allowing multiple applications to work together whatever language or framework they're written in. Once you have your system broken down into conceptual chunks each represented by a robust application, it doesn't matter whether it's Rails or Django or Node because they're small enough to be manageable and scalable and replaceable. The only thing that matters is that they're written in a language and framework suitable for the specific focused task at hand. If that task is serving dynamic webpages, Rails does you fine, even for big websites.
|
|
#
?
Apr 24, 2014 23:59
|
|
|
I never said "Ruby isn't good enough to serve big sites", I said it wasn't good enough to avoid a polyglot architecture, i.e. an architecture that is tolerant of decoupled services, or a service-oriented architecture if you want to use that terminology. Which is really kind of true of nearly everything (Java in particular seems to embrace it IMO). I don't really understand what we're arguing about.Doctor w-rw-rw- posted:Definitely not a startup-only thing, but after a certain size it won't be good enough to avoid a polyglot architecture or JVM layer (i.e. JRuby).
|
|
#
?
Apr 25, 2014 00:28
|
|
|
More money towards OpenSSL would not fix Heartbleed. The issues with OpenSSL are systematic. It's a result of poor maintenance, poor code structure, and a complete lack of testing and proper review and auditing. The maintainers of OpenSSL are both happily employed other companies. What would sending them money even fix?
|
|
#
?
Apr 25, 2014 06:34
|
|
|
Doctor w-rw-rw- posted:I never said "Ruby isn't good enough to serve big sites", I said it wasn't good enough to avoid a polyglot architecture, i.e. an architecture that is tolerant of decoupled services, or a service-oriented architecture if you want to use that terminology. Which is really kind of true of nearly everything (Java in particular seems to embrace it IMO). I don't really understand what we're arguing about. No, me neither. I think we agree. It seemed like you were making out that Rails was unique in that sense, which as you say it isn't, it's the same for everything. There's plenty of cultural issues with Rails, no need to worry about something like scale.
|
|
#
?
Apr 25, 2014 13:32
|
|
|
Doctor w-rw-rw- posted:I never said "Ruby isn't good enough to serve big sites", I said it wasn't good enough to avoid a polyglot architecture, i.e. an architecture that is tolerant of decoupled services, or a service-oriented architecture if you want to use that terminology. Which is really kind of true of nearly everything (Java in particular seems to embrace it IMO). I don't really understand what we're arguing about. Now you have me confused. Why would one wish to avoid an architecture that is tolerant of decoupled services?
|
|
#
?
Apr 25, 2014 15:08
|
|
|
Hughlander posted:Now you have me confused. Why would one wish to avoid an architecture that is tolerant of decoupled services? Deployment should be no more complex than it needs to be, and the longer you can avoid that without ridiculous contortions, the simpler the ops will be, IMO. Two services means two application lifecycles and some sort of pooling or connection management, which in turns means more statefulness, and possibly more ways undeployment+redeployment can go wrong. Basically, a service-oriented architecture (IMO) usually goes from a net negative to a sizable net positive relatively quickly. So if you're starting from nothing and quickly rewriting a fairly conventional prototype into a fairly conventional version 0.1 of a product, engineering a decoupled service is liable to impede early progress, because it's easier to experiment on a newborn project when it lives in a simple world.
|
|
#
?
Apr 25, 2014 19:04
|
|
|
pr0zac posted:So I'm looking to see if anyone has any experience making this kind of transition and what their experience was? Is the management side of things fun and challenging? What are people's opinions on the better decision for long term career outlook? In 5 years am I going to have a harder time finding a job if I move to management? Theres no financial reason to do it or not, my company has good equal promotion tracks for both engineers and managers, so its totally a career style choice. As a coder-turned-management (and now middle management, as a director) I've really enjoyed it. (I worked at Amazon for almost a decade, left and moved to various startups and now work at a large-ish company as a director of a sizeable team that does tools for internal productivity: dev tools, QA frameworks, the intranet, etc.) Just like there's a lot of really terrible technical jobs out there, there's a lot of really terrible management jobs. It can be a lot harder to figure out ahead of time, so I've prioritized networking to find future jobs. I'm at my current job because a friend left his job at Google to go there and be a VP -- it was a surprising move, so we grabbed coffee and he did a very good job recruiting me on their cool technology, great culture, and interesting challenges. A lot of companies do a terrible job of training their leadership, so you get a lot of technical people who are promoted because they are loud and good at technical things. When this percolates high enough up the management chain, you have no one with experience mentoring anyone else, and can end up with a very unpleasant atmosphere for everyone involved. Why I'm happy in management: I like that the challenge changes not just with every new project, as mentioned earlier, but also literally with every person on your team (and everyone above you in the chain, and every peer, and every other person in the organization that you want to influence.) If you're really doing it right, you're tailoring your message (presentation of yourself/team/ideas/goals) to the individual every time. I often quip in interviews that I don't like puzzles but I do like problems. Puzzles are things where there's an answer, you know there's an answer, and your only job is to find that answer. Systems administration has a lot of puzzles. Programming even has too many of them. But problems? Problems you don't even know if they're answerable -- heck, you might not even know the question. And you don't know necessarily what the best answer looks like, and sometimes part of what you have to do is define the criteria for "best answer". Higher levels of technical skill deal with a lot of problems, but even a first level manager's job is full of problems. That ambiguity drives a lot of technical people away, but really attracts people like me. I love that my efforts are multiplied across everyone on my team. If I'm a good leader, my entire team is more productive. I like that my success is judged on the basis of how good I hire and how good I lead: because it's judged on the strength of my team. I can directly influence that, whereas being part of a large software team you can't always have the level of control you need to actually feel like you're producing a good product. (Good) Management is all about "if you win, I win" and I like that. As a gamer I loved pve more than pvp simply because I preferred to be part of something where everyone was building to a common outcome (win or lose) rather than an encounter in which the end game was by definition loss for one party. So getting one of my technical guys up in front of a group of managers is more of a win for me than if I was doing the same presentation. Many first-level management positions expect you to be quite technical and able to code. At a sane company, this means you're still productive while also having the bulk of your duties be leadership. At an insane company, this means you're judged on your ability to produce the same volume as your most senior engineer and yet still be able to lead the entire time. See above about networking before you take a job. Why I'm happy even in middle management: In addition to all of the above, now I get to mentor managers -- which as I said before is really rare and I like being that good boss that actually cares about the careers of my team, that is helping them get more visibility and be promoted. Whenever I can have someone on my team present instead of me, it's a win for us both. My scope is broader, and much more about strategy: What business value am I adding, how do we fit into the big picture, what is coming down the road that we need to anticipate, where is their duplication of effort, etc. The "vision" matters, and I'm responsible for how engaged my teams feel in terms of being able to really make a difference day to day (this is more challenging in an internal team like mine.) If my teams don't have road maps I push them to make them. If my teams aren't presenting their best foot forwards, I work with them on that. I'm working with my entire division now to present what our story is: how do we work together, what unites us other than "our customers are internal" and are we doing everything we can to find places that two teams working together will produce a better product than if they were ignoring each other. I like that I can't just totally neglect the technical bits. I need the respect of my engineers. It's okay that I don't code daily, but I could code, I often am writing helper scripts for myself, and I need to be able to give feedback on designs. Even at my level, I challenge my managers on their team's direction and architecture, though sometimes it's less-technical things like "let's look at the calculations for the cost savings your tool is providing -- oh wait, you aren't even tracking usage? Let's figure out what to track add it to a sprint soon." It's often not just about what you control, but what others control: currently I own internal productivity/automation tools, but I don't own all the tools. I'm still asking my managers to ask their managers to understand where they fit in the total picture of the company. And I expect us to be creating and driving a coherent vision for where "tools" should be in the future, which will involve partnering with other teams and influencing their roadmaps, etc. Cutting it off here because this is getting long, but it's a topic I'm pretty passionate about. We don't have enough good managers out there, if I can help encourage some to take the jump I'd love to. (And selfishly: I'm hiring, so I have skin in the game of there being more good managers out there.) Kyth fucked around with this message at 18:21 on Apr 27, 2014 |
|
#
?
Apr 27, 2014 18:14
|
|
|
Doctor w-rw-rw- posted:Deployment should be no more complex than it needs to be, and the longer you can avoid that without ridiculous contortions, the simpler the ops will be, IMO. Two services means two application lifecycles and some sort of pooling or connection management, which in turns means more statefulness, and possibly more ways undeployment+redeployment can go wrong. Basically, a service-oriented architecture (IMO) usually goes from a net negative to a sizable net positive relatively quickly. So if you're starting from nothing and quickly rewriting a fairly conventional prototype into a fairly conventional version 0.1 of a product, engineering a decoupled service is liable to impede early progress, because it's easier to experiment on a newborn project when it lives in a simple world. Only if you do deployment it in such a way. Oft times I'll start with a SOA architecture where everything lives in a single deployment, whether JAR or executable. The height of which was an MMO I worked on about 12 years ago where you could run the server in any role and the client in a single executable tracing from player input through the network layer, to the server RPG mechanic in a debugger with no external dependencies. The current MMOs I work on have one JAR for every role and it's just what role the machine is configured for on startup that gets loaded. Again developers run all the roles in a single JVM, while production will usually run just 1-2 roles on a single instance.
|
|
#
?
Apr 27, 2014 18:58
|
|
|
DevOps / continuous delivery has been something I've been focused on for the past year or so. I always say "Deploying software is easy. Installing and configuring software is a nightmare." The fun starts when you have to deploy a web application that communicates with several pieces of middleware, and each piece is maintained by a different team with a different release schedule, and everything is backed by the same database where both teams are trying to make concurrent schema changes. Thank god there are decent tools to help with it.
|
|
#
?
Apr 27, 2014 19:27
|
|
|
Hughlander posted:Only if you do deployment it in such a way. Oft times I'll start with a SOA architecture where everything lives in a single deployment, whether JAR or executable. The height of which was an MMO I worked on about 12 years ago where you could run the server in any role and the client in a single executable tracing from player input through the network layer, to the server RPG mechanic in a debugger with no external dependencies. The current MMOs I work on have one JAR for every role and it's just what role the machine is configured for on startup that gets loaded. Again developers run all the roles in a single JVM, while production will usually run just 1-2 roles on a single instance. It's all variations, but for debugging multiple services I infinitely prefer deploying the production environment with config overrides (reduce the number of concurrent instances, turn on a debug port, override other services with your local versions, etc.) as is. Might be overkill when the number of services and applications is known and small, but for working with many external teams its so nice to be able to deploy and vend my own personal environment as if it were just another host in our fleet.
|
|
#
?
Apr 27, 2014 23:44
|
|
|
Ithaqua posted:DevOps / continuous delivery has been something I've been focused on for the past year or so. I always say "Deploying software is easy. Installing and configuring software is a nightmare." The last time I was at a large company (Amazon), these tools didn't exist so I actually ended up creating them (it was running on a quarter million target hosts by the time they started to phase it out for new technology, ~15 years later.) It definitely is easier these days to actually have them available. That said, they're only as good as the teams configuring and running them, and it's definitely possible to do it wrong even with a tool. The smallish company prior to this one tried to bring chef in several times but you really need buy-in from the dev teams to do so, and if they're still stuck in "of course we'll just copy a subset of the directory around to all the machines, what could possibly go wrong?" you struggle.
|
|
#
?
Apr 28, 2014 00:16
|
|
|
Having decided to eventually move on from academic research, it is time to dust off the old Resume. I'd appreciate any critiques - the biggest question in my mind currently being whether to trim to 1 page or expand to a more verbose 2 pages.
|
|
#
?
May 10, 2014 01:09
|
|
|
Go for 2. Add an "intent" section describing your generic ideal job and tailor it to every place you send.
|
|
#
?
May 10, 2014 01:24
|
|
|
Paolomania posted:Having decided to eventually move on from academic research, it is time to dust off the old Resume. I'd appreciate any critiques - the biggest question in my mind currently being whether to trim to 1 page or expand to a more verbose 2 pages. Stick to one. Lots or white space there. Consider a two-column layout, and remember that it's not about writing everything in the resume, it's about getting the resume read. It's okay to leave some information out that you can explain in person. JawnV6 posted:Go for 2. Add an "intent" section describing your generic ideal job and tailor it to every place you send. Doctor w-rw-rw- fucked around with this message at 01:28 on May 10, 2014 |
|
#
?
May 10, 2014 01:24
|
|
|
Yeah, I'd see if you can't buy yourself some space back with two columns on the expertise section at the very least, so as to get everything on one page. Publications might be something I'd consider striking too. It's great street cred in academia but in industry it's something you can just bring up in an interview conversation. Also I always can't help but  at "taint tracking."
|
|
#
?
May 10, 2014 02:01
|
|
|
One page, two page. Who cares? It does not matter. If you have a qualified applicant you're going to read their resume regardless.
|
|
#
?
May 10, 2014 02:05
|
|
|
Strong Sauce posted:One page, two page. Who cares? It does not matter. If you have a qualified applicant you're going to read their resume regardless. Applicants aren't qualified until their resume passes the filter, so saying that he resume will be read in full because its qualification is known a priori is not true.
|
|
#
?
May 10, 2014 02:18
|
|
|
15 years of work experience should be two pages IMO.
|
|
#
?
May 10, 2014 03:11
|
|
|
Hughlander posted:15 years of work experience should be two pages IMO. That is another question: do I include the early part of my career that I spent as an artist/animator or omit it? The really relevant stretch is the 8 years of gigs post-grad school.
|
|
#
?
May 10, 2014 04:44
|
|
|
I don't know why the length thing ever even comes up anymore. I get an email with an attachment. I open the attachment and look at it. Whether it's 1 page or 2 pages or 5 pages doesn't matter. [edit] I think mine is 3, but my current job takes up about 3/4ths of a page on its own because I wear a lot of hats.
|
|
#
?
May 10, 2014 04:56
|
|
|
I think what looks weird about it being two pages now is that the second page is nearly blank. You could either strike unnecessary stuff and better manage space to get it to one page, or flesh it out to use more of the second page.
|
|
#
?
May 10, 2014 05:13
|
|
|
kitten smoothie posted:Also I always can't help but I was at a con once where one of the talks was about taint analysis and taint propagation, taint was also the speaker's drinking word. Hilarity ensued.
|
|
#
?
May 10, 2014 09:11
|
|
|
|
| # ? Apr 24, 2024 15:40 |
|
|
I want to get into embedded systems. I think I am qualified. In general I am a very experienced programmer. I know a lot about programming in assembly w/ several different types of instruction sets, low level optimizations, operating systems, concurrent programming. However, I have never actually done anything with embedded systems, and don't know much about electric engineering. Should I bother applying for embedded systems jobs?
|
|
#
?
May 14, 2014 23:43
|
|
