4 ms·
Why would a company choose someone who doesn’t know the stack they are using over someone who does? Let alone 6-12 months. Knowing the ecosystem, best practice
by scarface74 7y ago
Why would a company choose someone who doesn’t know the stack they are using over someone who does? Let alone 6-12 months.
Knowing the ecosystem, best practices, frameworks, etc takes longer than a few weeks.
Sure I could learn Java or Swift in a few weeks, but does that mean I would be a competent Android or iOS developer?
- Retric 7y agoI have made the jump from one language to another several times on the job. It’s considered lower risk to use someone on the team than higher someone new. It’s also going to take a while to higher someone and get them up to speed. After a year you might not be as experienced as someone using the same stack for 5+ years but the difference is not that extreme. Further stacks tend to share quite a bit. Knowing SQL, HTML, JavaScript, CSS, jQuery etc takes a long time and has little to do with Java vs .Net.
- scarface74 7y agoLanguages are easy. But do you think you could jump from systems programming in C to mobile programming on Android “easily”. I’ve seen people jump from Java to C#. You can see the difference in their coding style, them not taking advantage of the features of the language, reinventing the wheel because they didn’t know there were popular packages that would do it for them, creating horrible inefficient queries and practices using EF etc.
- Retric 7y agoSure, jumps like that are not as bad as you might think. Style wise the code is often poor, but users don’t really care about the code. I started with Object Pascal on Mac OS 8/9 which forced you to do a lot of low level tasks like deal with Handles and suffer through a cooperative multithreaded OS. Rewriting the network layer from AppleTalk to TCP/IP was closer to systems programming in C than you might think. I have also been paid to write C programs, Windows application pre .Net in Visual C++ and post in C# and VB, written Java and C# websites, and most recently Angular SPA’s. Add to that a few random oddities like XSLT. PS: I even turned down working on Android, but could have made that jump.
- scarface74 7y agoThe jump is huge. Web development has a completely different model than mobile development. In web development, losing a connection is a failure case. Developing for mobile it is expected. Especially for enterprise apps. Logic is usually distributed, syncing is required, knowing which is the source of truth etc. This is really an example of hiring someone who has experience and hiring someone who thinks “it’s easy and just like what I did before”. Then I could get into all of the old netops guys who think AWS is just like what they did on prem and end up costing the company more...
- Retric 7y agoThey are different, but I don’t think networking is one of the major differences. Website backends often have significant dependence on other services that can be down. Further standalone apps can have zero network dependence or a lot. So having written both, and similar code for each, I don’t feel the networking side is all that different. If anything networking is probably the largest similarly between them. PS: I did some iOS development in my spare time even worked with some old J2ME, so I assume Android is fairly similar.
- rectangletangle 7y ago>In web development, losing a connection is a failure case. That's changing with the advent of progressive web apps. It's possible to write web apps now that are robust in the face of network problems, e.g., the web app renders and is functional even without a connection.
- scarface74 7y agoI’m not questioning whether you can do it with web technology. But you still have to design it where both your business logic and your data can live on the device and it can queue submitted data until there is a connection available and if multiple people are updating the same record, knowing how to merge the information or knowing which one takes precedence. I’ve even seen cases where the logic depended not on just the record, it depended on knowing which fields were updated by the device when it was disconnected and the prior value of the field to make sure that the user had seen the most recent value before they updated it. If not, someone on the back end had to do a “manual merge conflict resolution” by calling both people.
- raverbashing 7y ago> You can see the difference in their coding style, them not taking advantage of the features of the language, reinventing the wheel because they didn’t know there were popular packages that would do it for them, creating horrible inefficient queries and practices using EF etc. That's why code reviews are a mentoring opportunity. Help your colleague level up. (Also there should be team dialogue about how to solve some problems more efficiently, I mean, if I was unsure of something I would go and ask colleagues for some guidance, or you would go and do some research on your own)
- scarface74 7y agoNow not only do you have a less effective developer who takes time to ramp up, you’re also taking time from senior developers. I’m not saying this is always the right course and of course this doesn’t scale to larger companies, but my first project at my current company a week in was to develop a feature from scratch which ended up involving coding an API for the front end developers, writing an ETL process that used both Redshift (AWS OLAP database), and MySQL designing the schemas, configuring the AWS resources with CloudFormation, setting up queues, messages, lambdas, dealing with the vendor we were integrating with and learning the business vertical. How much longer would it have taken if instead of already knowing their stack - C# WebAPI, MySQL, and AWS - I came from a Java/Mongo/GCP background? They needed someone useful now to get a feature out that they wanted to charge customers for.
- LastZactionHero 7y agoI've hired for the stack specifically and it really narrowed our applicant pool. Lately I've relaxed this filter. We're a Rails app, but what if there was a really great engineer that uses Python or Node? They might not know the right libraries to use, but would probably know there are libraries, have a good concept of application design in general, and know how to ask the right questions. And we're hoping to hire someone for years.
- williamdclt 7y ago> Why would a company choose someone who doesn’t know the stack they are using over someone who does? If the two candidates are perfectly identical on every other criteria, sure. But that's not the case, the point is that most other criteria are more important that the specific tech experience when dealing with competent people. > Sure I could learn Java or Swift in a few weeks, but does that mean I would be a competent Android or iOS developer? If you have web experience, I'm not too worried about your productivity on Android or iOS. If I'm hiring for Django and you have experience with any of Symfony, Rails, Spring, .Net or NestJS, the tech expertise is the least of my concerns.
- scarface74 7y agoYou’re not worried about their mobile experience until the app fails to work in a tunnel because they are expecting an always on connection or they have half their records on their mobile and half on their server and need to figure out which to use or they never had to figure out a syncing algorithm because you usually don’t have to sync things on the web since you never expect long periods of not being connected. Edit: and I forgot to mention the biggest UI failure I made when I did mobile development a decade ago on WinCE ruggedized devices. I didn’t even think about actually taking the device out into the sun where are all of the field service techs would be working and seeing how the screen, colors and contrast looked.
- thestepafter 7y agoI would argue that this isn’t solely the responsibility of the programmer. This is where testing and QA comes in. Also, these are pretty basic considerations when building mobile apps and I would argue that they aren’t language specifics.
- scarface74 7y agoSo now we are going to spend more on development and probably rework because the developer decided he was just going to wrap the website in a webview and call it a “mobile app” because he didn’t think through these scenarios. Anytime that you have to go through the development -> QA cycle more then once it costs money and time. That’s kind of my point. Learning a language is easy and for the most part useless without knowing the frameworks and architectural best practices. I don’t know what Android has, but iOS has built in frameworks for handling syncing. If an Android developer didn’t know all of the built in frameworks available to them and re-invented the wheel, that would also be a waste of money.
- SpicyLemonZest 7y agoIt’s like saying “well, we’re an agile company, so we prefer to hire people who know standups and sprint planning”. In principle they will ramp up faster, sure. But it’s just not a useful thing to select on, and if you do select on it you’ll end up with a weird monoculture.
- scarface74 7y agoKnowing an ecosystem well on a senior level won’t happen in a few weeks. But to your example, I’ve seen developers who couldn’t adjust to developing where they had a rapid release cycle because they were use to the big design up front. How you develop software where you don’t have all of the requirements for the next year is a completely different mindset. Even on comments on this post, I see people who aren’t actually willing to actually talk to the customer to decide what they should work on.
- SpicyLemonZest 7y agoI agree the problems you're describing are real and important to select for, but just requiring agile experience won't help you there. Lots of people practice a weird form of agile development, where they go through all the motions of fast iteration, but almost all tasks consist of non-negotiable internal dependencies and not value delivered to a customer. I'd argue a similar thing is true for tech stacks. If there's some correlation between knowledge of all the fiddly bits of C++ and ability to write clean, performant systems-level code, I've yet to see it.
- Aeolun 7y ago> almost all tasks consist of non-negotiable internal dependencies and not value delivered to a customer It’s especially nice when you realize that as soon as you’ve completed all the non-negotiable features, a bunch of other things magically becomes non-negotiable.
- virgilp 7y agoReally senior level is about talking to people; “Knowing the ecosystem well” (if by that you mean a specific tech stack) is almost irrelevant. Senior is more about solving the right problem than solving the problem “right”
- candybar 7y ago> Why would a company choose someone who doesn’t know the stack they are using over someone who does? Let alone 6-12 months. > Knowing the ecosystem, best practices, frameworks, etc takes longer than a few weeks. > Sure I could learn Java or Swift in a few weeks, but does that mean I would be a competent Android or iOS developer? I would separate this into two separate categories - big tech companies (or others with similarly large shared internal infra) and others. For big tech companies, for most positions, the tech stack is proprietary internal stuff such that knowing the language and best practices only get you about 10% of the way. For other types of companies, it's generally more important that you hire people with the right business context than the exact tech stack. With that said, native mobile development isn't just a stack - it's more of a different, though overlapping, career path - the main reason not to hire non-mobile developers into a mobile role isn't that the stack is different and takes time to learn, but that the workflow is so different that they may or may not know what it is that they are even signing up for. Hypothetically, you'd rather hire someone with Xamarin background with no Java experience for an Android java role, than someone with no mobile dev experience, but lots of Java backend experience. My first mobile development experience was a moderately complex Android project on an app that was used by tens of millions of people daily. There was no ramp-up - I had zero prior experience before signing up for this project, never even played around with any mobile development before and I had never professionally programmed in Java - and I was the sole engineer working on both mobile and backend. It was a little painful but everything shipped on time.
- abtinf 7y ago> It was a little painful but everything shipped on time. Your last sentence would seem to contradict the entire body of your comment. You describe mobile development as a fundamentally different thing, but then your first gig was to work on a large project, the result of which was shipping on time at the cost of a “little” pain.
- candybar 7y agoThat's fair - what I meant by it being a different career isn't that skills aren't transferable at all but that you solve different kinds of problems entirely, as distinct from a difference in stack. And if you hired me then as a mobile developer, I'd have quit, so you don't want to hire non mobile engineers into mobile roles. Language/stack is a red herring here in the sense that Java backend development is a lot closer to Ruby backend development than it is to Android development.
- booleandilemma 7y agoAgreed. The job doesn’t exist to train you.
- scarface74 7y agoIt’s more nuanced than that. Most job reqs have “must haves” and “nice to haves”. My current job had one must have - C#/MVC/Web API and some Javascript experience - their current tech stack. Nice to haves were React, the fiddly bits of AWS, and Python. I was immediately useful because I was strong in the must haves, had a little AWS experience, and knew nothing about Python or React. I’ve since leveled up on the nice to haves except for React, I refuse to jump on the $cool_kids bandwagon of front end development. Especially seeing that everything else I listed, pays more, and doesn’t change as often.
- WiseWeasel 7y agoI find React's relatively simple architecture makes it easy to get into for JS devs; I've always found it extremely productive and maintainable. My sickness is to the point that it’s hard for me to imagine an immediate future where it hasn’t entirely consumed the front-end web environment like jQuery did 15 or so years ago. Maybe you're onto something, but I'm loving the new digs so far.
- scarface74 7y agoI don’t necessarily have anything against React as a technology, but as the old saying goes, a man with one ass can’t dance at two weddings. Almost everyone else at the company are better at the front end than I am but most haven’t taken the time to learn the backend. Why would I learn something, that I still wouldn’t be as good at in a year and couldn’t negotiate a higher pay based on bringing an above average amount of value to the company than other employees? I also wouldn’t be as competitive in the overall job market. This isn’t directed toward you, just a general comment.
- coldtea 7y ago>Why would a company choose someone who doesn’t know the stack they are using over someone who does? Because a stack defined programmer is a limited programmer...
- scarface74 7y agoAs a person who was responsible for hiring in a previous role and I still sit on a lot of technical interviews, I didn’t care if you were “limited” to the stack we were using as long as you could get the job done without six months to a year of training. If you started your career learning Java 20 years ago and you kept up with the latest trends of Java, you could still find a job now. The same is true for C# around 15 years ago.
- nscalf 7y agoBecause someone is smarter or better at the job in general that the lower quality specialists they can find. It's very hard to find a good generalist. If you can be a good generalist, you're exponentially more valuable than a generic specialist. I have done mostly backend work in Python, but in my current role I've worked in the frontend with Angular, I've debugged critical issues on our iOS app, I've fixed our AWS infrastructure, I've moved us away from dragging files into windows boxes to CI/CD, I've helped architect solutions, I've worked with our customers, I've led efforts across multiple teams, etc. Being a "Python Programmer" in no way measures my value to the company, in fact, it necessarily limits what I'm actually capable of contributing to the company. If you can show that you can add value outside of a specific tech stack, you're worth the few weeks to learn a new technology. And yeah, it is basically a few weeks to get up to speed and be a contributing member of a code base if you don't know the technology. 6-12 months to know all of the details if you're a solid engineer.
- scarface74 7y agoI would say just the opposite. When it comes to the entire “web developer” tech stack, sure I can put together a website good enough for internal use, but I’m not very good at the front end. I consider myself to be very good at C#, okay at backend JS development and passable at Python. I have enough experience from doing a lot of ETL in a previous life to know how to automize queries and schemas for speed and to not lock up a database. I’ve set up CI/CD solutions from scratch with what is now called “Azure Devops”, AWS’s CodeBuild/CodeDeploy/CodePipeline, OctopusDeploy and Jenkins I could just as easily and competitively apply for jobs as an “AWS Architect” who knows most of the popular AWS offerings for developers, Devops, netops, and system administrators and I have experience with them. But in many of those areas - especially on the front end and with the netops, system administration stuff, at any scale. It wouldn’t make any sense to hire someone who “kind of” knows what they are doing over hiring a specialist. If I need something now I’m not going to want to wait for you to get up to speed in year. Do you really think you’re as good at any of those areas as a specialist? AWS alone announces dozens of new things every month on their podcast.
- nscalf 7y ago