7 ms·
I don't think engineers use frameworks because they think it makes them look cool. They may choose between multiple, competing frameworks based on community per
by karaterobot 7y ago
I don't think engineers use frameworks because they think it makes them look cool. They may choose between multiple, competing frameworks based on community perception, but not between using a framework and not using one.
It's usually an argument from business needs that leads to application bloat. For example, what PM on a startup would sign off on spending two weeks to implement a custom date library and date picker component when there are excellent, if sizable options available to install right now? What small- or medium-sized company is going to create an in-house application framework when React is very good, and there are thousands of people you can hire who already know how to use it? A major reason applications get bloated is because development time costs the company more than client bandwidth and CPU do.
So, in that sense, it's not a matter of convenience either: It's a selection made based on weighing the constraints and options and selecting the best one. If poor application performance started costing companies a lot of revenue, I'm sure the result of the calculation would change.
- flukus 7y ago> A major reason applications get bloated is because development time costs the company more than client bandwidth and CPU do. If development time was really the biggest issue then I think we'd be seeing far fewer SPAs and complicated frameworks along with many more drag and drop VB apps or simple cgi sites, especially on internal tooling for small-medium companies. These also generally perform much better. But between developers wanting to work on more exciting things and PHBs wanting sexier looking apps we end up with the current nightmare.
- flomo 7y ago> I don't think engineers use frameworks because they think it makes them look cool. I've been in this line of work for a long time, and I'll admit its a huge factor. Webdev is a very fad-driven business, and I've seen numerous technologies come-and-go where they were hot, but not they are now super not. And to some degree this is logical, because you want people who can adapt to whatever is latest-and-greatest and not just a particular platform. But when recruiting it certainly helps. Plus the old chestnut that 50%-80% of software projects are failures, and looking out for number one means resume-oriented development for the next job. You are probably better off with a failed project on a 'cool' framework than a successful project on an uncool one.
- mathgladiator 7y ago> Webdev is a very fad-driven business I think it is like this because it is kind of awful on so many dimensions, and we have hacked the browser to do all sorts of things. It's kind of amazing, but now we have to deal with it.
- thinkingemote 7y agoI've talked to several really enthusiastic developer advocates about technology in the corridor and when it comes to describing the actual product or use of the technology they tend to mumble, or wave hands, or state some product that HN would argue about (e.g. advertising). All the developer advocates I've met who promote latest tech really don't seem to like their companys product. I would say that developers try to make it a fad to keep themselves interested and engaged in an otherwise boring, soul destroying or socially harmful job.
- gdev20 7y agoDon't forget that a dev may need another job in the future. And it doesn't help finding one when spending the last 2 years building an app with some "oldschool" tech stack that performs well. Most companies are looking for dev's who worked with the cool fancy stuff like react because that's already the tech stack they are using. And what is so bad trying to make my job interesting? If it's boring I wouldn't care much about the end result. Just another 9 to 5 job then.
- thinkingemote 7y agoNothing bad about making something interesting. You would be interested in the end result - but for developers who do this the end result is the technology they produce and not the product that the tech is being used for. I think it would be one of the reasons why "putting devs before users" happens. Also - look at the HN Who's Hiring threads. Almost all postings don't talk about the product or their users. They talk about the technology.
- 7y ago
- tracer4201 7y ago> It's a selection made based on weighing the constraints and options and selecting the best one. If poor application performance started costing companies a lot of revenue, I'm sure the result of the calculation would change. This is not often the case, where even senior level engineers at FAANGs end up selecting tools based on their personal tastes with complete disregard for long term maintenance and operational costs. I’ve seen projects with 6 different LISPY languages to solve the most trivial problem, including a couple developed in house. I’ve seen all kinds of complexity from people choosing a random design pattern and using it to solve literally every problem. Recently I came across a code base where every call to every other class uses the visitor pattern... just because. I’ve seen teams building internal platforms that picked one of most unknown languages for their customer teams to write business logic in... only because the senior engineer is an expert. These folks want other teams who have no background in this language — literally their main customer — to write production quality code, understand its nuances and write idiomatic code. You’re placing too much faith in PMs or any of these management functions having a worthwhile role technical decisions, especially ones they don’t understand and don’t have time to disambiguate. I’ve been at three different FAANGs where teams are literally drowning in idiotic decisions of some clown devs before them who invented technical complexity solely for the sake of getting promoted and then moved onto other teams. It’s not just devs... I’ve seen engineering managers sit comfortable and let people make dumbass decisions. They celebrate the new launch with beers or champagne but then leave within a couple months after because they know operational costs and escalations meant they’d get an unlubed pounding with a 2x4. And it’s not just at FAANGs - I commented on a Ask HN thread a few days ago where the poster asked what’s the best stack for web development. Without a single blink, every other post was some random framework the commenter likes working with - never mind any nuanced discussion of functional or non functional requirements or even beginning to think about support or organizational alignment.
- the_gipsy 7y agoI can attest to this problem. I like to call it, the "architect" who never sticks around. First, this guy gets rewarded for getting a project rolling and showing results as quick as possible. Second, he will get no reward (nor satisfaction) for sticking around. The company also seeks to reproduce point one, long term results are drowned. Rinse and repeat.
- stef25 7y ago> I don't think engineers use frameworks because they think it makes them look cool. Or, management tells engineers what tools to use because they think it makes them look cool. Our company recently got word from above that "all these old technologies go out the window, you're now all becoming JS devs and we'll use microservices for everything". The result is that for one project I could probably build entirely by myself with a LAMP stack in one week, we now have 15+ people in 3 teams & 3 different countries each working on tiny parts of a very small simple app barely used by anyone, using a whole shopping list of tech that you simply don't need. When I asked one of those management "evangelist" guys during an informal dinner "so why all of a sudden JS frameworks?" his answer was something honestly along the lines of "I don't know it seems the thing to do nowadays". Needless to say I'm job hunting like my life depends on it.
- DayDollar 7y agoBuisness schools need a new lecture: Management as a Liability: Tech debt, tech insolvency and what one can do to prevent it. I worked in automation for quite a while now- and seriously, i would be glad by now if management would push for new technology. The tech here is stagnant since the 80s. They start debates when you talk about object orientation.
- OGWhales 7y agoI recently graduated and majored in CIS as well as CS, and the CIS classes were in the business school. This is something we talked about a good deal in one of my CIS classes. Unfortunately, it was a specific major class, not something all business majors took. I am unsure if normal management majors were taught similar ideas but I thought the class I took was rather well done.
- shishy 7y agoYeah, have experienced this also, and I think there are competing interests between "best engineering option to get the product out fastest/most efficiently" and things like "hiring" and "getting investment funding" Using the latest and greatest buzzwords helps greatly with hiring and getting investment (it's a positive feedback loop that just feeds itself and increasingly adds pressure to make those decisions).
- JohnFen 7y ago> I don't think engineers use frameworks because they think it makes them look cool. They don't. They use frameworks because it reduces the cost of production. Unfortunately, they accomplish this at the cost of the user experience.
- edanm 7y agoYou say unfortunately, I say "lowering the cost of production means lower costs to consumers, and/or more valuable stuff getting created". Everything is a trade-off.
- JohnFen 7y agoI say unfortunately because the product quality is reduced as a result. So yes, we have may have more stuff, but most of it is seriously substandard. This may be a positive thing for some people. It's a negative thing for me, though, because it increases the difficulty of finding good solutions, and reduces the number of good solutions that exist.
- 0xEFF 7y ago> Unfortunately, they accomplish this at the cost of the user experience. This doesn't follow. At business startup frameworks provide a huge boost in user experience through code re-use.
- JohnFen 7y agoPerhaps it provides a boost in the initial development. But the applications that I've seen and used that use frameworks almost always provide an inferior user experience when compared to those that don't.