Hacker News Reader

Comment by philliphaydon | Hacker News Reader

Do people get annoyed by all the JavaScript frameworks and Databases coming out in regards to adoption from a company point of view? I mean every other day a new database comes out and claims to be better in one way or another than something else and then its like "fuck I picked X when now there's Y"

It seems over the last year technology has been growing more rapidly than any other period.

Fun times but so hard to keep track of everything!

View on HN

For those old timers who did distributed systems work there's not that much new under the sun. I look at something like this and say "ah, a quirky and somewhat dangerous cache layer." What's different is bloggers promoting it as a "database."

While I'm sure someone out there will see this and say "wow, that's exactly what I need!" chances are that if you have these sorts of scale issues you're going to have to figure it out on your own.

I'd rather see a write-up of how they arrived at this particular conclusion than another non-database.

I have found the way in the middle is to stick to the traditional things for the most critical work, things where the abilities and problems are well understood, until some of the new thing mature. Then, for less critical things that can run as separate services, try out the new things. That way people in a team can also get exposure to the new technologies and you can experience first hand the pros and cons of these technologies. Some people are more generally optimistic about new things which solves some of the frustrations they have with the old things (or perhaps their lack of experience and understanding of the old things), without realising there will be new problems that comes with the new solutions that haven't even been identified yet, never mind workarounds or solutions to these new, and yet unknown, problems. However, I think the biggest, while also avoidable, risk is usually adopting the new thing for a problem you don't even have. A lot of distributed databases exist because you are going to have a lot of concurrent users access similar data, related to themselves mostly, accessed at the same time, or have a lot of user data that you want to do big, ad-hoc searches and analysis across all the data. If you don't have the scale of 'a lot' or 'concurrent' or 'across all the data' that these things are designed to address, you might not even need this solution and especially not all the tradeoffs you have to make along with it. I agree it's an exciting time, and a time of discovery for everyone, but there will be many things learned the hard way and it's tricky to position yourself here especially when there's pressure around you.

While I understand that feeling, I've come to realize it's an inevitability, that shouldn't affect your work.

At any given time, you either have a need/problem, or you don't. If you DO, you evaluate the current tech available, and hopefully select something that fits your needs. You build out around said tech, and if your choice was correct, that means it's either solving your problem, or on it's way to.

If something comes along while you're implementing with your chosen solution, that looks similar, but better, it's only noise - because hey, you found a solution.

Just as we don't all re-write all of our code whenever a new language comes along (unless the thing in question was desperately in need of a re-write anyway) even if newer languages are nicer, we needn't switch DBs or frameworks for the same reasons.

Yes, I find it annoying because:

1) it's non-stop

2) there seldom sems to be anything truly novel in a broadly meaningful way (i.e. esoteric, if anything)

3) there is rarely an objective improvement on existing options

I no longer feel compelled to replace or adopt though, precisely for those reasons.

On the contrary, I find it awesome. Having choices is much better than not. It just shows the maturity of the entire tech ecosystem. Just take everything with a grain of salt, evaluate if any new tool/library matches your particular need/use case, do a quick proof of concept if so, and adapt if successful. Rinse and repeat. Change is good. Embrace it.

> It seems over the last year technology has been growing more rapidly than any other period.

I tend to agree with this statement. The entire stack appears to be going through a revolution.

The data layer in particular is seeing very rapid change after being largely (not entirely) static for decades.

Reply on news.ycombinator.com