Hacker News Reader

Comment by chabad360 | Hacker News Reader

Fragmentation is almost never a good thing and it's certainly not a good thing in this case. Allowing developers to use the same knowledge, the same interfaces, the se languages across systems means that consumers benefit. Having a de facto standard is good for everyone, and arguing that there is less innovation in programming because you don't own the rights to function names is ridiculous.

View on HN

Seems like it could be compared to breaking up a monopoly. When there's one player in town (Java & Android in the two examples I used), any competing ideas never get a chance to succeed: we're still using variations of Java, probably at this point for historical reasons, and Purism etc. is having a real rough go of it. Look at the failed attempts at competing OSes that various phone manufacturers put out. There were some duds, but I mentioned two that could have given Andriod a run for its money if there were more fragmentation in the market - and I think that would have been better for consumers.

> arguing that there is less innovation in programming because you don't own the rights to function names is ridiculous

I applaud the rhetorical flourish, but I think it misses the point. It's not the cause of the fragmentation, but the fragmentation caused by the lost rights that improves the innovation.

Let's take your proposal further and let's say we have 20 different languages (some even with non English characters), runtimes, compilers, build systems, UI guidelines, etc will uniform distribution across the world. So even if you want to publish a software for your home country you'll only be able to reach 5% of population with any single tool chain.

Are you willing to learn all those 20 different language/runtime/tool chains?

> Are you willing to learn all those 20 different language/runtime/tool chains?

No, and that's exactly my point. Because it would be impossible to support everything, developers and corporations might have to aim for more limited reach. I imagine that would lead to a situation where we have 20 different widely supported kernels, 20 different up-to-date web browsers, 20 different modern OSes, etc. I would love to be able to choose from such a diversity of software alternatives - wouldn't you?

I don’t see this resulting in 20 different web browsers or kernels or OSes that are anywhere near as good as the ones we have today. Making even just one is a shocking investment of resources and a massive ongoing maintenance responsibility. Many, many human lifetimes have gone into building something like Chrome. Wishing that this herculean effort was forced to be duplicated 19 times over just so that you could have alternatives to choose feels kinda perverse.

Talking about the complexity of modern software might be a tangent. However.

I think that if time and effort were limited by this fragmentation, we would find that we're able to do just fine with browsers that have fewer hours of effort put into them. The current range of second-tier browsers have a ton of strange and interesting ideas about them, but because the web standards they have to support are so complex they simply can't support much of the modern web. I imagine that those standards would be written much more conservatively if the writers knew that it would be impractical for anyone to implement them.

> Wishing that this herculean effort was forced to be duplicated 19 times over just so that you could have alternatives to choose feels kinda perverse.

There is absolutely no reason for you to get personal.

Having too many equally dominant options for computing platforms might sound good at first but in practice it means effectively no software is portable and the whole industry becomes inefficient as a whole with people doing redundant work most of the time.

Reply on news.ycombinator.com