Think in Go: Go's alternative to the multiple-inheritance mindset.

https://groups.google.com/forum/#!msg/golang-nuts/Rj5t1h_ztxI/Fgs6HfLqMHAJ

by peregrine • 15 years ago

92 26 15 years ago

groups.google.com

ggchappell• 15 years ago

This is an interesting discussion -- interesting in part because it addresses a mindset that I don't seem to have.

C++, for example, has multiple inheritance, and I have used it to good effect. OTOH, if I needed to implement Dijkstra's Alg in C++, then I don't think it would even occur to me to inherit from both a priority queue and a directed graph base class. (And even now that it does occur to me, it doesn't seem to be a good idea.) Rather, I'd use something akin to the interfaces presented here. Of course, interfaces are not formalized in C++, but the C++ culture does include the notion of a concept, to be used with a template.

If the language were Python, I'd proceed similarly. And then there's Java, which doesn't even have multiple inheritance. I've never heard anyone claim that Java is somehow ill-suited for the implementation of graph algorithms (but maybe it is -- never thought about it).

Perhaps what this discussion is overturning, is not so much a multiple-inheritance mindset, as an Eiffel mindset. (But I don't know much Eiffel, so it's hard to say.) And the discussion also suggests that Go's formalization of interfaces can be a Good Thing.

_delirium• 15 years ago

I also don't know much Eiffel, but I believe this 1988 article is the canonical pro-multiple-inheritance-in-Eiffel piece (by Eiffel's creator): http://archive.eiffel.com/doc/manuals/technology/bmarticles/...

sorbits• 15 years ago

I think there are many who doesn’t split types and algorithms, and something like Java does/did make it cumbersome plus its standard library doesn’t do it — but the C++ comity adopted STL into the standard library 13 years ago, which is all about separating algorithms from their types, and was actually born out of Alexander Stepanov’s work with graph algoritms, so I was also a little surprised by his generalizations and specifically calling out C++ (twice).

supersillyus• 15 years ago

Does what he's describing as the Go Way basically equate to generic programming in general, or is there something specific to Go about it?

Rusky3• 15 years ago

It's sort of Go-specific, but the same style is also used in Haskell using typeclasses, Scala/Ruby/etc using mixins/traits, and C++ using templates.

Rather than inheriting from some base class, you just make sure your class can do certain operations and then other stuff works on that- it's good way to decouple algorithms, containers, etc. from the types they work with.

dochtman• 15 years ago

It's also often described as duck typing...

caffeine2• 15 years ago

The comments in his code are reminiscent of programming by contract (an idea I believe Djikstra himself was fond of). In Go the types of the interfaces are statically checked, which is good (in Python or Ruby, this isn't the case), but the contract the interface represents has to remain in comments and unit tests.

It would be nice to be able to specify tests that every implementor of the Heap interface must pass?

discreteevent• 15 years ago

Alan Kay talks about field effects but maybe what he really means is that "everything is an object" is a mistake. Not everything is an object in real life. For example, a sound wave is not an object but objects can react to it (or it can manipulate them, whichever you prefer). Maybe emphasizing this in a programming language can be done as simply as you have described it is done in go. ( I have always found objects expressive and useful but when overused they feel like a straitjacket)

pointyhat5• 15 years ago

This requires me to sign in to a Google account (I don't have one). Bring back usenet!

dchest• 15 years ago

http://pastebin.com/iPibD66C

pointyhat5• 15 years ago

Thank you - much appreciated :)

bane3• 15 years ago

They're free, and it takes ~45 seconds to get one. You should probably just make one.

alnayyir2• 15 years ago

I'm sure they know how to get one, and whatever reason they have for not getting one, you've not addressed.

Given those circumstances, refrain from responding if you aren't going to be genuinely helpful.

Nobody likes a comment from the peanut gallery.

delinka• 15 years ago

Slightly off-topic: the web in general seems broken for mobile. To get this content on my iPhone, I had to concoct this URL:

https://groups.google.com/forum/?nomobile=true#!msg/golang-n...

Then, I discovered that to see more content, I had to swipe up with two fingers. There were no scroll indicators. This and the scaling problem ("you cannot rescale this web site. We have decided for you that this size is perfect. And no scrolling to see the words that run off the edge of the screen.") are...well, problems.

p9idf• 15 years ago

As with all of their other products, Google is in the process of breaking the Groups UI by adding useless frames and whitespace. Try this instead:

http://groups.google.com/group/golang-nuts/msg/7030eaf21d3a0...

zanst• 15 years ago

Russ, Ken, Rob and the other guys are making a great work in Go. It's weird that people are not yet paying much attention to it - and to Plan 9 as well.

Think in Go: Go's alternative to the multiple-inheritance mindset.

Related Stories