Live data from Hacker News

Visual Studio for Mac

msdn.microsoft.com

381–390 of 435 posts

Re: Visual Studio for Mac

#382

Earlier quoted context omitted.

The ram you were using 3 years ago is a lot slower than the ram they have put into the new MacBook pros. So there is that benefit at least. 32 GB is a lot of ram. What do you do where you need that much ram in a laptop? Is that the limiting factor in performance for you versus another component? Have you considered a desktop/tablet combo or anything like that?

There are a lot of use-cases for >16GB of RAM, but I'll share mine specifically. I develop network appliances for a "medium-to-large" enterprise. Specifically, these network appliances provide BGP, stateful packet filtering, and the other network services provided by our company's products. When working on these appliances, I tend to spawn hundreds(close to 1000, but not more on my laptop due to RAM) of VM's, and eac…

Yeah I can see that (though keep in mind that this is a very fringe case). But on the other hand I'm not complaining that my MacBook can't run any game available at 60 fps.

If I want to do that, I get a desktop. I think that has been a common theme for a long time. Power - desktop. Portability - laptop. We're asking Apple to make laptops as powerful as desktops. It just won't happen, unfortunately.

Re: Visual Studio for Mac

#383
post #298

Earlier quoted context omitted.

> IIRC the explanation is really "this is what Intel's stuff is supporting, and we're stuck with what Intel supports" All of the available CPUs support 32 GB; the i7s in the 15" model even support 64 GB. [0] http://ark.intel.com/products/91156/Intel-Core-i5-6360U-Proc... [1] http://ark.intel.com/products/88967/Intel-Core-i7-6700HQ-Pro...

They support > 16GB if you use DDR4. LPDDR3 is limited to 16GB. Given that Apple maximises for long battery life/power efficiency/thin-ness (where heat = bad), I think they made the right decision. There are very few people who genuinely need more than 16GB of RAM in a laptop computer. Would I get 32GB if it was available? Yes. It would make my work a little easier (multiple VM environments), but it's hardly the end…

Yeah. I mean look, they could design a laptop for the vast majority of their users, or they could design one that is best for <1% if their users. It's a clear choice.

Re: Visual Studio for Mac

#384
post #371

Earlier quoted context omitted.

Do most developers really need anything with a CPU more powerful than an iPhone 5 to build apps or web pages? In most cases if they do they're doing it wrong.

It helps. No reason to artificially limit processing power available. If I spend seconds more waiting for my Go command line program to compile, or Rails to boot up, or a complex web page to load, or my VMs to respond, or Elasticsearch to boot etc. etc. it all adds up. It's time that I can spend doing other stuff.

True, but we were all using most of that same technology when a core2 duo was a top end laptop, and it's not as though productivity has gone through the roof just because we can get an i7 laptop.

Mobile CPUs are approach mobile core 2 duo performance.

Re: Visual Studio for Mac

#385
post #63

Earlier quoted context omitted.

Apple is not saying anything like this, just a bunch of whiners twisting things. Just like they did after any Mac upgrade.

> WRONG. The removal of a dozen keys from an already gimped keyboard is decidedly anti-developer. I am a developer and I could not care less about function keys. Why the duck would developers need function keys? Even for Vim, the age-old advice is to remap Esc so that you keep your hands on the home row. A flexible multi-touch strip of context-aware keys can do much more things -- e.g. map debugger step moves when I'…

"Why the duck would developers need function keys"

Are you serious? They need them THE MOST. F10 = step over F11 = step into Shift-F11 step out

Standard VS debugging keys. And on the asinine Mac laptop keyboard, you could make F12 = Delete (or at least you could until Sierra, when Apple broke Karabiner)

Then there are other apps that you might want to fill glaring functionality holes in. Like Photoshop, which still has no "paste as new image" command. So you have to record a macro and assign it to a key, like F2. And for some reason you can't transform a selection rectangle after you draw it, so map F3 to allow that. And on and on.

Think it through. That's no a strip OF KEYS. It's a non-palpable strip that changes its contents willy-nilly and forces you to look down at it every few seconds. Your examples are pretty weak strawmen.

Re: Visual Studio for Mac

#386

Earlier quoted context omitted.

Do most developers really need anything with a CPU more powerful than an iPhone 5 to build apps or web pages? In most cases if they do they're doing it wrong.

yes. it's not needed if you're building a JavaScript app with Sublime text. But compiled environments such as .Net take a lot of cpu and memory. It's not uncommon to have 2-4 visual studios open which can take up 10 gigs of ram. Couple that up with continuous build/unit test frameworks and 5-10 VS add-ons and you need a powerful machine to make it responsive

I'd argue that some of what you describe should probably be an anti-pattern, or is at least a byproduct of very fast, cheap CPU power. Emacs has worked for years on much lower-end hardware and has been used to build some very elaborate systems.

These days a typical developer laptop is 5x more powerful than the system the code will be used on (mobile device, virtual server, etc.)

Re: Visual Studio for Mac

#387

Earlier quoted context omitted.

Do most developers really need anything with a CPU more powerful than an iPhone 5 to build apps or web pages? In most cases if they do they're doing it wrong.

Until you need to work with something that requires babel/gulp/etc builds, it's fine. The second either is added to the project, you will need a much more powerful computer.

True those are slow. But I think most projects could compile a lot faster if the dependencies were compiled into a separate bundle once a day and the app code into its own bundle when it changed.

Re: Visual Studio for Mac

#388

Earlier quoted context omitted.

I just started looking into Xamarin Forms and couldn't figure if it was just blindly typing xaml or there was a way to preview it. Thanks for the posts.

Yeah, for now you have to type the XAML but a designer is coming. There is a preview of a previewer but I couldn't get the one in VS to do anything. The XS previewer did work for me though. Make sure Android Studio / ADK is updated if only iOS is working.

I've only tried in VS so far. I'll check XS. Thanks.

Re: Visual Studio for Mac

#389
post #73

Most interesting part for me: """For the functional programmers among you, it includes excellent F# support, powered by the same F# compiler used in Visual Studio.""" I've heard that F# is great from multiple people I trust a lot (and a quick cross check showed it does indeed look very cool) so I might give it a try once this is released. I do some C# development (Unity Engine stuff) on my Powerbook so this is also g…

OCaml guys don't seem to like F#, as its implementation differs from many in the FP world believe is "correct", but any form of FP on .NET is a good thing, I think.

Re: Visual Studio for Mac

#390

Earlier quoted context omitted.

Lately, my main machine has been a mid-2014 mbp with 16GB of RAM and I recently purchased the new mbp with 8GB of RAM. Comparing them side by side is a very strange experience. _Technically_ it should be slower but it doesn't "feel" slower. However, at times, there is a bit of shutter that I can't put my finger on - am I just looking for an excuse to say its slow or was the loading time on opening this project always…

Thanks, that's a helpful review. It has me wondering if it might benefit from the hardware improvements enough that swapping would be noticeably faster.

I'm thinking along the same lines. If you're interested in other tests/simulations and I can run them side by side and report back. Today I've ran 2 VMs, about 20 tabs, a large project loaded into an IDE and two shell sessions without any noticeable slowdown.
Post reply on HN