Earlier quoted context omitted.
Well, besides libraries, performance. .NET supports proper multithreading (no GIL) and it is also much faster than CPython. If IronPython gets ported to .NET Core (maybe it already has, I don't know), then you'd get those benefits for free.
In the public repos, there no substantial new work taking place with IronPython.
Announcing .NET Core 1.0
261–270 of 327 posts
Re: Announcing .NET Core 1.0
#262Earlier quoted context omitted.
I went full circle, from a critic how they cloned Java back when my employer had the privilege to beta test .NET as a Microsoft partner, to someone that enjoys delivering solutions on the .NET stack. For me the sweet spot will be when .NET Native becomes more mature and I can get Delphi/Modula-3 back.
What do you mean by "getting Delphi/Modula-3 back" here? There are some implementations for .NET? or that C# is kinda similar? or something else?
Re: Announcing .NET Core 1.0
#263Earlier quoted context omitted.
The best way to make sure they (we, since I work at Microsoft) get the feedback is by finding out where the product is listening. A lot of products have UserVoice set up, a lot of dev product teams (dotnet, asp.net, etc.) are looking for GitHub issues. For Windows 10 I'd recommend using the Windows 10 Feedback Hub - if you hate something about Windows 10, post feedback in the Feedback Hub before you uninstall it. Vis…
Well I tried that once or twice in the Connect era. But if you say it will make a difference this time, next time I have issues I will contribute. And yeah, that cancer line was meant to snarky and cute, not factual. It does not really matter more than me indicating the obvious, that this is a big culture shift if you look at the timeline. I think MS has ways to go too, and the long game is where we will see how it f…
If you look at any of the projects that are on GitHub now, the public issues list is the issues list. You can see all the dev comments, scheduling via labels, pull requests, code reviews, etc.
https://github.com/aspnet/mvc/issues?q=is%3Aissue+is%3Aclose...
https://github.com/dotnet/coreclr/issues?q=is%3Aissue+is%3Ac...
For products that aren't on GitHub, public issue lists like UserVoice have still been a big improvement (in my opinion) because they usually keep the bugs / issues / feedback around until it's fixed, and the votes accumulate so the important issues keep bubbling up.
Re: Announcing .NET Core 1.0
#264Earlier quoted context omitted.
And 100x quicker than setting the clearly documented environment variable that disables the feature. > You can opt-out of the telemetry feature by setting an environment variable DOTNET_CLI_TELEMETRY_OPTOUT (e.g. export on OS X/Linux, set on Windows) to true (e.g. “true”, 1). Doing this will stop the collection process from running.
It is still opt-out, and thus considered harmful (at least by me)
The way I see it, at the end of the day, the decision for Microsoft is really between not collecting data and an opt-out system. If Microsoft chooses not to collect data, then all developers have to live with tools that improve slowly and have issues (possibly security related that could be maliciously abused) that are not fixed as quickly as they could be.
If Microsoft chooses an opt-out system, they can collect the data they need to make sure their tools are working optimally and as intended. Some developers may not be comfortable sharing how they use Microsoft's tools even with no personally identifiable information collected. These people can opt-out while minimally compromising the quality of the data collected. Additionally, the tools are open source, so any developer that's skeptical of how and what data is being collected by the tools can verify Microsoft's claims.
Those are the two options I see. To me, the cost/benefit of the second option greatly outweighs the cost/benefit of the first for all involved. By not collecting data, security issues that could actually compromise your privacy could go unfixed for longer. By collecting data through and opt-out and open source system, Microsoft can fix issues ASAP and developers can verify that data is collected in way that preserves their own privacy.
It seems like a lot of people are knee-jerking to the idea of collecting data through an opt-out system and not actually weighing the cost/benefit of the realistic options. Can you explain how not collecting data has a lower practical cost/benefit ratio than an opt-out and open source system?
Re: Announcing .NET Core 1.0
#265Re: Announcing .NET Core 1.0
#266Earlier quoted context omitted.
Thanks for the information! Apparently I missed that in the news stream. Sad to hear about that though. I was hoping for a break. Will just have to continue reading that PostgreSQL book I've already started.
I'm starting to learn postgres as well, which book if you don't mind me asking?
Re: Announcing .NET Core 1.0
#267Earlier quoted context omitted.
It is still opt-out, and thus considered harmful (at least by me)
If it were opt-in, they'd collect significantly less data, and the data they do collect would likely by heavily skewed. Microsoft could be misled by the poor quality of data collected, and an opt-in system could actually be worse than not collecting any data in the first place. The way I see it, at the end of the day, the decision for Microsoft is really between not collecting data and an opt-out system. If Microsoft…
Re: Announcing .NET Core 1.0
#268Earlier quoted context omitted.
Well I tried that once or twice in the Connect era. But if you say it will make a difference this time, next time I have issues I will contribute. And yeah, that cancer line was meant to snarky and cute, not factual. It does not really matter more than me indicating the obvious, that this is a big culture shift if you look at the timeline. I think MS has ways to go too, and the long game is where we will see how it f…
Yeah, please try again. The problem with a lot of Connect programs is that they were tied to a specific release (MonkeyManager Pro 2008) so bugs that didn't make the cut for the product release just got lost in space (not picked up for MonkeyManager Pro 2010). Part of the reason was that there were always internal bug tracking systems, which was what the dev team was really working from day to day, and these other sy…
Re: Announcing .NET Core 1.0
#269Any ideas how to do GUI development with .NET Core? Is there a version of Xamarin? Are there plans to port Windows.Forms, WPF? There is this massive thread about it http://forums.dotnetfoundation.org/t/cross-platform-wpf/421/... and an issue here https://github.com/dotnet/corefx/issues/5766
You can have a Web UI in desktop app. Just run the app with embedded kestrel web server and make all UI in HTML. And it will look pretty good an all platforms. CatLight is one of the apps that do that - https://catlight.io
Re: Announcing .NET Core 1.0
#270Earlier quoted context omitted.
What do you mean by "getting Delphi/Modula-3 back" here? There are some implementations for .NET? or that C# is kinda similar? or something else?
Delphi binaries dont have any dependencies, right?
If you using COM or third party dlls, which is quite common on Windows, then you need those as well.