Research Based on the .NET Runtime (2019)
mattwarren.org
Research Based on the .NET Runtime (2019)
1–10 of 13 posts
Re: Research Based on the .NET Runtime (2019)
#2umm
Re: Research Based on the .NET Runtime (2019)
#3The research papers on manual memory management for .NET looked interesting. I wonder if anything ever came of it?
Re: Research Based on the .NET Runtime (2019)
#4The research papers on manual memory management for .NET looked interesting. I wonder if anything ever came of it?
The delete/free bits never made it in, but some related stuff definitely has made its way into the standard library and compiler over time. Span and Memory provide a bunch of affordances for safely mixing GC and non-GC pointers in the same workflow. Big chunks of the BCL can operate on strings and data at any address in memory now, which is really helpful for high-throughput parsing and streaming.
Re: Research Based on the .NET Runtime (2019)
#5The research papers on manual memory management for .NET looked interesting. I wonder if anything ever came of it?
Is there even a way to allow for manual memory management (not just invoking the GC) in a manner that would still be safe that won’t defeat the purpose of a managed language?
Re: Research Based on the .NET Runtime (2019)
#6The research papers on manual memory management for .NET looked interesting. I wonder if anything ever came of it?
The delete/free bits never made it in, but some related stuff definitely has made its way into the standard library and compiler over time. Span and Memory provide a bunch of affordances for safely mixing GC and non-GC pointers in the same workflow. Big chunks of the BCL can operate on strings and data at any address in memory now, which is really helpful for high-throughput parsing and streaming.
What's an affordance? Is it something with a technical meaning, like pragma?
Re: Research Based on the .NET Runtime (2019)
#7Earlier quoted context omitted.
The delete/free bits never made it in, but some related stuff definitely has made its way into the standard library and compiler over time. Span and Memory provide a bunch of affordances for safely mixing GC and non-GC pointers in the same workflow. Big chunks of the BCL can operate on strings and data at any address in memory now, which is really helpful for high-throughput parsing and streaming.
What's an affordance? Is it something with a technical meaning, like pragma?
On technical side, I think GP meant stuff like this https://docs.microsoft.com/en-us/dotnet/api/system.span-1.-c...
And this: https://docs.microsoft.com/en-us/dotnet/csharp/language-refe...
Re: Research Based on the .NET Runtime (2019)
#8The research papers on manual memory management for .NET looked interesting. I wonder if anything ever came of it?
Is there even a way to allow for manual memory management (not just invoking the GC) in a manner that would still be safe that won’t defeat the purpose of a managed language?
Re: Research Based on the .NET Runtime (2019)
#9Earlier quoted context omitted.
What's an affordance? Is it something with a technical meaning, like pragma?
On technical side, I think GP meant stuff like this https://docs.microsoft.com/en-us/dotnet/api/system.span-1.-c... And this: https://docs.microsoft.com/en-us/dotnet/csharp/language-refe...
I know about spans and stackalloc, I mean quite literally does the word affordance mean something technical here that I've not come across?
Re: Research Based on the .NET Runtime (2019)
#10Earlier quoted context omitted.
On technical side, I think GP meant stuff like this https://docs.microsoft.com/en-us/dotnet/api/system.span-1.-c... And this: https://docs.microsoft.com/en-us/dotnet/csharp/language-refe...
I know about spans and stackalloc, I mean quite literally does the word affordance mean something technical here that I've not come across?
I don’t know what you came across and what you didn’t so I can’t answer that question.
Technically, spans support affected many parts of modern .NET, both standard library, runtime and even language.