Archive of UserLand's first discussion group, started October 5, 1998.
Re: C# papers
Author: Ed Draper Posted: 6/26/2000; 9:20:09 PM Topic: C# papers Msg #: 18138 (In response to 18101) Prev/Next: 18137 / 18139
>It seems like they could have acheived 90% of the benefits of >C# v. C++ by writing a good standard garbage collector for VC++, >fully supporting the C++ standard, and getting a good STL >implementation.I guess I'm not clear on what you mean by a "standard garbage collector for VC++." The nature of C++ (and C for that matter) does not lend itself to any generalized garbage collection mechanism. The closest you could get is with a bolt-on library that could front-end a subset of memory management routines, such as overloading new(), but you'd still have to contend with a bunch of memory that was being allocated via different allocation methods. There is no low hanging fruit here. If there was, it would've been already done.
In terms of standards conformance, the holiest of the C++ religious wars, VC++ is not so bad. Version 6.0 scored 93% on Plum-Hall's C++ conformance suite. I've also found their STL implementation to be exceptionally strong.
I think of C# as a "supercharged" Java. Rather than continuing on in their pissing match with Sun, they chose to zig and add value where they know developers will appreciated it - performance and flexibility.
As to your arguments in regards to "why bother switching", I generally agree. I'm still rather puzzled by the legions of developers who jumped on the Java bandwagon.
The area that I think C# will add the most value is in server based web applications. Negotiating the treacherous COM threading models in C++ is tricky business, C# looks like it could be of some assistance here.
Ed
This page was archived on 6/13/2001; 4:55:30 PM.
© Copyright 1998-2001 UserLand Software, Inc.