Timeline for How come compilers are so reliable?
Current License: CC BY-SA 2.5
10 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Apr 19, 2017 at 20:02 | comment | added | Eric Lippert | @aij: This scenario comes to mind of course because the C# 3 compiler had a bug in cases where you had an interface that inherited both a property and a method of the same name, and overload resolution did not correctly report the ambiguity error required by the specification. But since 99.999% of C# programmers will never get into that situation in the first place, and would not know the compiler was wrong if they did, they don't notice the bug. | |
| Apr 19, 2017 at 19:59 | comment | added | Eric Lippert | @aij: Correct in the sense of "clearly legal C# code". For example, have you ever written a program containing an interface that inherited two interfaces where one interface had a property and the other had a method of the same name as the property? Quick, without looking at the spec: is that legal? Now suppose you have a call to that method; is it ambiguous? And so on. People write code that doesn't do what they mean all the time. But only rarely do they write code where you'd have to be a spec expert to say whether it is even legal C#. | |
| Apr 19, 2017 at 18:53 | comment | added | aij | Are you really saying 99.999% of the code people write is "clearly correct"? My experience is pretty much the opposite. | |
| May 28, 2016 at 8:40 | comment | added | Gianluca Ghettini | mmm...bugs which end up in error messages are "fine". It's always possible to fiddle the code a little bit to make it work. What about bugs in which the compiler accepts the source code and produce "wrong" assmebly output. That's scary | |
| Mar 1, 2011 at 7:12 | comment | added | Eric Lippert | @MKO: Of course. Lots of bugs don't get fixed. Sometimes the fix is so expensive and the scenario is so obscure that the cost isn't justified by the benefits. And sometimes enough people have come to rely on the "buggy" behaviour that you have to keep maintaining it. | |
| Feb 28, 2011 at 23:33 | comment | added | Homde | Sureöy there must be energy better spent then making compilers completely "fool"-proof :) | |
| Feb 28, 2011 at 23:26 | history | edited | Eric Lippert | CC BY-SA 2.5 | added 209 characters in body |
| Feb 28, 2011 at 23:08 | comment | added | user1249 | Good error messages after the first one are very hard to do. | |
| S Feb 28, 2011 at 22:13 | history | answered | Eric Lippert | CC BY-SA 2.5 | |
| S Feb 28, 2011 at 22:13 | history | made wiki | Post Made Community Wiki |