[Mono-devel-list] Some patches for Collection
2a5gjx302 at sneakemail.com
Mon Feb 21 14:54:17 EST 2005
At 08:39 PM 19/02/2005 +0100, Martin Baulig wrote:
>On Thu, 2005-02-17 at 12:12 -0500, Ben Maurer wrote:
>> Let's just recap the facts:
>> - This is a bug that has manifested itself in similar cases before. These
>> were not tests for pendantic correctness, but real life bugs filed by
>> - The patch is trivial
>> - Any app it breaks would not run on the msft runtime anyways.
>> So the solution should be obvious, put in the patch.
>It's not obvious at all.
>The same arguments which apply for putting it in also apply in the other
>direction, ie. revert it once it is in.
The first two arguments apply. However, the third argument does NOT apply
in the reverse direction. The purpose of mono is to be able to run any
application that the Microsoft runtime can run. It is NOT to achieve 100%
standards compliance, or to create its own separate niche in the linux
market. If a user writes a program and gets it to work under Microsoft
.NET, and subsequently, that program does not work in mono, that is a bug
in mono. The reverse is NOT true: a program which works in mono and does
not work in MS.NET is a buggy program.
The patch _should_ be accepted. The reverse patch should _not_ be accepted,
ever. Miguel agrees with my point of view, by the way.
More information about the Mono-devel-list