----- Original Message ---- From: Ian Lance Taylor <iant@xxxxxxxxxx> To: Arthur Schwarz <aschwarz1309@xxxxxxx> Cc: gcc-help@xxxxxxxxxxx Sent: Wed, July 11, 2012 5:40:42 AM Subject: Re: Help Message Manual On Tue, Jul 10, 2012 at 8:06 PM, Arthur Schwarz <aschwarz1309@xxxxxxx> wrote: > Is there any book, manual, notes, etc. that contains a listing of all >diagnostic > messages with some description of their meanings. No. But feel free to ask about GCC messages here. > Less we get into some odd state where someone says "the messages say what they > mean" or the "message meaning is obvious", there is at least one loader message > where this is not the case. A document would be helpful. I'm not sure what you mean by a "loader message." GCC is a compiler, not a loader. Sorry. The loader message indicates that the vtable is undefined and then provides an erroneous fault location (source code line). Previous dialogs on gcc-help acknowledged that this was a fault in the loader and stated that it was an issue toupdate the loader and fix the problem. Given this statement my assumption wasthat gcc takes responsibility for both the loader an the compiler suite. My current means of dealing with errors that I don't understand is to compile my code with MSVC++(10) and do Google searches. Neither is totally effective and as has been pointed out before, MSVC++(10) does not follow the C++ standard at times. The result of this is that there are occasions where I spend 4+ hours in tracking downand fixing a bug - partly because I like to understand the fault and not just develop a workaround - when I should be "pumping out code". I would feel much better if I were able to look at a manual. Going forward, perhaps when a developer outputs a diagnostic message then the developer provides a description of the message meaning to gcc(?). Collecting these explanations becomes the manual? Ian
