This is the mail archive of the mailing list for the binutils project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: PATCH: More stdbool tweaking

> If you've reverted your change, is there any immediate need to change 
>> GDB?  Wasn't it all ``working'' before?  I guess I need to re-read the 
>> thread. Anyway, I think BFD should be fixed before GDB 5.2 gets branched 
>> (I got the what wrong but I don't think I got the how or why).
> I've reverted my change, mostly.  But the point was that on my setup -
> a recent ncurses installed makes the difference - GDB wasn't building
> in the first place.  Obviously I want GDB to build.

Yes.  But is it better to incure a little more short term pain for a 
better longer term gain?

> The current situation may be a little hackish, but the true/false code
> always has been.  I'm all in favor of coming up with a more thorough
> fix, but I'd rather put something in place now than leave GDB not
> building.

I can see two possible things happening.

- we fix it now.  ??.12 gets delayed a week

- we fix it later.  ??.12 doesn't get delayed

The problem I see with later is just how much later it might be.

Once the branch is cut I'd assume that such changes won't be considered 
again until xx.?? has been released (in a month?).  The GDB 5.2 branch 
is also ment to be cut soon and that too will (but to a much lesser 
extent) put pressure on BFD to not change radically.

To me this suggests that the necessary changes won't go in for months 
and that, in turn, makes this very different to normal fixes which are 
localized and can go in as soon as a branch is cut :-(

I guess an argument against this is that BFD has survived up until now 
so one more release won't hurt :-/


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]