On Fri, 2008-01-18 at 11:53 -0500, Lennart Sorensen wrote: > On Fri, Jan 18, 2008 at 01:22:40PM +0100, David Fokkema wrote: > > Wow, I'll have to sort through all those options... Since I want debian > > packages, I'll have to check which ones are already being used, which > > ones are against debian policy (the prefix one, for example) and which > > ones may conflict with my original intent, and which ones I don't yet > > understand, ;-) I had hoped to compile mysql-server-5.0 7etch3 version > > and remove the patches one by one until I'm back at 7etch1. Hopefully, I > > can determine which patch breaks my system. Unfortunately, this will > > probably cost a lot of time, even with the --with-low-memory option, I > > guess. Right now, it seems I'm not able to do it before next week, and > > I'll be away then. So in a week or two... But very nice to know it is > > possible to compile this beast on a slug, thanks! > > Remove half the patches, then if it fixes it, add half back, otherwise > remove a further half of the remaining patches, etc until you narrow it > down. > > One by one is a silly method.
Sure it is. > Divide and conquor. Split the problem in half at each step and you will > get there quicker. > > 1024 patches can be narrowed down in 10 compiles after all. Nice, thanks for the idea. I hadn't approached this with an algorithmic view, yet. All in all, we're talking about six patches, IIRC. Still, three compiles is better than six! David > > -- > Len Sorensen > > -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
