From: Carsten Haitzler (The Rasterman) <raster@ra...> - 2010-11-29 10:53:20
On Fri, 26 Nov 2010 12:42:15 -0200 Iván Briano (Sachiel) <sachieru@...>
> 2010/11/26 Carsten Haitzler <raster@...>:
> > On Fri, 26 Nov 2010 09:56:28 -0200 Gustavo Sverzut Barbieri
> > <barbieri@...> said:
> >> On Fri, Nov 26, 2010 at 8:45 AM, Carsten Haitzler <raster@...>
> >> wrote:
> >> > On Fri, 26 Nov 2010 08:08:52 -0200 Gustavo Sverzut Barbieri
> >> > <barbieri@...> said:
> >> >
> >> > because i dont want to turn it on until 1.1. i want to keep the current
> >> > alloc scheme precisely as-is for now. no no mp created, no counting of
> >> > items or anything else. the macros reduce to exactly the current code as
> >> > it stands right now.
> >> Yes, I understand that. However passthrough is exactly that with the
> >> exception the functions get called, so the binary result is not the
> >> same, but the behavior is (passthrough shouldn't do anything such as
> >> "counting of items or anything else")
> >> But yeah, you're working around an SVN workflow problem with that. In
> >> an ideal world you'd be working on some branch that would be merged
> >> into trunk as soon as it's open for features... which is possible with
> >> SVN, but a major PITA.
> > an ifdef is less work than pretty much any branching and merging :)
> But with branching and merging you can work on several features separately,
> not having any of them influence the other and then forget which
> branch had what.
> Can't you see you are missing all the excitement that comes from
> forgetting about
> that awesome thing you had working some months ago and find out about it one
> day listing all your branches out of curiosity, just to have the
> smartest of merges
> fail and having to do it manually anyway?
------------- Codito, ergo sum - "I code, therefore I am" --------------
The Rasterman (Carsten Haitzler) raster@...
Get latest updates about Open Source Projects, Conferences and News.