Re: [Wesnoth-dev] Choosing a new build system

2008-01-25 Thread Martin Renold
On Thu, Jan 24, 2008 at 08:51:51AM +0100, Jens Seidel wrote: > I'm curious: How many people do you expect to become familiar with a new > system? I'm already somewhat familiar with SCons now. I started using it for new (smaller) projects, mostly c++ plus some code generation during the build. A b

Re: [Wesnoth-dev] Choosing a new build system

2008-01-25 Thread allefant
SilverTree does have a scons build (at least for unix and windows/mingw): http://silvertree.googlecode.com/svn/trunk/SConstruct So that way, everyone can have a glimpse what a build script comparable to Wesnoth would look like. What should be noted, this single script builds both game and editor w

Re: [Wesnoth-dev] Choosing a new build system

2008-01-25 Thread Eric S. Raymond
allefant <[EMAIL PROTECTED]>: > SilverTree does have a scons build (at least for unix and > windows/mingw): http://silvertree.googlecode.com/svn/trunk/SConstruct Thanks! I can copy several of those tests into the scons build for Wesnoth I'm putting together (see SConstruct in the root directory).

Re: [Wesnoth-dev] Choosing a new build system

2008-01-25 Thread allefant
On Jan 25, 2008 5:31 PM, Bruno Wolff III <[EMAIL PROTECTED]> wrote: > > Problems with particular build scripts are also less likely to get > noticed. If you don't have someone who knows all of the build systems in use, > you have to have different people modify them for new additions and I suspect

Re: [Wesnoth-dev] Choosing a new build system

2008-01-25 Thread David Philippi
Am Freitag 25 Januar 2008 schrieb Eric S. Raymond: > What's different today is that autotools has gotten to be a worse and > worse tarpit over time. Thus, its challengers have a better competitive > position today then they did back when they were less mature and > autotools was less encrusted. T

Re: [Wesnoth-dev] Choosing a new build system

2008-01-25 Thread Eric S. Raymond
David Philippi <[EMAIL PROTECTED]>: > I think you'll find lots of arguments against any system. I've seen lots of > different systems appear over the years and vanish again. The dominance of > autotools never got a real challenge... I think it's getting one now. A few years back I thought scons

Re: [Wesnoth-dev] Choosing a new build system

2008-01-25 Thread David Philippi
Am Freitag 25 Januar 2008 schrieb Martin Renold: > I've been looking for arguments against SCons a bit. There are. It looks > like SCons did really fall hard when the KDE project tried to use it: I think you'll find lots of arguments against any system. I've seen lots of different systems appear

Re: [Wesnoth-dev] Choosing a new build system

2008-01-25 Thread Eric S. Raymond
Bruno Wolff III <[EMAIL PROTECTED]>: > Please announce when it is ready to use with some minimal instructions. > I do lots (over one a day on average) of builds from source of Wesnoth trunk > on > Fedora and would like to try testing what will probably be the new build > system. I'll do that. --

Re: [Wesnoth-dev] Choosing a new build system

2008-01-25 Thread Eric S. Raymond
allefant <[EMAIL PROTECTED]>: > Indeed. Instead of one system working good, there are multiple > half-working ones (e.g. in my case, I maintain the scons build, so > it's almost untested for Windows with VC and for OSX). I think we'll probably end up scrapping that one in favor of a single cross-p

Re: [Wesnoth-dev] Choosing a new build system

2008-01-25 Thread Eric S. Raymond
Martin Renold <[EMAIL PROTECTED]>: > I'm already somewhat familiar with SCons now. That's good, because I'm just learning it now. It will be useful to have someone else on the project who understands it. Please have a look at SConstruct in the root directory, which is me trying learn scons by tr

Re: [Wesnoth-dev] Choosing a new build system

2008-01-25 Thread Eric S. Raymond
David Philippi <[EMAIL PROTECTED]>: > There are new versions of autotools from time to time, I wonder wheter they > don't add new nice features most people don't know anything about... They add new macros and tests, sure. The problem is that the underlying architecture of autotools is so bad, an

Re: [Wesnoth-dev] Choosing a new build system

2008-01-25 Thread Eric S. Raymond
Martin Renold <[EMAIL PROTECTED]>: > I've been looking for arguments against SCons a bit. There are. It looks > like SCons did really fall hard when the KDE project tried to use it: > > http://lwn.net/Articles/188693/ (long article but insightful) > > At the end of the WAF manual (pdf) there is a

Re: [Wesnoth-dev] Choosing a new build system

2008-01-25 Thread Bruno Wolff III
On Fri, Jan 25, 2008 at 13:47:55 +0100, allefant <[EMAIL PROTECTED]> wrote: > 3) Maintaining more than one build system is stupid, it doubles (or in > the case of that project, triples) the amount of work :P Problems with particular build scripts are also less likely to get noticed. If you don't

Re: [Wesnoth-dev] Choosing a new build system

2008-01-25 Thread Martin Renold
On Fri, Jan 25, 2008 at 01:18:00PM -0500, Eric S. Raymond wrote: > (Our release manager is going to need to learn scons too. Alas I > suspect Ivanovic still carrying a bit of a torch for cmake...so sad, > too bad.) I've been looking for arguments against SCons a bit. There are. It looks like SCon

Re: [Wesnoth-dev] Choosing a new build system

2008-01-25 Thread Bruno Wolff III
On Fri, Jan 25, 2008 at 13:18:00 -0500, "Eric S. Raymond" <[EMAIL PROTECTED]> wrote: > > There are a few obvious missing features in scons: environment and > options dictionaries should be directly accessible via the [] > operator, and I'm puzzled that there's not a built-in > command-existence

Re: [Wesnoth-dev] Choosing a new build system

2008-01-25 Thread Alfredo Beaumont
Or, 2008eko Urtren 25a(e)an, Eric S. Raymond(e)k idatzi zuen: > If the WAF architect is is bright as I suspect he is, WAF is about a > year of development away from being the best choice in this space, > with a good shot at overtaking scons in deployments once it matures. > But that year has not ha