Re: Re-thinking bootstrapping yi

2012-03-29 Thread Jeremy Wall
On Thu, Mar 29, 2012 at 1:36 PM, Corey O'Connor wrote: > On Thu, Mar 29, 2012 at 7:53 AM, Jeremy Wall wrote: >> I would add that making the build cabal-dev friendly as well would be >> a good idea. >> >> Having yi-contrib and yi be separate build makes compiling both with >> cabal-dev less >> eas

Re: Re-thinking bootstrapping yi

2012-03-29 Thread Corey O'Connor
On Thu, Mar 29, 2012 at 7:53 AM, Jeremy Wall wrote: > I would add that making the build cabal-dev friendly as well would be > a good idea. > > Having yi-contrib and yi be separate build makes compiling both with > cabal-dev less > easy than it could be. I haven't used cabal-dev. What problems do

Re: Re-thinking bootstrapping yi

2012-03-29 Thread Jeremy Wall
On Wed, Mar 28, 2012 at 11:02 PM, Corey O'Connor wrote: > On Wed, Mar 28, 2012 at 1:23 AM, Jean-Philippe Bernardy > wrote: >> I'll just throw my two cents. >> >> I am not sure what are your goals, but my main issue with developing >> was the time it took to compile the sources. I think at the mom

Re: Re-thinking bootstrapping yi

2012-03-29 Thread Marc Weber
a) cabal compiling twice: Yes, I dream about cabal building debug and non debug versions in parallel as well b) compile time rest: yesod is using ghci interpreter to speed up recompilation. Whether this way is usable for a console editor I don't know. Consider this being my thoughts only Marc We

Re: Re-thinking bootstrapping yi

2012-03-28 Thread Jean-Philippe Bernardy
I think I get it now. It's a bit "scary" to require every user to have a copy of the sources. What you propose seems to amount to having all sources become part of the "configuration" we have now. Would it be possible to have Yi look for "configuration files" in the source tree directly? (but onl

Re: Re-thinking bootstrapping yi

2012-03-28 Thread Corey O'Connor
On Wed, Mar 28, 2012 at 1:23 AM, Jean-Philippe Bernardy wrote: > I'll just throw my two cents. > > I am not sure what are your goals, but my main issue with developing > was the time it took to compile the sources. I think at the moment > "cabal install" compile (nearly) every module twice. This s

Re: Re-thinking bootstrapping yi

2012-03-28 Thread Jean-Philippe Bernardy
I'll just throw my two cents. I am not sure what are your goals, but my main issue with developing was the time it took to compile the sources. I think at the moment "cabal install" compile (nearly) every module twice. This should be avoidable by using the new cabal functionality to depend on a li

Re: Re-thinking bootstrapping yi

2012-03-28 Thread Corey O'Connor
On Tue, Mar 27, 2012 at 7:48 PM, Jeremy Wall wrote: > I would hesitate to require yi to be able to clone a git location. > This feels like a bad idea to me. > Having yi understand where the modules are located though that seems > reasonable. Hackage > integration might also be reasonable. How feas

Re: Re-thinking bootstrapping yi

2012-03-27 Thread Jeremy Wall
On Tue, Mar 27, 2012 at 8:32 PM, Corey O'Connor wrote: > I have been considering how to improve the developer experience. The > goal is to make the compile/test cycle shorter. There are a lot of way > to do this. However I'm thinking of going a bit nuts and integrating > the process into yi itself

Re-thinking bootstrapping yi

2012-03-27 Thread Corey O'Connor
I have been considering how to improve the developer experience. The goal is to make the compile/test cycle shorter. There are a lot of way to do this. However I'm thinking of going a bit nuts and integrating the process into yi itself. Code on Yi directly from within Yi and have the re-config proc