Re: Making compilation results deterministic (#4012)

2015-09-18 Thread Simon Marlow
On 17/09/2015 17:30, Simon Peyton Jones wrote: | We don't currently have a good way to reproduce the problem from a | completely clean build, however it's easy to reproduce by doing two | builds and leaving the .hi files from the first build in place while | removing the .o files Indeed. But i

RE: Making compilation results deterministic (#4012)

2015-09-17 Thread Simon Peyton Jones
| We don't currently have a good way to reproduce the problem from a | completely clean build, however it's easy to reproduce by doing two | builds and leaving the .hi files from the first build in place while | removing the .o files Indeed. But is determinism in the face of .hi file changes Par

Re: Making compilation results deterministic (#4012)

2015-09-17 Thread Bartosz Nitka
2015-09-16 12:18 GMT+01:00 Simon Peyton Jones : > > | My understanding is that currently, if you build a Haskell project > | from clean sources with ghc --make, then wipe all the .o/.hi files, > | and rebuild again from clean, with all the same flags and environment, > | you are unlikely to en

Re: Making compilation results deterministic (#4012)

2015-09-17 Thread Simon Marlow
On 16/09/2015 12:18, Simon Peyton Jones wrote: | My understanding is that currently, if you build a Haskell project | from clean sources with ghc --make, then wipe all the .o/.hi files, | and rebuild again from clean, with all the same flags and environment, | you are unlikely to end up with

RE: Making compilation results deterministic (#4012)

2015-09-16 Thread Simon Peyton Jones
| My understanding is that currently, if you build a Haskell project | from clean sources with ghc --make, then wipe all the .o/.hi files, | and rebuild again from clean, with all the same flags and environment, | you are unlikely to end up with identical binaries for either the .o | or .hi f

Re: Making compilation results deterministic (#4012)

2015-09-16 Thread Malcolm Wallace
On 16 Sep 2015, at 10:43, Simon Peyton Jones wrote: > I’ll hypothesise that you mean > · In –make mode, with unchanged source I want to get the same output > from compiling M.hs if M’s imported interface files have not changed. > But even then I’m confused. Under those circumstances, wh

RE: Making compilation results deterministic (#4012)

2015-09-16 Thread Simon Peyton Jones
You’ve described the problem well. Indeed I think you should make a wiki page to articulate the problem, and point to this email (perhaps among others) for more detail. It’s so easy to lose track of email, and so helpful to have a wiki page that always reflects the latest understanding. You

Re: Making compilation results deterministic (#4012)

2015-09-15 Thread Sven Panne
2015-09-14 19:04 GMT+02:00 Bartosz Nitka : > [...] Uniques are non-deterministic [...] > Just out of curiosity: Why is this the case? Naively, I would assume that you can't say much about the value returned by getKey, but at least I would think that in repeated program runs, the same sequence of