Our problem is also in dealing with RHEL/CentOS machines. While it's easy enough to install a newer compiler and switch to that (just by setting CC and CXX), the process of using a non-default Python installation is more onerous and intimidating to infrequent users.
I'm also not very happy in general with the process of saying we support older versions of something, unintentionally using newer features, and then getting backed into a corner with respect to whether we continue to support those old versions. In general, the decision process for changing major dependencies like this should be more intentional. If you want to compromise, the wiki says we still support Python 2.5 as well, and I'm happy to agree to forgo that requirement and bump the minimum required version up to 2.6 ;). Steve On Thu, Apr 14, 2016 at 6:44 AM Jason Lowe-Power <[email protected]> wrote: > Hi Andreas, > > As another data point, we are stuck on RHEL6 for most of our machines here > at Wisconsin. We do have access to Python 2.7.3, but we have to jump > through hoops to use it (e.g., the build breaks when you set python 2.7 as > the default because scons uses 2.6). There are workarounds, but to many of > the users here (people in our Comp Arch classes) it's not very > straightforward. This isn't an argument in either direction, just another > data point ;). > > Jason > > On Thu, Apr 14, 2016 at 8:03 AM Andreas Sandberg <[email protected] > > > wrote: > > > Hi Joe, > > > > It seems like this was more complicated than expected. We already depend > > on argparse, which requires Python 2.7, in quite a few places. This is > > true both for the style checker and other code. > > > > We could always rewrite code to use the deprecated optparse module (which > > isn’t exactly a good option), but this change would obviously be more > > intrusive than getting rid of the autojunk parameter. > > > > Would it be possible for you to upgrade to Python 2.7 instead? > > > > I know that this isn’t ideal for you, but all non-EOL versions of Ubuntu > > already ship with it. RHEL6 ships with Python 2.6, but contains a too old > > gcc version, so that’s not really supported anyway. > > > > Regards, > > Andreas > > > > On 14/04/2016, 13:48, "Andreas Sandberg" <[email protected]> > wrote: > > > > >Hi Joe, > > > > > >Sorry about that. I¹ll look into it and push a patch that fixes this. > > > > > >In practice, we seem to depend on Python 2.6 for things like declaring > > >abstract methods in interfaces, the with statement, the multiprocessing > > >package, advanced string formatting, and probably a lot of other things > as > > >well. I have updated the wiki to reflect this. > > > > > >We should probably consider changing the requirement to Python 2.7 since > > >that¹s what most users are likely to use. None of the currently > supported > > >Ubuntu versions ship with anything older than 2.7.3. Let¹s save that for > > >another thread though. > > > > > >//Andreas > > > > > >On 13/04/2016, 21:45, "Gross, Joe" <[email protected]> wrote: > > > > > >>Hi Andreas, > > >> > > >>This checkin recently made it downstream to our internal repo and I > > >>notice that you've added a SequenceMatcher with the autojunk parameter. > > >>Currently gem5 claims to require python 2.5-2.7 > > >>(http://www.m5sim.org/Dependencies). However, adding this parameter > now > > >>requires > > >>python2.7.1-2.7.11(https://docs.python.org/2/library/difflib.html) and > > >>thus the style checker breaks whenever it is run on our > > >>system(python2.7). > > >> > > >>Could you update the style checker script to use calls that are > supported > > >>by python2.5-2.7.11? We would appreciate it and can then re-enable the > > >>style checker. Thanks! > > >> > > >>Joe > > >> > > >>-----Original Message----- > > >>From: gem5-dev [mailto:[email protected]] On Behalf Of Andreas > > >>Sandberg > > >>Sent: Wednesday, March 30, 2016 10:23 AM > > >>To: [email protected] > > >>Subject: [gem5-dev] changeset in gem5: misc: Don't build region.py as a > > >>PySource > > >> > > >>changeset 1f92ce16e171 in /z/repo/gem5 > > >>details: http://repo.gem5.org/gem5?cmd=changeset;node=1f92ce16e171 > > >>description: > > >>misc: Don't build region.py as a PySource > > >> > > >>The style refactor change (style: Refactor the style checker as a > > >>Python package) moved region.py from src/python/m5/util/ to > > >>util/style/. The SConscript update accidentally got lost in that > > >>commit. This commit removes region.py from src/python/SConscript. > > >> > > >>Signed-off-by: Andreas Sandberg <[email protected]> > > >> > > >>diffstat: > > >> > > >> src/python/SConscript | 1 - > > >> 1 files changed, 0 insertions(+), 1 deletions(-) > > >> > > >>diffs (11 lines): > > >> > > >>diff -r 2201a8c86538 -r 1f92ce16e171 src/python/SConscript > > >>--- a/src/python/SConscriptWed Mar 30 15:56:02 2016 +0100 > > >>+++ b/src/python/SConscriptWed Mar 30 16:21:27 2016 +0100 > > >>@@ -59,7 +59,6 @@ > > >> PySource('m5.util', 'm5/util/jobfile.py') > > >> PySource('m5.util', 'm5/util/multidict.py') > > >> PySource('m5.util', 'm5/util/orderdict.py') > > >>-PySource('m5.util', 'm5/util/region.py') > > >> PySource('m5.util', 'm5/util/smartdict.py') > > >> PySource('m5.util', 'm5/util/sorteddict.py') > > >> PySource('m5.util', 'm5/util/terminal.py') > > >>_______________________________________________ > > >>gem5-dev mailing list > > >>[email protected] > > >>http://m5sim.org/mailman/listinfo/gem5-dev > > > > > > > IMPORTANT NOTICE: The contents of this email and any attachments are > > confidential and may also be privileged. If you are not the intended > > recipient, please notify the sender immediately and do not disclose the > > contents to any other person, use it for any purpose, or store or copy > the > > information in any medium. Thank you. > > _______________________________________________ > > gem5-dev mailing list > > [email protected] > > http://m5sim.org/mailman/listinfo/gem5-dev > > > -- > > Jason > _______________________________________________ > gem5-dev mailing list > [email protected] > http://m5sim.org/mailman/listinfo/gem5-dev > _______________________________________________ gem5-dev mailing list [email protected] http://m5sim.org/mailman/listinfo/gem5-dev
