From bwaters at nrao.edu Thu Nov 1 22:59:57 2007 From: bwaters at nrao.edu (Boyd Waters) Date: Thu, 1 Nov 2007 20:59:57 -0600 Subject: [IPython-dev] more tab completion issues. In-Reply-To: References: Message-ID: <1AF1D9DA-4AE1-49A2-98DA-73048AE2AC22@nrao.edu> I get even more duplicates than that... >>> i.__ Display all 121 possibilities? (y or n) i.__abs__ i.__abs__ i.__add__ i.__add__ i.__and__ i.__and__ i.__class__ i.__class__ i.__class__ i.__class__ i.__cmp__ i.__cmp__ i.__coerce__ i.__coerce__ i.__delattr__ i.__delattr__ i.__delattr__ i.__div__ ... Funny. On Nov 1, 2007, at 7:44 PM, Michael VanLandingham wrote: > Following along the whole rlcompleter/readline/editline thread, I > noticed today that when I do tab completion on a class, I get > everything listed twice. > Happens in python and ipython. > Example: > >>> i = int(5) > >>> i > 5 > >>> i.__ > Display all 121 possibilities? (y or n) > i.__abs__ i.__abs__ i.__add__ i.__add__ > i.__and__ i.__and__ i.__class__ > i.__class__ > ...etc. > > > I don't see duplicates if I do 'dir(i)' > > Is anyone else seeing this behavior? Any ideas why? Something in > the rlcompleter or readline... Seems to occur with any class. > > > -Michael > > _______________________________________________ > Pythonmac-SIG maillist - Pythonmac-SIG at python.org > http://mail.python.org/mailman/listinfo/pythonmac-sig From m.vanland at gmail.com Thu Nov 1 23:25:52 2007 From: m.vanland at gmail.com (Michael VanLandingham) Date: Thu, 1 Nov 2007 20:25:52 -0700 Subject: [IPython-dev] more tab completion issues. In-Reply-To: <1AF1D9DA-4AE1-49A2-98DA-73048AE2AC22@nrao.edu> References: <1AF1D9DA-4AE1-49A2-98DA-73048AE2AC22@nrao.edu> Message-ID: Yeah after looking at it again, I noticed certain things, like __class__, had 4, and not 1, entries in the listing. It's a bit annoying, since with something like numpy's 'array', it'll fill up your window with 320+ items. I thought about debugging it, but I don't know how to intercept the 'tab' keystroke in the tab completion. -mvl On Nov 1, 2007, at 7:59 PM, Boyd Waters wrote: > I get even more duplicates than that... > >>> i.__ > Display all 121 possibilities? (y or n) > i.__abs__ i.__abs__ i.__add__ > i.__add__ i.__and__ i.__and__ > i.__class__ i.__class__ i.__class__ > i.__class__ i.__cmp__ i.__cmp__ > i.__coerce__ i.__coerce__ i.__delattr__ > i.__delattr__ i.__delattr__ i.__div__ > ... > > Funny. > > > > On Nov 1, 2007, at 7:44 PM, Michael VanLandingham wrote: > >> Following along the whole rlcompleter/readline/editline thread, I >> noticed today that when I do tab completion on a class, I get >> everything listed twice. >> Happens in python and ipython. >> Example: >> >>> i = int(5) >> >>> i >> 5 >> >>> i.__ >> Display all 121 possibilities? (y or n) >> i.__abs__ i.__abs__ i.__add__ i.__add__ >> i.__and__ i.__and__ i.__class__ >> i.__class__ >> ...etc. >> >> >> I don't see duplicates if I do 'dir(i)' >> >> Is anyone else seeing this behavior? Any ideas why? Something in >> the rlcompleter or readline... Seems to occur with any class. >> >> >> -Michael >> >> _______________________________________________ >> Pythonmac-SIG maillist - Pythonmac-SIG at python.org >> http://mail.python.org/mailman/listinfo/pythonmac-sig > From bwaters at nrao.edu Fri Nov 2 18:12:23 2007 From: bwaters at nrao.edu (Boyd Waters) Date: Fri, 2 Nov 2007 16:12:23 -0600 Subject: [IPython-dev] more tab completion issues. In-Reply-To: References: <1AF1D9DA-4AE1-49A2-98DA-73048AE2AC22@nrao.edu> Message-ID: I've reported this in Apple Bug 5563035. On Nov 1, 2007, at 9:25 PM, Michael VanLandingham wrote: > Yeah after looking at it again, I noticed certain things, like > __class__, had 4, and not 1, entries in the listing. It's a bit > annoying, since with something like numpy's 'array', it'll fill up > your window with 320+ items. > > I thought about debugging it, but I don't know how to intercept the > 'tab' keystroke in the tab completion. > > -mvl > > On Nov 1, 2007, at 7:59 PM, Boyd Waters wrote: > >> I get even more duplicates than that... >> >>> i.__ >> Display all 121 possibilities? (y or n) >> i.__abs__ i.__abs__ i.__add__ >> i.__add__ i.__and__ i.__and__ >> i.__class__ i.__class__ i.__class__ >> i.__class__ i.__cmp__ i.__cmp__ >> i.__coerce__ i.__coerce__ i.__delattr__ >> i.__delattr__ i.__delattr__ i.__div__ >> ... >> >> Funny. >> >> >> >> On Nov 1, 2007, at 7:44 PM, Michael VanLandingham wrote: >> >>> Following along the whole rlcompleter/readline/editline thread, I >>> noticed today that when I do tab completion on a class, I get >>> everything listed twice. >>> Happens in python and ipython. >>> Example: >>> >>> i = int(5) >>> >>> i >>> 5 >>> >>> i.__ >>> Display all 121 possibilities? (y or n) >>> i.__abs__ i.__abs__ i.__add__ >>> i.__add__ >>> i.__and__ i.__and__ i.__class__ >>> i.__class__ >>> ...etc. >>> >>> >>> I don't see duplicates if I do 'dir(i)' >>> >>> Is anyone else seeing this behavior? Any ideas why? Something >>> in the rlcompleter or readline... Seems to occur with any class. >>> >>> >>> -Michael >>> >>> _______________________________________________ >>> Pythonmac-SIG maillist - Pythonmac-SIG at python.org >>> http://mail.python.org/mailman/listinfo/pythonmac-sig >> From fperez.net at gmail.com Fri Nov 2 20:04:24 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Fri, 2 Nov 2007 18:04:24 -0600 Subject: [IPython-dev] Strange behavior of history with multi-line recalls Message-ID: Hi all, I was trying to reuse some previous input and kept running into strange problems. It turns out that my history looks like this: 49: for k in outcnew[2]: nn = norm(outcnew[k]-outcold[k]) print k,'->',nn 50: outcnew[2].keys() 51: sn=set(outcnew[2].keys()) 52: so=set(outcold[2].keys()) 53: sn-so 54: so-sn 55: exec In[48] 56: print In[48] 57: print In[49] 58: exec In[49] 59: for k in outcnew[2]: 60: nn = norm(outcnew[2][k]-outcold[2][k]) 61: print k,'->',nn 62: for k in outcnew[sc]: 63: nn = norm(outcnew[sc][k]-outcold[sc][k]) 64: print k,'->',nn 65: sc etc. Things like lines 59-60 or 62-64 shouldn't happen. Normally whenever you type multi-line input, it registers as a single input line number. The way these ended in there was because I'd recalled previous multi-line input and edited it, point at which somehow the history list ended up filled with more than one entry. Ideas? Cheers, f From gael.varoquaux at normalesup.org Sat Nov 3 10:54:34 2007 From: gael.varoquaux at normalesup.org (Gael Varoquaux) Date: Sat, 3 Nov 2007 15:54:34 +0100 Subject: [IPython-dev] Strange behavior of history with multi-line recalls In-Reply-To: References: Message-ID: <20071103145434.GA5048@clipper.ens.fr> On Fri, Nov 02, 2007 at 06:04:24PM -0600, Fernando Perez wrote: > Things like lines 59-60 or 62-64 shouldn't happen. Normally whenever > you type multi-line input, it registers as a single input line number. > The way these ended in there was because I'd recalled previous > multi-line input and edited it, point at which somehow the history > list ended up filled with more than one entry. I have noticed this too. It happens for lines that have been entered in a previous session, but not for lines entered in the running session. So it seems that the information that these lines form a block does not survive saving and reloading. Ga?l From jorgen.stenarson at bostream.nu Fri Nov 9 14:35:27 2007 From: jorgen.stenarson at bostream.nu (=?ISO-8859-1?Q?J=F6rgen_Stenarson?=) Date: Fri, 09 Nov 2007 20:35:27 +0100 Subject: [IPython-dev] iplib.py is full of tabs Message-ID: <4734B67F.6090901@bostream.nu> Hi, I just updated ipython and saw that iplib.py is full of tab indentations. It's been a week or two since I last updated so I don't know when it occured. Is this an intended change? /J?rgen From fperez.net at gmail.com Fri Nov 9 15:15:20 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Fri, 9 Nov 2007 13:15:20 -0700 Subject: [IPython-dev] iplib.py is full of tabs In-Reply-To: <4734B67F.6090901@bostream.nu> References: <4734B67F.6090901@bostream.nu> Message-ID: On Nov 9, 2007 12:35 PM, J?rgen Stenarson wrote: > Hi, > > I just updated ipython and saw that iplib.py is full of tab > indentations. It's been a week or two since I last updated so I don't > know when it occured. Is this an intended change? Most certainly not. In http://ipython.scipy.org/moin/Developer_Zone/Developer_Guidelines: """ This is just a set of simple guidelines aimed at ensuring that we all spend our limited time on the code itself and not struggling with patch management. Any comments on this are very much welcome (except for any discussion on #1). 1. NEVER, EVER, EVER USE HARD TABS. USE 4 SPACES FOR INDENTATION. I want to make sure that this one is as loudly and clearly spelled out as possible. It is absolutely not open to discussion. I've already had bugs creep in because of patches with hard tabs in them, so in the future I'll simply reject any patch with a single tab in it. IPython is indented (following python.org policy) with 4 spaces per level. """ Brian, I'm afraid your editor is misconfigured. In this case, your commit was the entire file: http://projects.scipy.org/ipython/ipython/changeset/2854 Hopefully this can be cleaned up without too much hassle. Brian, it might be a good idea for you to look at any other commits you've made recently, since you may have introduced hard tabs elsewhere. As a suggestion for the future (for all of us, this could happen to anyone), *please* always look at the actual diff you're about to commit before making the commit. Something quick like svn diff | less will spot this type of problem immediately, since you'll see a multithousand-line diff instead of a small one. I actually have this alias: planck[IPython]> which svd svd: aliased to svn diff | colordiff | less That I use always before commits on any codebase. Colordiff is a really handy little utility that syntax highlights diffs at the command line, available for ubuntu with a simple apt-get install, and probably trivial to install on any *nix-based box. Cheers, f From ellisonbg at gmail.com Fri Nov 9 15:53:21 2007 From: ellisonbg at gmail.com (Brian Granger) Date: Fri, 9 Nov 2007 13:53:21 -0700 Subject: [IPython-dev] iplib.py is full of tabs In-Reply-To: References: <4734B67F.6090901@bostream.nu> Message-ID: > Brian, I'm afraid your editor is misconfigured. In this case, your > commit was the entire file: Yep, I made that edit on a clean Leopard system - where I didn't have my usual editor :) Sorry about that, I will clean it up. Brian -- Brian E. Granger, Ph.D. Research Scientist Tech-X Corporation phone: 720-974-1850 bgranger at txcorp.com ellisonbg at gmail.com From fperez.net at gmail.com Fri Nov 9 15:58:18 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Fri, 9 Nov 2007 13:58:18 -0700 Subject: [IPython-dev] iplib.py is full of tabs In-Reply-To: References: <4734B67F.6090901@bostream.nu> Message-ID: On Nov 9, 2007 1:53 PM, Brian Granger wrote: > > Brian, I'm afraid your editor is misconfigured. In this case, your > > commit was the entire file: > > Yep, I made that edit on a clean Leopard system - where I didn't have > my usual editor :) Sorry about that, I will clean it up. No worries, thanks for the commit with the fix, that was quick ;) Cheers, f From steve at shrogers.com Fri Nov 16 21:39:30 2007 From: steve at shrogers.com (Steven H. Rogers) Date: Fri, 16 Nov 2007 19:39:30 -0700 Subject: [IPython-dev] Enhanced IPython Help Message-ID: <473E5462.9010203@shrogers.com> During the IPython sprint last Spring, a couple of us looked at interactively searching the documentation for all available modules, but this project has languished other than some ideas on the `wiki `_. I'm ready to devote more time to this now. I've submitted a paper proposal on it for PyCon 2008, and if it's accepted, I'll need to get something working or I'll be terribly embarrassed. Unless someone has a better idea, I'm planning to extend "?" such that if the search string isn't found in the local namespace, modules in PYTHONPATH are searched for a match. I'm looking at `pyndexter `_ for indexing and searching. Any ideas welcome. Should work from the IPython or IPython1 svn repositories? # Steve From vivainio at gmail.com Mon Nov 19 01:50:53 2007 From: vivainio at gmail.com (Ville M. Vainio) Date: Mon, 19 Nov 2007 08:50:53 +0200 Subject: [IPython-dev] Enhanced IPython Help In-Reply-To: <473E5462.9010203@shrogers.com> References: <473E5462.9010203@shrogers.com> Message-ID: <46cb515a0711182250t42ef99b3j20654a5c58a9febc@mail.gmail.com> On Nov 17, 2007 4:39 AM, Steven H. Rogers wrote: > Unless someone has a better idea, I'm planning to extend "?" such that > if the search string isn't found in the local namespace, modules in > PYTHONPATH are searched for a match. I'm looking at `pyndexter > `_ for indexing and searching. Any ideas > welcome. Should work from the IPython or IPython1 svn repositories? As I see it, you should make it as IPython-independent as possible, and then we'll add small wrappers to both IPython branches (probably as Extensions). -- Ville M. Vainio - vivainio.googlepages.com blog=360.yahoo.com/villevainio - g[mail | talk]='vivainio' From fperez.net at gmail.com Tue Nov 20 12:48:45 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Tue, 20 Nov 2007 10:48:45 -0700 Subject: [IPython-dev] Enhanced IPython Help In-Reply-To: <46cb515a0711182250t42ef99b3j20654a5c58a9febc@mail.gmail.com> References: <473E5462.9010203@shrogers.com> <46cb515a0711182250t42ef99b3j20654a5c58a9febc@mail.gmail.com> Message-ID: Hi Steven, On Nov 18, 2007 11:50 PM, Ville M. Vainio wrote: > On Nov 17, 2007 4:39 AM, Steven H. Rogers wrote: > > > Unless someone has a better idea, I'm planning to extend "?" such that > > if the search string isn't found in the local namespace, modules in > > PYTHONPATH are searched for a match. I'm looking at `pyndexter > > `_ for indexing and searching. Any ideas > > welcome. Should work from the IPython or IPython1 svn repositories? Many thanks for keeping an eye on this idea. I really think it will make a big difference in usability for everyone to have nice, well integrated help available interactively. I see from http://swapoff.org/browser/pyndexter/trunk/COPYING that the pyndexter license is BSD, which is great. Given that the current interactive response to things like '?' is only implemented in the main IPython (not '1') repo, you should work from there. I'll be happy to give you SVN access to make a branch if you find that you need it, until the work is ready for merging into trunk. > As I see it, you should make it as IPython-independent as possible, > and then we'll add small wrappers to both IPython branches (probably > as Extensions). The bulk of it should definitely be as self-contained and modular as possible, but if you find that depending on any of ipython makes life easier for you, don't be afraid to. By this I mean that if from IPython import something helps your project, go ahead and do it freely. As far as wiring it into the interactive behavior, there it definitely makes sense to have your code be in its own module(s), and then simply wire a line or two into the ipython help code to call out to your module. Cheers, f From fperez.net at gmail.com Tue Nov 20 12:53:20 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Tue, 20 Nov 2007 10:53:20 -0700 Subject: [IPython-dev] IPython and SDL/PyGame In-Reply-To: References: Message-ID: Hi Jim, sorry for the delayed reply, family business... On Nov 13, 2007 12:14 PM, Jim Burnes wrote: > Fernando, > > I'm writing a book on programming in Python. I'd like to use IPython > with SDL/PyGame/OpenGL. > > I noticed that you have GUI interfaces to PyLab / GTK / QT, WxWindow > event loops. Do you know of any special issues with SDL/OpenGL? > > Thanks, > > Jim Burnes I'm cc-ing the ipython dev list because someone else might have something else to add on this front. The first thing you should look at is: http://ipython.scipy.org/moin/Cookbook/EmbeddingInOpenGL I should add that at one point, I hacked a bit a -gl flag similar to the -wx/-qt/etc ones in my private copy, as I wanted to interactively run OpenGL apps from ipython for testing GPU accelerated codes interactively. It seemed to work kind of OK, but I kept running into problems with re-initialization of the OpenGL main loop. Since that project rapidly moved over to a pure C++ codebase, I didn't pursue it further. Honestly I only spent maybe one hour testing this, so I suspect a full solution is actually feasible. Cheers, f From steve at shrogers.com Wed Nov 21 08:12:55 2007 From: steve at shrogers.com (Steven H. Rogers) Date: Wed, 21 Nov 2007 06:12:55 -0700 Subject: [IPython-dev] Enhanced IPython Help In-Reply-To: References: <473E5462.9010203@shrogers.com> <46cb515a0711182250t42ef99b3j20654a5c58a9febc@mail.gmail.com> Message-ID: <47442ED7.8070004@shrogers.com> Fernando Perez wrote: > Hi Steven, > > On Nov 18, 2007 11:50 PM, Ville M. Vainio wrote: > >> On Nov 17, 2007 4:39 AM, Steven H. Rogers wrote: >> >> >>> Unless someone has a better idea, I'm planning to extend "?" such that >>> if the search string isn't found in the local namespace, modules in >>> PYTHONPATH are searched for a match. I'm looking at `pyndexter >>> `_ for indexing and searching. Any ideas >>> welcome. Should work from the IPython or IPython1 svn repositories? >>> > > Many thanks for keeping an eye on this idea. I really think it will > make a big difference in usability for everyone to have nice, well > integrated help available interactively. > > I see from > > http://swapoff.org/browser/pyndexter/trunk/COPYING > > that the pyndexter license is BSD, which is great. > > Given that the current interactive response to things like '?' is only > implemented in the main IPython (not '1') repo, you should work from > there. I'll be happy to give you SVN access to make a branch if you > find that you need it, until the work is ready for merging into trunk. > > >> As I see it, you should make it as IPython-independent as possible, >> and then we'll add small wrappers to both IPython branches (probably >> as Extensions). >> > > The bulk of it should definitely be as self-contained and modular as > possible, but if you find that depending on any of ipython makes life > easier for you, don't be afraid to. By this I mean that if > > from IPython import something > > helps your project, go ahead and do it freely. As far as wiring it > into the interactive behavior, there it definitely makes sense to have > your code be in its own module(s), and then simply wire a line or two > into the ipython help code to call out to your module. > > Cheers, > > f > > I don't see any need for any IPython dependency beyond the desire to use the '?' symbol. It should be usable in a standard Python shell. Now for a module name. I called the wiki page SearchDocs, so is searchdocs or perhaps docsearch a reasonable module name? Happy Thanksgiving, # Steve From fperez.net at gmail.com Wed Nov 21 12:17:51 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Wed, 21 Nov 2007 10:17:51 -0700 Subject: [IPython-dev] Enhanced IPython Help In-Reply-To: <47442ED7.8070004@shrogers.com> References: <473E5462.9010203@shrogers.com> <46cb515a0711182250t42ef99b3j20654a5c58a9febc@mail.gmail.com> <47442ED7.8070004@shrogers.com> Message-ID: On Nov 21, 2007 6:12 AM, Steven H. Rogers wrote: > I don't see any need for any IPython dependency beyond the desire to use > the '?' symbol. It should be usable in a standard Python shell. Now > for a module name. I called the wiki page SearchDocs, so is searchdocs > or perhaps docsearch a reasonable module name? +1 for docsearch from me. Simple, short and clear. > Happy Thanksgiving, > # Steve Likewise. Cheers, f From fperez.net at gmail.com Thu Nov 22 15:54:10 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Thu, 22 Nov 2007 13:54:10 -0700 Subject: [IPython-dev] IPython 0.8.2 and PyReadline 1.5, getting ready for release... Message-ID: Hi all, we've been delaying 0.8.2 for a while (my fault), so I wanted to touch bases with everyone to see if there's anything in your pipelines that you absolutely want to get in, else we should probably go ahead and make a release. I've put builds of ipython and pyreadline from current trunk SVN in the usual location: http://ipython.scipy.org/dist/testing/ so anyone can test them. Would this weekend be a good time to make a release, or does anyone want a bit longer delay? I know we have a lot of pending tickets, but holding a release further won't help that, and there are many fixes already in the codebase that users who don't track SVN can benefit from, so I think we should just go ahead and make the release. Thoughts? Cheers, f From fperez.net at gmail.com Thu Nov 22 15:58:36 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Thu, 22 Nov 2007 13:58:36 -0700 Subject: [IPython-dev] [IPython-user] IPython 0.8.2 and PyReadline 1.5, getting ready for release... In-Reply-To: References: Message-ID: On Nov 22, 2007 1:57 PM, Noah Gift wrote: > > > > > so anyone can test them. Would this weekend be a good time to make a > > release, or does anyone want a bit longer delay? > > > > I will help run through the paces on Leopard. Is there an dev egg yet? Thanks! Yes, here's the egg we put out with the standard build script: http://ipython.scipy.org/dist/testing/ipython-0.8.2.svn.r2848-py2.5.egg Let us know if you hit any problems with it. Cheers, f From fperez.net at gmail.com Thu Nov 22 16:08:22 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Thu, 22 Nov 2007 14:08:22 -0700 Subject: [IPython-dev] [Enthought-dev] pausing in IPython in a Chaco based wx app In-Reply-To: References: Message-ID: On Oct 18, 2007 10:30 AM, Tom Denniston wrote: > I am cross posting to IPython and Enthought because I don't know > whether this is an IPython or ets question. > > I am driving a wx-traits-chaco app from IPython using ipython in > wx-thread mode. The one problem is I sometimes want to loop in > IPython doing something that updates the data in the gui and then > pauses till the user is done looking at the new data, then the user > hits enter and it continues updating the gui with yet more data. > > I was doing the dumb thing of using the python input function but that > seems to lock the wx thread so that the gui doesn't update and becomes > unresponsive. Is there a way to code a > that doesn't lock the wx thread? Off the top of my head I can't think of a solution, but if you find one (you may already have), it would be nice to have it listed in the ipython cookbook: http://ipython.scipy.org/moin/Cookbook as I'm sure you're not the only one with this need. Cheers, f From fperez.net at gmail.com Thu Nov 22 16:11:51 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Thu, 22 Nov 2007 14:11:51 -0700 Subject: [IPython-dev] Continuing from an exeption In-Reply-To: <20070920202124.GB26651@clipper.ens.fr> References: <20070920202124.GB26651@clipper.ens.fr> Message-ID: On Sep 20, 2007 1:21 PM, Gael Varoquaux wrote: > OK, this is a stupid question ,the answer is probably somewhere in the > manual, and the question should be asked on the user mailing list. But... > > How can I continue running a script after it bailed out because of an > exception ? > > For instance I have > > """ > b = do_long_calculation() > print a > a = other_long_calculation(b) > """ > > This will fail because a does not exist. I can "%debug" to go in debug > mode and find that out. I could correct the problem by defining a, and > then continue the script, or I could just decide to skip the line and > continue the script. > > Is there a way to do this ? No, the python language doesn't really support execution resumption after an exception. I imagine one could hack up something very nasty that would involve an exec-like function that takes the existing stack and restarts it with modified pointers to new code, but it's not anything that I've seen done, nor that I plan on writing anytime soon :) Cheers, f From fperez.net at gmail.com Thu Nov 22 16:13:05 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Thu, 22 Nov 2007 14:13:05 -0700 Subject: [IPython-dev] ipy_traits_completer In-Reply-To: <200709190952.15047.dd55@cornell.edu> References: <200709190952.15047.dd55@cornell.edu> Message-ID: Hey Darren, On Sep 19, 2007 6:52 AM, Darren Dale wrote: > I just activated ipy_traits_completer in my ipy_user_conf.py file: > > def main(): > from ipy_traits_completer import activate > complete_threshold = 3 > activate(complete_threshold) > > import ipy_stock_completers > > o = ip.options > main() > > When the traits completer is activated, my readline_omit__names setting in > ipythonrc is ignored. could you file a ticket on this ? Or even better, fix it if you can, since you now have write access :) Cheers, f From gael.varoquaux at normalesup.org Thu Nov 22 16:13:31 2007 From: gael.varoquaux at normalesup.org (Gael Varoquaux) Date: Thu, 22 Nov 2007 22:13:31 +0100 Subject: [IPython-dev] Continuing from an exeption In-Reply-To: References: <20070920202124.GB26651@clipper.ens.fr> Message-ID: <20071122211330.GA18460@clipper.ens.fr> On Thu, Nov 22, 2007 at 02:11:51PM -0700, Fernando Perez wrote: > > Is there a way to do this ? > No, the python language doesn't really support execution resumption > after an exception. I imagine one could hack up something very nasty > that would involve an exec-like function that takes the existing stack > and restarts it with modified pointers to new code, but it's not > anything that I've seen done, nor that I plan on writing anytime soon > :) OK, thanks. I had come to this conclusion too. (you really are going through your backlog :->). Ga?l From fperez.net at gmail.com Thu Nov 22 16:15:20 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Thu, 22 Nov 2007 14:15:20 -0700 Subject: [IPython-dev] Continuing from an exeption In-Reply-To: <20071122211330.GA18460@clipper.ens.fr> References: <20070920202124.GB26651@clipper.ens.fr> <20071122211330.GA18460@clipper.ens.fr> Message-ID: On Nov 22, 2007 2:13 PM, Gael Varoquaux wrote: > On Thu, Nov 22, 2007 at 02:11:51PM -0700, Fernando Perez wrote: > > > Is there a way to do this ? > > > No, the python language doesn't really support execution resumption > > after an exception. I imagine one could hack up something very nasty > > that would involve an exec-like function that takes the existing stack > > and restarts it with modified pointers to new code, but it's not > > anything that I've seen done, nor that I plan on writing anytime soon > > :) > > OK, thanks. I had come to this conclusion too. > > (you really are going through your backlog :->). Well, travel and family matters are kind of over for a bit, so I'm trying. But it's a lost battle... :) Cheers, f From gael.varoquaux at normalesup.org Thu Nov 22 16:17:12 2007 From: gael.varoquaux at normalesup.org (Gael Varoquaux) Date: Thu, 22 Nov 2007 22:17:12 +0100 Subject: [IPython-dev] [Enthought-dev] pausing in IPython in a Chaco based wx app In-Reply-To: References: Message-ID: <20071122211712.GC18460@clipper.ens.fr> On Thu, Nov 22, 2007 at 02:08:22PM -0700, Fernando Perez wrote: > > I was doing the dumb thing of using the python input function but > > that > > seems to lock the wx thread so that the gui doesn't update and > > becomes > > unresponsive. Is there a way to code a > > that doesn't lock the wx thread? > Off the top of my head I can't think of a solution, but if you find > one (you may already have), it would be nice to have it listed in the > ipython cookbook: > http://ipython.scipy.org/moin/Cookbook > as I'm sure you're not the only one with this need. Yes, this would be very important for interaction with say pylab in data processing scripts. For instance imagine you need to code something that asks the user to select a region of interest on a graph, and then go on with the processing. This a weak point compared to matlab and it hurts us (my lab) in our day-to-day work. Ga?l From vivainio at gmail.com Fri Nov 23 02:45:19 2007 From: vivainio at gmail.com (Ville M. Vainio) Date: Fri, 23 Nov 2007 09:45:19 +0200 Subject: [IPython-dev] ipy_traits_completer In-Reply-To: <200709190952.15047.dd55@cornell.edu> References: <200709190952.15047.dd55@cornell.edu> Message-ID: <46cb515a0711222345y7ef22db8s8926fa7f502c68ff@mail.gmail.com> On Sep 19, 2007 3:52 PM, Darren Dale wrote: > I just activated ipy_traits_completer in my ipy_user_conf.py file: > > def main(): > from ipy_traits_completer import activate > complete_threshold = 3 > activate(complete_threshold) > > import ipy_stock_completers > > o = ip.options > main() > > When the traits completer is activated, my readline_omit__names setting in > ipythonrc is ignored. Try changing it in your ipy_user_conf.py after activate(). This is not strictly a bug, python-side configuration overrides ipythonrc. -- Ville M. Vainio - vivainio.googlepages.com blog=360.yahoo.com/villevainio - g[mail | talk]='vivainio' From vivainio at gmail.com Fri Nov 23 08:30:44 2007 From: vivainio at gmail.com (Ville M. Vainio) Date: Fri, 23 Nov 2007 15:30:44 +0200 Subject: [IPython-dev] IPython 0.8.2 and PyReadline 1.5, getting ready for release... In-Reply-To: References: Message-ID: <46cb515a0711230530u17dd66b1r1c17a2d32499bc7d@mail.gmail.com> On Nov 22, 2007 10:54 PM, Fernando Perez wrote: > we've been delaying 0.8.2 for a while (my fault), so I wanted to touch > bases with everyone to see if there's anything in your pipelines that > you absolutely want to get in, else we should probably go ahead and > make a release. I don't have anything that requires a fix right now. -- Ville M. Vainio - vivainio.googlepages.com blog=360.yahoo.com/villevainio - g[mail | talk]='vivainio' From darren.dale at cornell.edu Fri Nov 23 11:11:42 2007 From: darren.dale at cornell.edu (Darren Dale) Date: Fri, 23 Nov 2007 11:11:42 -0500 Subject: [IPython-dev] ipy_traits_completer In-Reply-To: <46cb515a0711222345y7ef22db8s8926fa7f502c68ff@mail.gmail.com> References: <200709190952.15047.dd55@cornell.edu> <46cb515a0711222345y7ef22db8s8926fa7f502c68ff@mail.gmail.com> Message-ID: <200711231111.42830.darren.dale@cornell.edu> On Friday 23 November 2007 02:45:19 am Ville M. Vainio wrote: > On Sep 19, 2007 3:52 PM, Darren Dale wrote: > > I just activated ipy_traits_completer in my ipy_user_conf.py file: > > > > def main(): > > from ipy_traits_completer import activate > > complete_threshold = 3 > > activate(complete_threshold) > > > > import ipy_stock_completers > > > > o = ip.options > > main() > > > > When the traits completer is activated, my readline_omit__names setting > > in ipythonrc is ignored. > > Try changing it in your ipy_user_conf.py after activate(). This is not > strictly a bug, python-side configuration overrides ipythonrc. Thank you, Ville, your suggestion worked. I notice, though, that ip_traits_completer does not respect readline_omit__names. I would like to fix this before the release, I think I can get it done today. I guess I have not been following the list close enough. I didn't realize ipython was already moving away from rc files, I thought that was only in ipython1. I like the new default prompt. Darren From darren.dale at cornell.edu Fri Nov 23 12:59:19 2007 From: darren.dale at cornell.edu (Darren Dale) Date: Fri, 23 Nov 2007 12:59:19 -0500 Subject: [IPython-dev] ipy_traits_completer In-Reply-To: <200711231111.42830.darren.dale@cornell.edu> References: <200709190952.15047.dd55@cornell.edu> <46cb515a0711222345y7ef22db8s8926fa7f502c68ff@mail.gmail.com> <200711231111.42830.darren.dale@cornell.edu> Message-ID: <200711231259.19525.darren.dale@cornell.edu> On Friday 23 November 2007 11:11:42 am Darren Dale wrote: > On Friday 23 November 2007 02:45:19 am Ville M. Vainio wrote: > > On Sep 19, 2007 3:52 PM, Darren Dale wrote: > > > I just activated ipy_traits_completer in my ipy_user_conf.py file: > > > > > > def main(): > > > from ipy_traits_completer import activate > > > complete_threshold = 3 > > > activate(complete_threshold) > > > > > > import ipy_stock_completers > > > > > > o = ip.options > > > main() > > > > > > When the traits completer is activated, my readline_omit__names setting > > > in ipythonrc is ignored. > > > > Try changing it in your ipy_user_conf.py after activate(). This is not > > strictly a bug, python-side configuration overrides ipythonrc. > > Thank you, Ville, your suggestion worked. I notice, though, that > ip_traits_completer does not respect readline_omit__names. I would like to > fix this before the release, I think I can get it done today. This is done. Here is the change I made: Index: IPython/Extensions/ipy_traits_completer.py =================================================================== --- IPython/Extensions/ipy_traits_completer.py (revision 2868) +++ IPython/Extensions/ipy_traits_completer.py (working copy) @@ -99,6 +99,13 @@ attr_start = symbol_parts[-1] if attr_start: attrs = [a for a in attrs if a.startswith(attr_start)] + + # Let's also respect the user's readline_omit__names setting: + omit__names = ipget().options.readline_omit__names + if omit__names == 1: + attrs = [a for a in attrs if not a.startswith('__')] + elif omit__names == 2: + attrs = [a for a in attrs if not a.startswith('_')] #print '\nastart:<%r>' % attr_start # dbg Darren From darren.dale at cornell.edu Fri Nov 23 12:59:55 2007 From: darren.dale at cornell.edu (Darren Dale) Date: Fri, 23 Nov 2007 12:59:55 -0500 Subject: [IPython-dev] [IPython-user] IPython 0.8.2 and PyReadline 1.5, getting ready for release... In-Reply-To: References: Message-ID: <200711231259.55409.darren.dale@cornell.edu> On Thursday 22 November 2007 03:54:10 pm Fernando Perez wrote: > Hi all, > > we've been delaying 0.8.2 for a while (my fault), so I wanted to touch > bases with everyone to see if there's anything in your pipelines that > you absolutely want to get in, else we should probably go ahead and > make a release. > > I've put builds of ipython and pyreadline from current trunk SVN in > the usual location: > > http://ipython.scipy.org/dist/testing/ > > so anyone can test them. Would this weekend be a good time to make a > release, or does anyone want a bit longer delay? > > I know we have a lot of pending tickets, but holding a release further > won't help that, and there are many fixes already in the codebase that > users who don't track SVN can benefit from, so I think we should just > go ahead and make the release. > > Thoughts? I'm all set. Darren From fperez.net at gmail.com Sat Nov 24 00:12:15 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Fri, 23 Nov 2007 22:12:15 -0700 Subject: [IPython-dev] [IPython-user] IPython 0.8.2 and PyReadline 1.5, getting ready for release... In-Reply-To: <200711231259.55409.darren.dale@cornell.edu> References: <200711231259.55409.darren.dale@cornell.edu> Message-ID: On Nov 23, 2007 10:59 AM, Darren Dale wrote: > I'm all set. Sounds good. We'll give the rc another day or so in case someone finds any last-minute nasties, and I'll package up the release files for Sunday or so. Cheers, f From fperez.net at gmail.com Sat Nov 24 14:26:25 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Sat, 24 Nov 2007 12:26:25 -0700 Subject: [IPython-dev] [IPython-user] IPython 0.8.2 and PyReadline 1.5, getting ready for release... In-Reply-To: <2E502158-11E3-4555-AD17-110A8AEA0E7B@gmail.com> References: <200711231259.55409.darren.dale@cornell.edu> <2E502158-11E3-4555-AD17-110A8AEA0E7B@gmail.com> Message-ID: On Nov 24, 2007 11:29 AM, Noah Gift wrote: > So far I like %doctest_mode > > It is a good start, nice work!! No other weirdness I have noticed, > other than a simple user error problem, of not knowing how to get a > four space tab to work when I define a function: > > def function: > > I am sure this is user error, but I am stuck with doing four spaces, > as a tab will automatically try to autocomplete. Two things: 1. why do you need the explicit tab in the above case? Is the auto-indenting not working for you? That would certainly qualify as a bug. 2. By default, we bind ctrl-i to inserting an explicit tab for indentation purposes. That can be changed to any other key in the configuration. Cheers, f From fperez.net at gmail.com Sat Nov 24 15:00:19 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Sat, 24 Nov 2007 13:00:19 -0700 Subject: [IPython-dev] [IPython-user] IPython 0.8.2 and PyReadline 1.5, getting ready for release... In-Reply-To: References: <200711231259.55409.darren.dale@cornell.edu> <2E502158-11E3-4555-AD17-110A8AEA0E7B@gmail.com> Message-ID: On Nov 24, 2007 12:51 PM, Noah Gift wrote: > It appears auto-indention is not working for me. Here is a default session: > > > [ngift at Macintosh-6 ][H:10426][J:0]> ipython > Leopard libedit detected. Ah. Could you please try import readline readline. and paste in the output? With GNU readline, we get: In [2]: readline. readline.__builtins__ readline.get_current_history_length readline.__class__ readline.get_endidx readline.__delattr__ readline.get_history_item readline.__dict__ readline.get_history_length readline.__doc__ readline.get_line_buffer readline.__file__ readline.have_readline readline.__getattribute__ readline.insert_text readline.__hash__ readline.parse_and_bind readline.__init__ readline.read_history_file readline.__name__ readline.read_init_file readline.__new__ readline.redisplay readline.__reduce__ readline.remove_history_item readline.__reduce_ex__ readline.replace_history_item readline.__repr__ readline.set_completer readline.__setattr__ readline.set_completer_delims readline.__str__ readline.set_history_length readline._rl readline.set_pre_input_hook readline.add_history readline.set_startup_hook readline.clear_history readline.sys readline.get_begidx readline.uses_libedit readline.get_completer readline.write_history_file readline.get_completer_delims And the critical one is readline.insert_text which we use to rewrite the new line you're about to type into via the pre_readline method: http://projects.scipy.org/ipython/ipython/browser/ipython/trunk/IPython/iplib.py#L1281 If that function doesn't exist, or is a no-op, we're stuck without auto-indent in Leopard. If it does work, then we may be doing something else incorrectly though. I unfortunately don't have a leopard box to test on yet (waiting for my wife's laptop to arrrive...). Cheers, f From fperez.net at gmail.com Sat Nov 24 16:24:12 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Sat, 24 Nov 2007 14:24:12 -0700 Subject: [IPython-dev] [IPython-user] IPython 0.8.2 and PyReadline 1.5, getting ready for release... In-Reply-To: References: <200711231259.55409.darren.dale@cornell.edu> <2E502158-11E3-4555-AD17-110A8AEA0E7B@gmail.com> Message-ID: On Nov 24, 2007 1:48 PM, Noah Gift wrote: > Here are the results: > I do see a readline.insert_text. One other symptom I notice is that if I > use the up arrow to scroll back through my command history, that the lines > become confused and the spacing will be odd like follows: > > > In [2]: clear > > In [3]: def function(): > ...: print "This is a test" > ...: > > In [4]: def function(): clear > > The clear from line 2 appears in line 3. > > I am always happy to help do QA testing for Leopard, as it is killing two > birds with one stone for Jeremy and I. OK, thanks for the info. It looks to me like we have a bit of real work still to do on Leopard, but I'd rather not delay this release too much longer because of it. There will always be 0.8.3, and we can put that out soonish if we figure out what's going on with Leopard. I will get access to a Leopard machine soon (roughly a week, hopefully), so I should be able to dig into this myself. Unless someone has a fix they can send us, I think I'll cut out the 0.8.2 release tomorrow, and we'll continue to work on this one in the upcoming weeks so we can have a fully Leopard-happy ipython. It seems to me that the multi-line history is getting mangled up in leopard's libedit, which may behave differently than readline in some subtle aspects. That may be a bug in our part or one in libedit, but even if it's the latter, we might be able to work around it somehow. Cheers, f From fperez.net at gmail.com Sat Nov 24 23:29:08 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Sat, 24 Nov 2007 21:29:08 -0700 Subject: [IPython-dev] Fwd: [sage-support] Re: install problem with pickleshare In-Reply-To: <46cb515a0710172347p11425253tb60ec0132161b03f@mail.gmail.com> References: <1192514681.204773.36580@i13g2000prf.googlegroups.com> <85e81ba30710160006m2c2f3ae0g215a1eeae57095ed@mail.gmail.com> <1192548768.188492.94600@v23g2000prn.googlegroups.com> <85e81ba30710161612n1f5907dey924e0120a2989639@mail.gmail.com> <96de71860710170322qbb1bae6yd5fcb905267db4a9@mail.gmail.com> <46cb515a0710172347p11425253tb60ec0132161b03f@mail.gmail.com> Message-ID: On Oct 17, 2007 11:47 PM, Ville M. Vainio wrote: > Perhaps this is due to sys.path not containing Extensions yet for some > reason (I'm not aware why this is not a problem on my box). > > To get over it quickly, change line: > > > python2.5/site-packages/IPython/iplib.py", line 58, in > > import pickleshare > > > to: > > from IPython.Extensions import pickleshare. Mmh, I hadn't realized this bug hadn't been fixed in trunk. I've just committed the fix as r2871. Cheers, f From fperez.net at gmail.com Sun Nov 25 01:04:13 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Sat, 24 Nov 2007 23:04:13 -0700 Subject: [IPython-dev] Patch to fix string exceptions raised by DPyGetOpt.py In-Reply-To: <20070509002739.GA5111@localhost.localdomain> References: <20070509002739.GA5111@localhost.localdomain> Message-ID: Hi Paul, On May 8, 2007 5:27 PM, Paul Mueller wrote: > Hi, > > There are some string exceptions in IPython/DPyGetOpt.py (in r2314), some of > which are raised when passing bad options to ipython on the command line. Since > string exceptions are deprecated in python 2.5, this raises a > DeprecationWarning (usage message redirected to /dev/null for clarity): [...] Many thanks for your patch, and a huge apology that this seemed to fall through the cracks for such a long time. But it was still nagging me from the bottom of my inbox, so I finally got to it :) I just applied it and it will go into 0.8.2, to be released very soon. Best regards, f From fperez.net at gmail.com Sun Nov 25 01:08:00 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Sat, 24 Nov 2007 23:08:00 -0700 Subject: [IPython-dev] Unicode bug in %whos, with patch In-Reply-To: <20070503202243.GA7261@localhost.localdomain> References: <20070503202243.GA7261@localhost.localdomain> Message-ID: On May 3, 2007 1:22 PM, Paul Mueller wrote: > Hi, > > If you have a unicode variable with a character whose ordinal is >= 128, you > get a UnicodeEncodeError on using %whos (this is using r2304): > > In [1]:u = u'\x80' > > In [2]:%whos > Variable Type Data/Info > ------------------------------- > u unicode --------------------------------------------------------------------------- > Traceback (most recent call last) > > /home/yuurei/src/mods/ipython/ in () > > /home/yuurei/src/mods/ipython/IPython/iplib.py in ipmagic(self, arg_s) > 962 else: > 963 magic_args = self.var_expand(magic_args,1) > --> 964 return fn(magic_args) > 965 > 966 def ipalias(self,arg_s): > > /home/yuurei/src/mods/ipython/IPython/Magic.py in magic_whos(self, parameter_s) > 1057 print '(%s Mb)' % (vbytes/Mb,) > 1058 else: > -> 1059 vstr = str(var).replace('\n','\\n') > 1060 if len(vstr) < 50: > 1061 print vstr > > : 'ascii' codec can't encode character u'\x80' in position 0: ordinal not in range(128) > > > I've attached a patch that fixes this. Thanks again, Paul, things are now OK for the upcoming release: In [1]: u = u'\x80' In [2]: %whos Variable Type Data/Info ------------------------------- u unicode \x80 In [3]: Sorry again for the ridiculous delay in dealing with your contributions. That's no way to treat your users, esp. those who pitch in with patches! Mea culpa. Cheers, f From fperez.net at gmail.com Sun Nov 25 01:13:13 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Sat, 24 Nov 2007 23:13:13 -0700 Subject: [IPython-dev] Problems with physics profile In-Reply-To: <949BD319240E484B984463501CD7D17C010881F8@IMAGOEXCH1.corp.imago.com> References: <949BD319240E484B984463501CD7D17C010881F8@IMAGOEXCH1.corp.imago.com> Message-ID: [ Forgive the top-posting, but this is so old that I want to leave the full context in...] Duane, did you ever resolve this issue? I see no replies to you on-list, so I'd like to know if you sorted out the problem. It's working fine on my linux box, but I could set things up under VMWare to test this if you're still stuck. Cheers, f On Apr 11, 2007 10:40 AM, Duane Kaufman wrote: > Hi, > > I am trying to get the Physics profile working under Windows, but I get > an error when starting (see below). > I am running Ipython 0.8.0 and Scientific 2.6. Any hints as to what to > look into? > > Thanks, > Duane > > Error Follows: > > > C:\Console2>C:\Python24\python.exe C:\Python24\scripts\ipython-script.py > -pylab > -p physics > *** Simplified input for physical quantities enabled. > ------------------------------------------------------------------------ > --- > exceptions.TypeError Traceback (most > recent call > last) > > C:\Console2\ > > c:\python24\lib\site-packages\ipython-0.8.0-py2.4.egg\IPython\Extensions > \Physica > lQInteractive.py > 20 __license__ = Release.license > 21 > ---> 22 from Scientific.Physics.PhysicalQuantities import > PhysicalQuantity > 23 > 24 # This code can be set up to work with Numeric or with math for > providin > g the > > c:\python24\lib\site-packages\ScientificPython-2.6-py2.4-win32.egg\Scien > tific\Ph > ysics\PhysicalQuantities.py > 591 > 592 _addUnit('Hz', '1/s') # Hertz > --> 593 _addUnit('N', 'm*kg/s**2') # Newton > 594 _addUnit('Pa', 'N/m**2') # Pascal > 595 _addUnit('J', 'N*m') # Joule > > c:\python24\lib\site-packages\ScientificPython-2.6-py2.4-win32.egg\Scien > tific\Ph > ysics\PhysicalQuantities.py in _addUnit(name, unit) > 573 raise KeyError('Unit ' + name + ' already defined') > 574 if type(unit) == type(''): > --> 575 unit = eval(unit, _unit_table) > 576 for cruft in ['__builtins__', '__args__']: > 577 try: del _unit_table[cruft] > > C:\Console2\ > > c:\python24\lib\site-packages\ScientificPython-2.6-py2.4-win32.egg\Scien > tific\Ph > ysics\PhysicalQuantities.py in __pow__(self, other) > 367 raise TypeError("cannot exponentiate units with > non-zero off > set") > 368 if type(other) == type(0): > --> 369 return PhysicalUnit(other*self.names, > pow(self.factor, other > ), > 370 map(lambda x,p=other: x*p, > self.powers)) > > 371 if type(other) == type(0.): > > TypeError: unsupported operand type(s) for *: 'int' and 'NumberDict' > WARNING: Loading of IPython.Extensions.PhysicalQInteractive failed. > > ------------------------------------------------------------------------ > --- > exceptions.NameError Traceback (most > recent call > last) > > C:\Console2\ > > NameError: name 'PhysicalQuantityInteractive' is not defined > WARNING: Failure executing code: 'q = PhysicalQuantityInteractive' > ------------------------------------------------------------------------ > --- > exceptions.NameError Traceback (most > recent call > last) > > C:\Console2\ > > NameError: name 'PhysicalQuantityInteractive' is not defined > WARNING: Failure executing code: "g = PhysicalQuantityInteractive('9.8 > m/s**2')" > > *** q is an alias for PhysicalQuantityInteractive > *** g = 9.8 m/s^2 has been defined > *** rad = pi/180 has been defined > Python 2.4.4 (#71, Oct 18 2006, 08:34:43) [MSC v.1310 32 bit (Intel)] > Type "copyright", "credits" or "license" for more information. > > IPython 0.8.0 -- An enhanced Interactive Python. > ? -> Introduction to IPython's features. > %magic -> Information about IPython's 'magic' % functions. > help -> Python's own help system. > object? -> Details about 'object'. ?object also works, ?? prints more. > > IPython profile: physics > > Welcome to pylab, a matplotlib-based Python environment. > For more information, type 'help(pylab)'. > > NOTICE: This e-mail communication may contain privileged or other confidential information for the sole use of the designated recipients. If you have received it in error, please advise the sender by reply email and immediately delete the message and any attachments without reviewing, copying or disclosing the contents. > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://lists.ipython.scipy.org/mailman/listinfo/ipython-dev > From hans_meine at gmx.net Sun Nov 25 07:50:05 2007 From: hans_meine at gmx.net (Hans Meine) Date: Sun, 25 Nov 2007 13:50:05 +0100 Subject: [IPython-dev] [IPython-user] IPython 0.8.2 and PyReadline 1.5, getting ready for release... In-Reply-To: References: Message-ID: <200711251350.06165.hans_meine@gmx.net> On Samstag 24 November 2007, Fernando Perez wrote: > It seems to me that the multi-line history is getting mangled up in > leopard's libedit, which may behave differently than readline in some > subtle aspects. That may be a bug in our part or one in libedit, but > even if it's the latter, we might be able to work around it somehow. Just a reminder that there's another problem which I mentioned in the "Preparing for 0.8.2 release" thread in September: In autoindent-mode, one cannot paste sth. like the following: def x(): return "Test %d me %d" % ( 23, 42) (The last line will be truncated at the left.) Playing around with this, I am now able to provide an extra hint: The culprit seems to be the "return" - if you replace that with print, it works properly. Obviously, the code handles return specially, but mixes up the notions of /next input line/ and /next command line/. HTH Ciao, / / .o. /--/ ..o / / ANS ooo -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: This is a digitally signed message part. URL: From fperez.net at gmail.com Sun Nov 25 13:03:45 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Sun, 25 Nov 2007 11:03:45 -0700 Subject: [IPython-dev] Fwd: [sage-support] Re: install problem with pickleshare In-Reply-To: <85e81ba30711242322v1831589fhaa40f8ba7c7d79ad@mail.gmail.com> References: <1192514681.204773.36580@i13g2000prf.googlegroups.com> <85e81ba30710160006m2c2f3ae0g215a1eeae57095ed@mail.gmail.com> <1192548768.188492.94600@v23g2000prn.googlegroups.com> <85e81ba30710161612n1f5907dey924e0120a2989639@mail.gmail.com> <96de71860710170322qbb1bae6yd5fcb905267db4a9@mail.gmail.com> <46cb515a0710172347p11425253tb60ec0132161b03f@mail.gmail.com> <85e81ba30711242322v1831589fhaa40f8ba7c7d79ad@mail.gmail.com> Message-ID: On Nov 25, 2007 12:22 AM, William Stein wrote: > I tried updating Sage to this version of Ipython, and > when I start sage it just immediately exits: > > rank4:~ was$ sage > ---------------------------------------------------------------------- > | SAGE Version 2.8.13, Release Date: 2007-11-21 | > | Type notebook() for the GUI, and license() for information. | > ---------------------------------------------------------------------- > rank4:~ was$ > > > I.e., I get immediately dumped to the prompt. Do you > have any idea why? (I haven't tried at all to figure out why.) > > If anybody wants to try this, just do > > wget http://sage.math.washington.edu/home/was/tmp/ipython-0.8.1.r2871.spkg > sage -f ipython-0.8.1.r2871.spkg Sorry about that! Could you redo the process with the current SVN (r2873, if it matters)? Because of how Sage starts ipython, it was getting confused and going into 'batch mode' right away, thus never entering the interactive loop. Fixed now. Cheers, f From fperez.net at gmail.com Sun Nov 25 13:07:13 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Sun, 25 Nov 2007 11:07:13 -0700 Subject: [IPython-dev] [IPython-user] IPython 0.8.2 and PyReadline 1.5, getting ready for release... In-Reply-To: <200711251350.06165.hans_meine@gmx.net> References: <200711251350.06165.hans_meine@gmx.net> Message-ID: On Nov 25, 2007 5:50 AM, Hans Meine wrote: > On Samstag 24 November 2007, Fernando Perez wrote: > > It seems to me that the multi-line history is getting mangled up in > > leopard's libedit, which may behave differently than readline in some > > subtle aspects. That may be a bug in our part or one in libedit, but > > even if it's the latter, we might be able to work around it somehow. > > Just a reminder that there's another problem which I mentioned in > the "Preparing for 0.8.2 release" thread in September: > In autoindent-mode, one cannot paste sth. like the following: > > def x(): > return "Test %d me %d" % ( > 23, 42) > > (The last line will be truncated at the left.) > > Playing around with this, I am now able to provide an extra hint: The culprit > seems to be the "return" - if you replace that with print, it works properly. > Obviously, the code handles return specially, but mixes up the notions > of /next input line/ and /next command line/. Yup, and for now that will be considered a 'limitation' of the auto-pasting. The problem is that to recognize something like your example in auto-pasting mode would require actually parsing the input to identify incomplete lines, at pasting time. I don't know how to do that in a simple and sane way that doesn't complicate the already horrible auto-pasting logic further. I'd love it for that logic to be better, but I'm afraid I don't have a good solution at this point. Patches welcome, as usual. Cheers, f From fperez.net at gmail.com Sun Nov 25 14:32:40 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Sun, 25 Nov 2007 12:32:40 -0700 Subject: [IPython-dev] [IPython-user] IPython 0.8.2 and PyReadline 1.5, getting ready for release... In-Reply-To: References: <46cb515a0711230530u17dd66b1r1c17a2d32499bc7d@mail.gmail.com> Message-ID: On Nov 24, 2007 11:51 PM, wang frank wrote: > > Just curious. Has anyone tested this release with xemacs 21.4.21 and emacs > 22.1 in linux and window XP? I assume that it works out of box in linux. > However, it will really help normal user if some details on how to make it > work in window with xemacs and emacs. Actually, under gutsy emacs22 has problems loading python-mode at all, so I can't easily test it. There are fixes for the issue (it was discussed at length on the numpy list) but I'm using an outside emacs23 snapshot so I can get proper XFT font support, so I don't really care (and can't spend time on it). I used to be an XEmacs guy and had used this on and off without any problems, but I haven't used XEmacs for a few months (since I switched to CVS emacs for the font support). Now I just noticed that while XEmacs seems to load ipython just fine, it won't connect to the X server for matplotlib use (my default mode loads ipython -pylab). I have no idea why emacs is OK with X11 but not XEmacs, I'd never seen the error before. It would be *really nice* if our emacs/windows users could collect the wisdom on how to make it correctly work in that platform, and post it on the wiki. The cookbook would be a good place for this: http://ipython.scipy.org/moin/Cookbook Note that for now, I've re-enabled wiki account creation in case anyone wants to pitch in. The message on the page still says that it's disabled, but it's not :) I'll turn it off again when the next wave of spammers comes, perhaps for now the message will distract a few of these cockroaches. But if any of you is willing to help on the wiki with the emacs info, you should be able to make an account right away. > One more question. Ipython is much better than the shell comes with python > release. Why does python includes the ipython in it release? ipython is too big and too complex for realistic inclusion in the standard library, I think. I've never tried to lobby for its inclusion because I doubt they'd go for it, but feel free to ask them. I should add that if I were Guido, I'd probably say no :) (the reasons being that I wouldn't want such a large maintenance burden inside the core code just for the interactive usage cases). Cheers, f From fperez.net at gmail.com Sun Nov 25 14:35:47 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Sun, 25 Nov 2007 12:35:47 -0700 Subject: [IPython-dev] [IPython-user] IPython 0.8.2 and PyReadline 1.5, getting ready for release... In-Reply-To: References: <200711231259.55409.darren.dale@cornell.edu> <2E502158-11E3-4555-AD17-110A8AEA0E7B@gmail.com> Message-ID: On Nov 24, 2007 2:24 PM, Fernando Perez wrote: > I will get access to a Leopard machine soon (roughly a week, > hopefully), so I should be able to dig into this myself. Unless > someone has a fix they can send us, I think I'll cut out the 0.8.2 > release tomorrow, and we'll continue to work on this one in the > upcoming weeks so we can have a fully Leopard-happy ipython. > > It seems to me that the multi-line history is getting mangled up in > leopard's libedit, which may behave differently than readline in some > subtle aspects. That may be a bug in our part or one in libedit, but > even if it's the latter, we might be able to work around it somehow. I made a ticket for this, so we can track progress on the issue: http://ipython.scipy.org/ipython/ipython/ticket/199 Cheers, f From fperez.net at gmail.com Sun Nov 25 17:42:26 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Sun, 25 Nov 2007 15:42:26 -0700 Subject: [IPython-dev] [IPython-user] IPython 0.8.2 and PyReadline 1.5, getting ready for release... In-Reply-To: References: <46cb515a0711230530u17dd66b1r1c17a2d32499bc7d@mail.gmail.com> Message-ID: On Nov 25, 2007 3:36 PM, wang frank wrote: > > I really want to put something to let window users to configure the ipython > inside xemacs and emacs. Unfortunately, it is beyond my capability. The > python shell works fine with minimal effort. ipython is really tricky. After > searching the answer for several days, I am still stuck at the output > command. The ipython seems stopping to generate output for some command on > the display. Since I do not know lisp and only used the emacs and xemacs for > very short time, I really could not solve the problem. > > I may have to use the default python shell and hopping someday I could > solve this problem. I certainly didn't mean my comments as a direct request to you, and I'm sorry that Pete's suggestions haven't worked out in your setup. It was a general request to our user community, since there's a chance that (now or in the future) someone can pitch in and contribute the right bits of advice and configuration details needed to make it work. It's a shame it doesn't, and I'd love to include any fixes from our side we can to help, but until a windows+emacs guru tells us what we're doing wrong, we won't know exactly what we need to change. Cheers, f From fperez.net at gmail.com Sun Nov 25 17:48:31 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Sun, 25 Nov 2007 15:48:31 -0700 Subject: [IPython-dev] host completer patch In-Reply-To: <759C3283-7BF8-4418-97B5-E9FB78209975@afpy.org> References: <622E5C19-8EC3-47E1-B940-3E5F7DFAE02C@afpy.org> <0E51D11E-409F-400B-8A13-201647480762@afpy.org> <759C3283-7BF8-4418-97B5-E9FB78209975@afpy.org> Message-ID: On Jul 16, 2007 2:42 PM, Gael Pasgrimaud wrote: > > Le 16 juil. 07 ? 19:07, Fernando Perez a ?crit : > > > On 7/16/07, Gael Pasgrimaud wrote: > > > > This is NOT hypothetical, in fact the only breakin I've ever been a > > victim of had this pattern, and was made MUCH more serious thanks to > > an unhashed known_hosts (it was an old mismanaged Solaris box, years > > ago). > > > > Ouch ! But known_hosts is not the only file with host names. > What about ~/.ssh/config ? I don't think this one can be encrypted > one day. > > > > > But we're in the business of being useful to people, not enforcing > > policy on them. I'd be happy to include your patch if you modify it > > to include a docstring explaining the above, making very clear the > > consequences of enabling it. At least if people do this, they should > > know *exactly* the exposure they are generating. > > > > Ok, done. My english is quite poor but i've tried to be as clear as > possible. > I've also put a link to this thread since your example is really > meaningful. Thanks! I've just added this, and it will be part of 0.8.2 once we put it out (later today if I can finish flushing my queue). Cheers, f From fperez.net at gmail.com Sun Nov 25 19:25:48 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Sun, 25 Nov 2007 17:25:48 -0700 Subject: [IPython-dev] sort dynamic attributes on tab completion In-Reply-To: <96be7bb00707081558ub59b008s4ce418557560d545@mail.gmail.com> References: <96be7bb00707081558ub59b008s4ce418557560d545@mail.gmail.com> Message-ID: Hi Matthew, On Jul 8, 2007 3:58 PM, Matthew Neeley wrote: > Dynamic attributes from trait_names and _getAttributeNames are added > to the list of tab completions, but when this happens, the attribute > list is turned into a set, so the attributes are unordered when > printed, which makes it hard to find the right completion. Here's a > patch to turn this set back into a list and sort it. > > -Matthew Neeley sorry for never replying. Your patch (or at least the idea) did make it in when I added the traits support, though the function was in fact committed as genutils.dir2(): http://ipython.scipy.org/ipython/ipython/changeset/2568#file3 > PS- This is my first contribution to ipython, so I hope this is the > right place to send this. I think ipython is great software, and want > to thank you all for your hard work! Thanks to you for the kind words and the contribution. And yes, even if sometimes we do a poor job (typically me) of handling the mailing list and responding to user contributions in time, this is the right place to send them. If it's something more involved, making a trac ticket may be better, but we have a small enough project that for now I'm OK not enforcing (as larger projects often do, and with good reason) a strict rule of only working on things that have had a trac item created for them. Cheers, f From fperez.net at gmail.com Sun Nov 25 19:31:09 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Sun, 25 Nov 2007 17:31:09 -0700 Subject: [IPython-dev] Problems with external commands and help() In-Reply-To: <4681797E.7060307@users.sourceforge.net> References: <4681797E.7060307@users.sourceforge.net> Message-ID: Hi Andreas, [keeping your full original message for context, as this is old] Sorry for the delayed reply, I wonder if you found a solution to your problem. If not, you might want to try with the testing candidate for 0.8.2 from: http://ipython.scipy.org/dist/testing/ We'll be soon releasing 0.8.2 final, which will be very close to that. If your problems persist, please let us know and we'll do our best to help. Cheers, f On Jun 26, 2007 1:39 PM, Andreas M. wrote: > Hello, > > I use > > Py 2.4.3 (#69, Apr 11 2006, 15:32:42) [MSC v.1310 32 bit (Intel)] IPy 0.8.1 > > on WindowsXP Pro/SP2. > > I have defined the %HOME% env-var. I also have the %EDITOR% set. > None of the !commands work. When trying to invoke the external editor I > get this: > > [amix]|9> %edit > IPython will make a temporary file named: > h:\amix\temp\ipython_edit_ofs9wj.py > Editing... done. Executing edited code... > Could not open file for safe > execution. > WARNING: File not found. Did you forget to save? > > > No editor popped up! > > When I open the help utility it goes like this (example): > > > In [2]: import datetime > > help> modules datetime > > Here is a list of matching modules. Enter any module name to get more help. > > datetime - Fast implementation of the datetime type. > test.test_datetime - Test date/time type. > zope.interface.common.idatetime - Datetime interfaces. > zope.interface.common.tests.test_idatetime - Test for datetime interfaces > > help> datetime > > help> > > Also, doing help('datetime') or help(datetime) does not yield any > results, nor does it with any other module. > > My %PYTHONPATH% is set and so is %PYTHONDOCS%. > > I tried it both with CMD.EXE as backend-shell as well as Powershell 1.0 > > Any ideas ? > > -- > Andreas M. > > _______________________________________________ > IPython-dev mailing list > IPython-dev at scipy.org > http://lists.ipython.scipy.org/mailman/listinfo/ipython-dev > From fperez.net at gmail.com Sun Nov 25 19:32:24 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Sun, 25 Nov 2007 17:32:24 -0700 Subject: [IPython-dev] [IPython-user] IPython within wxshell In-Reply-To: References: Message-ID: Hi Tom, On May 29, 2007 8:49 PM, Tom Denniston wrote: > Has anyone had any success with IPython within pycrust or the wx pyshell. > > I tried the very naive: > > sh = Instance(IPython.Shell.IPythonShellEmbed) > sh = IPython.Shell.IPythonShellEmbed() > sh (['-pylab']) > > > > within the enthought pyface shell. I also tried > Shell.IPShellMatplotlibWX. Both start IPython but the escape > characters for colors don't get interpretted colors. If I turn the > colors off I get some working Ipython behavior, but most behavior > doesn't work. Tab completion doesn't work and command history, for > instance do not work. > > Moreover the standard output of command like help seems to go to my > text console rather than the wx shell console. > > Has anyone gotten this combination to work? Are there any wx related > examples or wiki pages? I noticed one on open gl and another on > pygtk--both in the cookbook but nothing on wx. > > Is there a reasonably easy way to do this now or is this something > that ipython does not yet support? Did you ever get this to work by modifying the GTK idea? If so, I'd love to hear about it and perhaps we could have a set of companion recipes on the cookbook for the different toolkits. Cheers, f From fperez.net at gmail.com Sun Nov 25 22:30:41 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Sun, 25 Nov 2007 20:30:41 -0700 Subject: [IPython-dev] [IPython-user] IPython 0.8.2 and PyReadline 1.5, getting ready for release... In-Reply-To: References: <46cb515a0711230530u17dd66b1r1c17a2d32499bc7d@mail.gmail.com> Message-ID: On Nov 25, 2007 12:32 PM, Fernando Perez wrote: > I used to be an XEmacs guy and had used this on and off without any > problems, but I haven't used XEmacs for a few months (since I switched > to CVS emacs for the font support). Now I just noticed that while > XEmacs seems to load ipython just fine, it won't connect to the X > server for matplotlib use (my default mode loads ipython -pylab). I > have no idea why emacs is OK with X11 but not XEmacs, I'd never seen > the error before. Just to clarify: I've realized the problem was not with XEmacs. Instead, my X11 session decided to stop accepting new connections, and XEmacs was the first program that happened to throw the error: Xlib: connection to ":0.0" refused by server Xlib: Maximum number of clients reached Error: Can't open display: :0 Later I realized that *nothing* was working (meaning, no new X11 programs could be started), and I just happened to blame Xemacs for the failure by coincidence. I'd never seen that in my life, and I had to do a full X11 restart so I could open new programs. I just started using Gutsy a few days ago, and I'm seeing a few funny things here and there... Anyway, the summary is that on gutsy, both XEmacs and Emacs23 are OK with ipython. Emacs22 has a problem with python-mode.el, but there are known workarounds for that (and the problem has nothing to do with ipython). Cheers, f From fperez.net at gmail.com Mon Nov 26 03:49:06 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Mon, 26 Nov 2007 01:49:06 -0700 Subject: [IPython-dev] 0.8.2.rc ready Message-ID: Hi all, I've managed to flush everything I had pending but I'm a bit tired to actually pay attention to release details now, so I've just uploaded 0.8.2 as a release candidate to: http://ipython.scipy.org/dist/testing/ Unless someone spots a problem or complains, I think we'll make this the final 0.8.2 tomorrow/Tuesday, along with the pyreadline 1.5 that's in there. The remaining Leopard issues, we'll deal with for 0.8.3 soon, but this already at least works better than 0.8.1 for Leopard (in addition to all the other changes). Cheers, f From dkaufman at imago.com Mon Nov 26 09:39:24 2007 From: dkaufman at imago.com (Kaufman, Duane) Date: Mon, 26 Nov 2007 08:39:24 -0600 Subject: [IPython-dev] Problems with physics profile In-Reply-To: References: <949BD319240E484B984463501CD7D17C010881F8@IMAGOEXCH1.corp.imago.com> Message-ID: <949BD319240E484B984463501CD7D17C0100F1FD@IMAGOEXCH1.corp.imago.com> Hello, Yes, this problem has been solved. I believe it was a compatibility problem between the Scientific package and the Numpy/Numeric installs I had on that machine. Thanks for getting back to me, Duane > -----Original Message----- > From: Fernando Perez [mailto:fperez.net at gmail.com] > Sent: November 25, 2007 00:13 > To: Kaufman, Duane > Cc: ipython-dev at scipy.org > Subject: Re: [IPython-dev] Problems with physics profile > > [ Forgive the top-posting, but this is so old that I want to > leave the full context in...] > > Duane, did you ever resolve this issue? I see no replies to > you on-list, so I'd like to know if you sorted out the > problem. It's working fine on my linux box, but I could set > things up under VMWare to test this if you're still stuck. > > Cheers, > > f > > On Apr 11, 2007 10:40 AM, Duane Kaufman wrote: > > Hi, > > > > I am trying to get the Physics profile working under Windows, but I > > get an error when starting (see below). > > I am running Ipython 0.8.0 and Scientific 2.6. Any hints as > to what to > > look into? > > > > Thanks, > > Duane > > > > Error Follows: > > > > > > C:\Console2>C:\Python24\python.exe > > C:\Python24\scripts\ipython-script.py > > -pylab > > -p physics > > *** Simplified input for physical quantities enabled. > > > ---------------------------------------------------------------------- > > -- > > --- > > exceptions.TypeError Traceback (most > > recent call > > last) > > > > C:\Console2\ > > > > > c:\python24\lib\site-packages\ipython-0.8.0-py2.4.egg\IPython\Extensio > > ns > > \Physica > > lQInteractive.py > > 20 __license__ = Release.license > > 21 > > ---> 22 from Scientific.Physics.PhysicalQuantities import > > PhysicalQuantity > > 23 > > 24 # This code can be set up to work with Numeric or with math > > for providin g the > > > > > c:\python24\lib\site-packages\ScientificPython-2.6-py2.4-win32.egg\Sci > > en > > tific\Ph > > ysics\PhysicalQuantities.py > > 591 > > 592 _addUnit('Hz', '1/s') # Hertz > > --> 593 _addUnit('N', 'm*kg/s**2') # Newton > > 594 _addUnit('Pa', 'N/m**2') # Pascal > > 595 _addUnit('J', 'N*m') # Joule > > > > > c:\python24\lib\site-packages\ScientificPython-2.6-py2.4-win32.egg\Sci > > en > > tific\Ph > > ysics\PhysicalQuantities.py in _addUnit(name, unit) > > 573 raise KeyError('Unit ' + name + ' already defined') > > 574 if type(unit) == type(''): > > --> 575 unit = eval(unit, _unit_table) > > 576 for cruft in ['__builtins__', '__args__']: > > 577 try: del _unit_table[cruft] > > > > C:\Console2\ > > > > > c:\python24\lib\site-packages\ScientificPython-2.6-py2.4-win32.egg\Sci > > en > > tific\Ph > > ysics\PhysicalQuantities.py in __pow__(self, other) > > 367 raise TypeError("cannot exponentiate units with > > non-zero off > > set") > > 368 if type(other) == type(0): > > --> 369 return PhysicalUnit(other*self.names, > > pow(self.factor, other > > ), > > 370 map(lambda x,p=other: x*p, > > self.powers)) > > > > 371 if type(other) == type(0.): > > > > TypeError: unsupported operand type(s) for *: 'int' and 'NumberDict' > > WARNING: Loading of IPython.Extensions.PhysicalQInteractive failed. > > > > > ---------------------------------------------------------------------- > > -- > > --- > > exceptions.NameError Traceback (most > > recent call > > last) > > > > C:\Console2\ > > > > NameError: name 'PhysicalQuantityInteractive' is not defined > > WARNING: Failure executing code: 'q = PhysicalQuantityInteractive' > > > ---------------------------------------------------------------------- > > -- > > --- > > exceptions.NameError Traceback (most > > recent call > > last) > > > > C:\Console2\ > > > > NameError: name 'PhysicalQuantityInteractive' is not defined > > WARNING: Failure executing code: "g = > PhysicalQuantityInteractive('9.8 > > m/s**2')" > > > > *** q is an alias for PhysicalQuantityInteractive > > *** g = 9.8 m/s^2 has been defined > > *** rad = pi/180 has been defined > > Python 2.4.4 (#71, Oct 18 2006, 08:34:43) [MSC v.1310 32 > bit (Intel)] > > Type "copyright", "credits" or "license" for more information. > > > > IPython 0.8.0 -- An enhanced Interactive Python. > > ? -> Introduction to IPython's features. > > %magic -> Information about IPython's 'magic' % functions. > > help -> Python's own help system. > > object? -> Details about 'object'. ?object also works, ?? > prints more. > > > > IPython profile: physics > > > > Welcome to pylab, a matplotlib-based Python environment. > > For more information, type 'help(pylab)'. > > > > NOTICE: This e-mail communication may contain privileged or > other confidential information for the sole use of the > designated recipients. If you have received it in error, > please advise the sender by reply email and immediately > delete the message and any attachments without reviewing, > copying or disclosing the contents. > > > > _______________________________________________ > > IPython-dev mailing list > > IPython-dev at scipy.org > > http://lists.ipython.scipy.org/mailman/listinfo/ipython-dev > > > From fperez.net at gmail.com Mon Nov 26 13:40:54 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Mon, 26 Nov 2007 11:40:54 -0700 Subject: [IPython-dev] Problems with physics profile In-Reply-To: <949BD319240E484B984463501CD7D17C0100F1FD@IMAGOEXCH1.corp.imago.com> References: <949BD319240E484B984463501CD7D17C010881F8@IMAGOEXCH1.corp.imago.com> <949BD319240E484B984463501CD7D17C0100F1FD@IMAGOEXCH1.corp.imago.com> Message-ID: On Nov 26, 2007 7:39 AM, Kaufman, Duane wrote: > Hello, > > Yes, this problem has been solved. I believe it was a compatibility > problem between the Scientific package and the Numpy/Numeric installs I > had on that machine. Excellent. At this point pretty much everyone has caught up with numpy and Numeric is truly behind us, so hopefully these issues will be soon gone. > Thanks for getting back to me, A bit late, but still :) Regards, f From fperez.net at gmail.com Tue Nov 27 19:17:59 2007 From: fperez.net at gmail.com (Fernando Perez) Date: Tue, 27 Nov 2007 17:17:59 -0700 Subject: [IPython-dev] IPython Crash Report In-Reply-To: <46cb515a0711271520u70ecd60dgbeb461708501d2c0@mail.gmail.com> References: <9279508.9577541195653401262.JavaMail.root@vms229.mailsrvcs.net> <46cb515a0711271520u70ecd60dgbeb461708501d2c0@mail.gmail.com> Message-ID: Thanks for the reports, Ville. The one from the user on 0.7.1 is baffling, because he's not using any of the gui or thread support, but that's also a fairly old version, so who knows. The one from 0.8.1 is using pylab is interesting, and I was just able to reproduce it. It's certainly an annoyance, but the interesting thing is that ipython doesn't actually crash! What happens is that if by chance the KeyboardInterrupt fires exactly at the time that the Tk event loop is running, the Tk exception handler raises the exception directly into sys.excepthook, which is our crash handler. So we generate a report as if there had been a crash, when in reality ipython is perfectly OK. I don't see a way to prevent this, but you can tell from the reports when this happens because you'll see *just* the Tkinter part of the traceback and nothing else: A problem occured executing Python code. Here is the sequence of function calls leading up to the error, with the most recent (innermost) call last. /home/fperez/test/lib-tk/Tkinter.py in __call__(self=, *args=('692', '??', '??', '??', '??', '16', '284484270', '??', '572', '324', '??', '0', '??', '??', '.147924620', '6', '654', '994', '??')) 1386 return self.tk.call('image', 'names') 1387 1388 def image_types(self): 1389 """Return a list of all available image types (e.g. phote bitmap).""" 1390 return self.tk.call('image', 'types') 1391 1392 1393 class CallWrapper: 1394 """Internal class. Stores function to call when some user 1395 defined Tcl function is called e.g. after an event occurred.""" 1396 def __init__(self, func, subst, widget): 1397 """Store FUNC, SUBST and WIDGET as members.""" 1398 self.func = func 1399 self.subst = subst 1400 self.widget = widget -> 1401 def __call__(self, *args): global __call__ = undefined self = args = ('692', '??', '??', '??', '??', '16', '284484270', '??', '572', '324', '??' , '0', '??', '??', '.147924620', '6', '654', '994', '??') 1402 """Apply first function SUBST to arguments, than FUNC.""" 1403 try: 1404 if self.subst: 1405 args = self.subst(*args) 1406 return self.func(*args) 1407 except SystemExit, msg: 1408 raise SystemExit, msg 1409 except: 1410 self.widget._report_exception() 1411 1412 1413 class Wm: 1414 """Provides functions for the communication with the window manager.""" 1415 1416 def wm_aspect(self, KeyboardInterrupt: This case isn't even caused by our own handling, since for Tk, the threads stuff doesn't activate. The Python interpreter has wired the Tk event loop into its own, so this race condition can probably be triggered by *any* python program that activates the Tkinter timer callbacks. If you get one of these that has wx, gtk or qt code in them, please send them also my way. I'd like to have a look at the structure of the traceback, to see if any ideas come up. Thanks again for the info, f From jorgen.stenarson at bostream.nu Fri Nov 30 14:03:06 2007 From: jorgen.stenarson at bostream.nu (=?ISO-8859-1?Q?J=F6rgen_Stenarson?=) Date: Fri, 30 Nov 2007 20:03:06 +0100 Subject: [IPython-dev] PyReadline 1.5 released Message-ID: <47505E6A.3020407@bostream.nu> Hi all, the release version of pyreadline 1.5 is now available at: http://ipython.scipy.org/dist/ Thanks to Fernando for building and uploading New feature: - Improved support for non-ascii characters by switching to the unicode functions for cummunication with the console. Thanks to pan for all help in getting this to work. This feature works for me in my day to day work but there may still be codepaths that have not been excercised by me. Especially in combination with ipython there can still be problems lurking. /J?rgen