On Wed, Sep 16, 2009 at 03:05:43PM +0200, Frank Lübeck wrote: > On Mon, Sep 14, 2009 at 07:10:29PM +0200, Bill Allombert wrote: > > Dear GAP team, > > > > I received the following bug report from a Debian user > > (This is Debian bug #546579 <http://bugs.debian.org/546579> > > hence public) > > > > The issue is that under the circumstance below, GAP display: > > > > Panic: tried to print a path of unknown type '72' at > > Print( "Group(", l, ")" ); > > > > The version of GAP used was 4.4.12 on amd64-linux (64bit). > > Dear Bill, Joachim and Others, > > This is a known effect in GAP, and I would classify it as "inconvenience" > rather than a "bug". It can occur when the printing of a large object is > interrupted with Ctrl-C. > > What happens is that GAP while printing objects with subobjects somehow > marks the subobjects which are currently printed. This is necessary to avoid > infinite loops because GAP objects can contain themselves as subobjects. > > If such printing is interrupted GAP has no chance to clean up these marks, > and this leads to corrupted representations of the corresponding objects > in memory. Further computations with these objects should be fine but you > can no longer print them. > > Maybe, at some stage, we will change the printing methods such that such > marking of objects on GAP-kernel level is avoided. > > > > Cheers, > > Bill > > > > ----- Forwarded message from Joachim Breitner <[email protected]> ----- > > > > gap> G := GL(4,2); > > gap> SG := Group(Filtered(G, m -> Z(2)*[1,1,1,1]*m=Z(2)*[1,1,1,1])); > > gap> rcs := RightCosets(G,SG);; > > gap> rcs[1]; > > I cannot resist to propose a much more efficient way to do this: > > G := GL(4,2); > v := Z(2)^0*[1,1,1,1]; > SG := Stabilizer(G, v); > rcs := RightCosets(G,SG); # (now it is difficult to interrupt during printing) > > Best regards, > > Frank
Hello Joachim, Given that upstream know about the problem and does not have plan to adress it, I suggest to close this report. Cheers, -- Bill. <[email protected]> Imagine a large red swirl here. -- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected]
