Headless Comanche Swiki?

Roger Whitney whitney at cs.sdsu.edu
Thu Sep 21 06:00:35 UTC 2000

Here is a stack trace of trying to run a Squeak image
headless on Solaris. 

eli 23-> squeak -headless test.image

Segmentation fault

11479668 CursorWithMask>beCursor
11479576 Cursor class>currentCursor:
11479484 InputSensor>currentCursor:
11479392 Cursor>show
11435120 SystemDictionary>snapshot:andQuit:
11435028 ScreenController>snapshotAndQuit
11434584 SelectionMenu>invokeOn:
11434492 ScreenController>controlActivity
11434400 Controller>controlLoop
11434308 Controller>startUp
11434096 [] in ControlManager>activeController:
11434188 [] in BlockContext>newProcess
eli 24-> 

Asking the cursor class to show a cursor is not going 
to much good in headless mode. Not clear to me why
headless mode is a VM feature. In VisualWorks there is a 
parcel (Smalltalk code) that one uses to make an image 
headless. It is a bit more work to start a headless image,
but the solution is more portable. 

> On Wed, Sep 20, 2000 at 08:15:35PM +0200, Cees de Groot wrote:
> > I'm about to install Comanche Swiki on our Intranet (Solaris box),
> > and the last little hurdle to take is to try to run it headless. The
> > -headless mode doesn't work, it results in a segmentation fault and a
> > stack(?) trace immediately.
> >
> I do not know what the problem is, but there is a good chance that it's
> my fault, as I contributed the hacks to Ian's sqXWindow.c to support
> the '-headless' option. Unfortunately, I do not have access to a Solaris
> machine, and I only tested the code on Linux.
> I recommend that you get a copy of Ian's source distribution, and build
> a headless VM from sources. This will be better for your purposes,
> because the X code will be removed entirely from the headless VM,
> and I'm sure you will not have any problems with that version. If
> you are not comfortable building the headless VM yourself, ask on
> the list and I'm sure someone with a Solaris box can send you one.
> If you have the time and patience to try running your existing VM
> under a debugger, please let me know where the segmentation fault
> is occurring, and I'll try to figure it out from there.
> Dave

Roger Whitney
San Diego State University
whitney at cs.sdsu.edu

More information about the Squeak-dev mailing list