Re: raising probes with heatbugs & swarm-1.0.3 (addedenum)

[Rick Riolo]

Glen,
Thanks for checking this out and supplying a patch.

>From my point of view, these instructions on how to fix is fine (though 
you might want to create a patch file so everyone else
can easily update their 1.0.3's).

re testing procedures:
Its a tough call, at this point, given the limited resources.
I'd say NOT use your time on testing on all the platforms.
(But don't go on vacation just after a release...)

And put on "to-do" list the goal of developing completely automated
testing, eg using Expect or some other tools.  Then you can ask some
lucky (aka gullible) volunteer to do the tests on the platforms
you don't have handy access to.

- r

Rick Riolo                           address@hidden
Program for Study of Complex Systems (PSCS)
4068 Randall Lab                
University of Michigan         Ann Arbor MI 48109-1120
Phone: 313 763 3323                  Fax: 313 763 9267
http://pscs.physics.lsa.umich.edu/PEOPLE/rlr-home.html
http://pscs.physics.lsa.umich.edu//pscs.html

On Fri, 26 Sep 1997, glen e. p. ropella wrote:

> Date: Fri, 26 Sep 1997 23:32:35 -0600
> From: glen e. p. ropella <address@hidden>
> To: address@hidden
> Subject: raising probes with heatbugs & swarm-1.0.3 (addedenum)
> 
> Rick Riolo writes:
>  > Looking at the compile and linking, I can't see any differences
>  > other than swarm going from 1.0.1 and 1.0.3 .
>  > 
>  > Glen, I assume you can bring up a heatbug with a click on your machine, 
>  > yes?  How about on other architectures?
> 
> Yep!  I can do it on both my linux box and my solaris box.
> 
> However, I changed the ZoomRaster.* handleButton method from:
> 
>   -handleButton: (int) n X: (long) x Y: (long) y 
> 
> to:
> 
>   -handleButton: (int) n X: (int) x Y: (int) y 
> 
> And in the ZoomRaster.m file, the method as a whole has to change
> from:
> 
> -handleButton: (int) n X: (long) x Y: (long) y {
>   return [super handleButton: n X: (x / (long)zoomFactor) Y: (y / 
> (long)zoomFactor)];
> }
> 
> To:
> 
> -handleButton: (int) n X: (int) x Y: (int) y {
>   return [super handleButton: n X: (x / (int)zoomFactor) Y: (y / 
> (int)zoomFactor)];
> }
> 
> I also checked the Alpha, and sure enough, it crashed when probing.  And
> the changes above fixed it.  Now, why?  I imagine it has something to 
> do with left over trash in memory addresses, but, I blame that for 
> everything. <grin>
> 
> So, is this something I need to fix immediately?  Or is it something that
> can wait until 1.0.4?  (I'll probably begin making 1.0.4 right after
> I finish with the swarmdocs-1.0.3.)
> 
> Also, this shows a hole in my testing process.  What I've been doing
> is a 3 phase process.
> 
> 1) once all the changes are in and seem to be working, I compile all
> the apps I can get my hands on for either the Solaris or the linux
> platforms (depends on whether I'm at home when I get to this step or
> at work),
> 
> 2) if those all worked, I begin compiling the package and the
> swarmapps on the other machines, otherwise I make fixes and revert to
> step 1 again,
> 
> 3) if step 2 went smoothly, I come back to the base platform (either
> linux or solaris) and run the swarmapps through their individual
> tests, like probing, drag&drop, etc.  And I have a few modified
> swarmapps that exercise certain features that I compile and run.
> 
> Obviously, this bug slipped through because the probing works on
> both linux and solaris, but fails on HPs and Alphas (I checked
> the Ultra, btw and it doesn't show the problem).  So, the natural
> thing to add would be to do all the tests on all the platforms.
> But, that's a *real* problem because even a single step of heatbugs
> on the alpha in australia takes a good 5 minutes.
> 
> Any ideas?
> 
> glen
> -- 
> {glen e. p. ropella <address@hidden> |  Send lawyers, guns, and money!  }
> {Hive Drone, SFI Swarm Project         |            Hail Eris!            }
> {http://www.trail.com/~gepr/home.html  |               =><=               }
> 
>                   ==================================
>    Swarm-Support is for discussion of the technical details of the day
>    to day usage of Swarm.  For list administration needs (esp.
>    [un]subscribing), please send a message to <address@hidden>
>    with "help" in the body of the message.
>                   ==================================
> 

                  ==================================
   Swarm-Support is for discussion of the technical details of the day
   to day usage of Swarm.  For list administration needs (esp.
   [un]subscribing), please send a message to <address@hidden>
   with "help" in the body of the message.
                  ==================================