[Magic-dev] Re: A magic bug

[R. Timothy Edwards]

Eric Liu Wong wrote:

> I think I found a bug in magic about highlighting connections.  My magic system is quite recent. Please open the
> attachment and see the readme file for detailed information on my system, techfile and the way to reproduce the
 > problem.

In or around line 701 of database/DBcontact.c, change:

    if (TTMaskEqual(rMask, DBResidueMask(ctype)))

to

    if (TTMaskIntersect(rMask, DBResidueMask(ctype)))

I think my original assumption was that overlapping contacts would
be replaced by a stacking type, but when the contacts are in separate
subcells, the contacts can overlap without generating a stacking
type.  So, TTMaskEqual() was producing a mask of non-connecting
types that included a lot of types that really connect to the
contact.  I didn't investigate why magic hangs and starts gobbling
memory;  the connectivity search routine is known to be very tricky
and is very hard to debug.  When the above fix seemed to fix the
problem, I stopped investigating it further.  If you find any other
problems, let me know.

I have posted the fix as part of magic-7.3.115.

					Regards,
					Tim

+--------------------------------+-------------------------------------+
| Dr. R. Timothy Edwards (Tim)   | email: tim.edwards at multigig.com     |
| MultiGiG, Inc.                 | web:   http://www.multigig.com      |
| 100 Enterprise Way, Suite A-3  | phone: (831) 621-3283               |
| Scotts Valley, CA 95066        | cell:  (240) 401-0616               |
+--------------------------------+-------------------------------------+