| View previous topic :: View next topic |
| Author |
Message |
ixle
Joined: 15 Sep 2011
Posts: 133
Location: United States
|
 Posted: Fri Jan 13, 2012 3:14 pm Post subject: #shortest path: unknown error since 2.00.8 |
 |
|
I'm getting an error since upgrading to 2.00.8 from 2.00.8b4: #SHORTEST PATH: UNKNOWN ERROR.
This happens when I attempt to find a path through a portal I've linked from my list of portals (used #map dig for each item to each room it goes to). It seems to be happening regardless of room name, area name, or vnum search.
I use a complex set of aliases to link/unlink portals and find a path to a room/area/vnum, so it's rather hard for me to easily provide more background without dumping a lot of code here. Hopefullly this is enough to get started on. |
|
| Back to top |
|
 |
Slysven
Joined: 10 Apr 2011
Posts: 254
Location: As "Jomin al'Bara" in WoTMUD or Wiltshire, UK
|
| Posted: Fri Jan 13, 2012 6:27 pm Post subject: |
|
|
Well, perusing the code it seems that this message occurs when, in the shortest_path() function the searchgrid_find() HAS found a route from the current room to the destination when given the current room and destination room directly but a second procedure using this function that checkes the intermediate rooms to be valid VOID rooms (NOW which are to have EXACTLY TWO exits) does NOT.
So, I'd check that your void rooms are set correctly and that any rooms or NOW EXITS that have AVOID flags on are set correctly. Incidentally the second use of searchgrid_find() in the code that I mention has a cryptic commant: "Slower than a backtrace, but works with mazes." I wonder what this means, could the code actually check the route in the reverse direction instead, what does this mean for your portal routes if they are set to be one way? |
|
| Back to top |
|
|
ixle
Joined: 15 Sep 2011
Posts: 133
Location: United States
|
| Posted: Fri Jan 13, 2012 7:52 pm Post subject: |
|
|
| Slysven wrote: | Well, perusing the code it seems that this message occurs when, in the shortest_path() function the searchgrid_find() HAS found a route from the current room to the destination when given the current room and destination room directly but a second procedure using this function that checkes the intermediate rooms to be valid VOID rooms (NOW which are to have EXACTLY TWO exits) does NOT.
So, I'd check that your void rooms are set correctly and that any rooms or NOW EXITS that have AVOID flags on are set correctly. Incidentally the second use of searchgrid_find() in the code that I mention has a cryptic commant: "Slower than a backtrace, but works with mazes." I wonder what this means, could the code actually check the route in the reverse direction instead, what does this mean for your portal routes if they are set to be one way? |
I don't have any VOID rooms that I know of. It still finds a path and runs it using the portal, just the last few rooms after the portal don't go to the correct room somehow. |
|
| Back to top |
|
|
Scandum
Site Admin
Joined: 03 Dec 2004
Posts: 3274
|
| Posted: Fri Jan 13, 2012 9:27 pm Post subject: |
|
|
Would you mind downloading the beta source code again and see if that fixes it?
If that's the case let me know so I can compare 2.00.8 with the beta, which should narrow things down quite a bit. |
|
| Back to top |
|
|
ixle
Joined: 15 Sep 2011
Posts: 133
Location: United States
|
| Posted: Fri Jan 13, 2012 10:23 pm Post subject: |
|
|
| Scandum wrote: | Would you mind downloading the beta source code again and see if that fixes it?
If that's the case let me know so I can compare 2.00.8 with the beta, which should narrow things down quite a bit. |
Yeah, in the 2.00.8 beta works without an error message, finds the right vnum. I still had the 2.00.8b4 src saved. |
|
| Back to top |
|
|
Scandum
Site Admin
Joined: 03 Dec 2004
Posts: 3274
|
| Posted: Fri Jan 13, 2012 11:36 pm Post subject: |
|
|
Unfortunately there are substantial differences between the two versions.
Would you mind emailing your map file to mudclient@gmail.com along with the required commands to reproduce the error? |
|
| Back to top |
|
|
ixle
Joined: 15 Sep 2011
Posts: 133
Location: United States
|
| Posted: Sat Jan 14, 2012 12:29 pm Post subject: |
|
|
| Scandum wrote: | Unfortunately there are substantial differences between the two versions.
Would you mind emailing your map file to mudclient@gmail.com along with the required commands to reproduce the error? |
Mail sent. |
|
| Back to top |
|
|
Scandum
Site Admin
Joined: 03 Dec 2004
Posts: 3274
|
|
| Back to top |
|
|
ixle
Joined: 15 Sep 2011
Posts: 133
Location: United States
|
| Posted: Sun Jan 15, 2012 2:13 am Post subject: |
|
|
| So far so good. |
|
| Back to top |
|
|
ixle
Joined: 15 Sep 2011
Posts: 133
Location: United States
|
| Posted: Sun Jan 15, 2012 2:35 am Post subject: |
|
|
Spoke too soon :(
Same error from vnum 20472 using command 'hf in ghostly fire' |
|
| Back to top |
|
|
Scandum
Site Admin
Joined: 03 Dec 2004
Posts: 3274
|
| Posted: Sun Jan 15, 2012 11:36 am Post subject: |
|
|
Looks like it wasn't properly handling the AVOID flag, I updated the beta with a fix.
Thanks for the help.  |
|
| Back to top |
|
|
|
FAQ
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
