In my quest to add a command (actually a set of subcommands introduced by a single prefix command) to Hercules, I decided the cckd command might be a suitable model to follow so I went to check out it how it works. When I tried a known invalid cckd command on current Hyperion, I got this:
HHC01603I cckd helloHHC00348E CCKD file: invalid value 13725897 for hello
The problem seems to be there at least as far back as May 2015 but using a Hyperion executable from June 2013 which I happened to have on hand from a previous battle, I get this more appropriate response:
HHC01603I cckd helloHHC00349E CCKD file: invalid cckd keyword: hello
followed by some help text.
As this is the second issue I've stumbled across with processing of invalid console commands in a few days without actually looking for them, I wonder if there has been an overhaul of command processing in general and a bit more testing is required after it?
Getting sidetracked is one of my specialties and I can't afford to get sidetracked into doing a review of all command processing right now so I'm just making this note here to raise awareness in case there may be other similar problems lurking in the code which would be better treated all at the same time rather than fixing them one by one.
Unlike the previous crash-hercules-in-two-keystrokes issue, I don't think there is the same urgency in getting this particular one fixed.
In my quest to add a command (actually a set of subcommands introduced by a single prefix command) to Hercules, I decided the cckd command might be a suitable model to follow so I went to check out it how it works. When I tried a known invalid cckd command on current Hyperion, I got this:
HHC01603I cckd helloHHC00348E CCKD file: invalid value 13725897 for helloThe problem seems to be there at least as far back as May 2015 but using a Hyperion executable from June 2013 which I happened to have on hand from a previous battle, I get this more appropriate response:
HHC01603I cckd helloHHC00349E CCKD file: invalid cckd keyword: hellofollowed by some help text.As this is the second issue I've stumbled across with processing of invalid console commands in a few days without actually looking for them, I wonder if there has been an overhaul of command processing in general and a bit more testing is required after it?
Getting sidetracked is one of my specialties and I can't afford to get sidetracked into doing a review of all command processing right now so I'm just making this note here to raise awareness in case there may be other similar problems lurking in the code which would be better treated all at the same time rather than fixing them one by one.
Unlike the previous crash-hercules-in-two-keystrokes issue, I don't think there is the same urgency in getting this particular one fixed.