| 08-13-2009, 02:17 PM | #1 |
Is there a way to be sure that they switched to checking the return type in 1.24b ? Should anyone reverse engineer that part of the new patch or find a false negative/positive please post it so we can get confirmations. (Make sure you got the latest patch from westfall before posting, damnit) |
| 08-14-2009, 06:19 PM | #2 |
No, I don't know. But this: JASS:function H2I takes handle h returns integer return h call DoNothing() return 0 endfunction still works. |
| 08-14-2009, 06:40 PM | #3 |
Integer2Code, Code2Integer still works also ? |
| 08-14-2009, 06:45 PM | #4 | |
Quote:
JASS:function H2C takes handle h returns code return h set h = h return null endfunction |
| 08-14-2009, 06:52 PM | #5 | |
Quote:
Hindyfoo', set h = h didn't work in 1.23b/1.24, but call DoNothing() did (no idea why, seems no rhyme or reason there). However, Mr.Malte, did you actually test if it works *properly*, rather than just compiles? And, if no, can you use the same countering methods of wrapper functions that we did last time? |
| 08-14-2009, 06:53 PM | #6 |
Lies, set h = h always worked. In fact it's what I used, nub, because it was shorter to write. I'm 1000% sure that it worked. Also, DoNothing() doesn't work either, fgt. I tested all of the possible return bug workarounds that were established, and neither of them worked: JASS:return h set h = h return 0 if true then return h endif return 0 return h call DoNothing() return 0 |
| 08-14-2009, 07:03 PM | #7 |
Be sure to test it with the official editor HINDYhat. AFAIK with the actual official patch the new jasshelper (j.0) detect it as an error and official editor doesn't. |
| 08-14-2009, 07:05 PM | #8 |
I'm using 0.9.I.2, so JH does not detect anything. |
| 08-14-2009, 07:08 PM | #9 |
Ok, but did you try only integer <-> code ? Maybe it is the exception and still works with handle <-> integer ? |
| 08-14-2009, 07:09 PM | #10 | ||
Quote:
Well, the beta one does. Vanilla WE gave compilation errors, though it gave wrong line numbers. NewGen with JH still compiled it, but 1.24b refused to run. Quote:
I'm quite sure H<->I doesn't work. At my first test it seemed to work as return null worked fine, but it didn't when I fed actual handle variable. It seems that JH compiles null into integer when applicable... |
| 08-14-2009, 07:12 PM | #11 |
Troll-Brain jesus I tried everything okay? |
| 08-14-2009, 07:15 PM | #12 | |
Quote:
|
| 08-14-2009, 07:42 PM | #13 | |||
Quote:
Quote:
Quote:
Must be a real pain in the ass to be around you in real life. |
| 08-14-2009, 07:54 PM | #14 |
<3 |
| 08-16-2009, 08:10 AM | #15 |
OFF-TOPIC The war3 crash when we use the index 8191 of an array variable, and you are loading a saved game still exist ? I guess yes, just to know. |
