|
Cheat Engine The Official Site of Cheat Engine
|
View previous topic :: View next topic |
Author |
Message |
shadowsinrain How do I cheat? Reputation: 0
Joined: 28 Jun 2015 Posts: 2
|
Posted: Sun Jun 28, 2015 3:58 pm Post subject: Feature request: pointerscan agreement-based heuristics |
|
|
Hello.
Feature #1.
When 80% of pointerscan results are pointing to same address, I think it is obvious which address I am goint to rescan for. Please, make most "popular" address to fill in automatically in "Rescan memory" window.
Then for most situations (I believe so) it will become no-brainer to spam "Rescan memory" + enter. Automatic incremental file naming will help too.
Feature #2.
Now, if pointerscan already understands what to scan for, unlike normal scan, why not allow automatic filtering in background?
Of course there should be some sanity checks, so when game temporarily goes into another state, where interesting locations are unavaliable, f.e. loading new level or quiting to main menu. I think these conditions may be stated as:
* At least X% of pointers are valid.
* At least Y% pointing to same address.
GUI might look like this:
Code: |
Current quess: 0x1A3B5C7D |Refresh|
Min % of valid pointers: [____]
Min % of agreement: [____]
|One step| |Start| |Stop| |Cancel|
|
Feature #3.
Sometimes first pointerscan with simple settings gives list of addreses, where each of them works with like 80% probability. If I want reliable result, I should waste N hours in pointerscan. But why not use same heuristics here too?
Let user load multiple pointer chains into one row of CE table, and let CE to guess which of those are correct.
|
|
Back to top |
|
|
mgr.inz.Player I post too much Reputation: 218
Joined: 07 Nov 2008 Posts: 4438 Location: W kraju nad Wisla. UTC+01:00
|
Posted: Mon Jun 29, 2015 7:32 am Post subject: Re: Feature request: pointerscan agreement-based heuristics |
|
|
Feature #1
I saw, many times while preparing do to first rescan, that 80% (or even more) of pointerscan results were pointing to the same address, THE WRONG ADDRESS. (e.g. 0000001C, or 00000000)
The most "popular" address is not always the right one.
Feature #2
* At least X% of pointers are valid.
* At least Y% pointing to same address.
And what X and Y will be? For different game, or even newer game version, those values can change drastically.
_________________
|
|
Back to top |
|
|
shadowsinrain How do I cheat? Reputation: 0
Joined: 28 Jun 2015 Posts: 2
|
Posted: Mon Jun 29, 2015 8:59 am Post subject: Re: Feature request: pointerscan agreement-based heuristics |
|
|
Well, I should agree, 80%-rule might kick in only after first rescan. Usually I am doing 3-5 rescans on simple games, I think, and reducing this number to 1 will be great neverthless.
Need research to say how often this happens. I guess, with some data on hands, it will be also possible to craft some more heuristics for filtering very first set of pointers in "average" game.
Perhaps I can contribute by collecting pointerscan samples and writing analysis scripts: I have 200+ games in my library and some Python knowlege. But before I started, can please someone give me quick direction to where I can find pointerscan file format details? I know there is repository, but more ready documentation would be nice.
mgr.inz.Player wrote: |
pointing to the same address, THE WRONG ADDRESS. (e.g. 0000001C, or 00000000)
|
I think all pointer paths leading to very first kilobytes of game memory may be invalidated. Same about paths with address 0x0 at any position. Of course these heuristics should be optional.
|
|
Back to top |
|
|
Gniarf Grandmaster Cheater Supreme Reputation: 43
Joined: 12 Mar 2012 Posts: 1285
|
Posted: Mon Jun 29, 2015 2:11 pm Post subject: |
|
|
I wanted to point out that finding the most frequent address implies reading all pointers, like when you do a "only filter out invalid pointers" rescan. That means feature#1 is roughly doubling the total rescan time: one "rescan" to make stats on the results, then the rescan dialog can pop up showing the most frequent address, then you can do the rescan that will actually remove results.
So it's definitely a no-go for any case with a large amount of results.
shadowsinrain wrote: | Then for most situations (I believe so) it will become no-brainer to spam "Rescan memory" + enter | I don't know the specifics of what you're hacking, but you might want to consider doing rescans in "value to find" mode with the "repeat rescan until stopped" option, that'd spare you having to manually type the address and would do the no-brainer part for you.
_________________
DO NOT PM me if you want help on making/fixing/using a hack. |
|
Back to top |
|
|
|
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum You cannot attach files in this forum You can download files in this forum
|
|