Binary comparision not correct
Binary comparision not correct
I got two binary files. ExamDiff Pro can not compare them correctly.
I tried to compare these file with UltraCompare and it can work correctly.
There are two differences:
1. The content from offset 2000H - 2FFFH is different.
2. Binary1.bin is larger.
I tried to compare these file with UltraCompare and it can work correctly.
There are two differences:
1. The content from offset 2000H - 2FFFH is different.
2. Binary1.bin is larger.
- Attachments
-
- BinaryFiles.rar
- (1.62 KiB) Downloaded 1349 times
Re: Binary comparision not correct
ExamDiff.png is the result of ExamDiff Pro and UltraCompare.png is the result of UltraCompare.
If you open the files with hex editor and jump to the offsets, you will find out the data in those offsets are the same.
If you open the files with hex editor and jump to the offsets, you will find out the data in those offsets are the same.
- Attachments
-
- UltraCompare.PNG (78.78 KiB) Viewed 25876 times
-
- ExamDiff.PNG (86.07 KiB) Viewed 25876 times
Re: Binary comparision not correct
Perhaps you are not seeing but in both places diffs are correct. Both FF/01 (in the first shown diff) and 01/19 (in the second) are colored as Changed.
psguru
PrestoSoft
PrestoSoft
Re: Binary comparision not correct
I think I can understand what your means is.
The binary comparision algorithm of ExamDiff Pro is not wrong but not clear enough.
The different content of these files are from 2000H to 2FFFH and it is not easy to get the result from ExamDiff Pro.
I think ExamDiff Pro is the best of the comparision tools in the world, but UltraCompare shows the result clearer to users than ExamDiff Pro in this case.
The binary comparision is a very important function to some users, perhaps you can consider to make it better.
The binary comparision algorithm of ExamDiff Pro is not wrong but not clear enough.
The different content of these files are from 2000H to 2FFFH and it is not easy to get the result from ExamDiff Pro.
I think ExamDiff Pro is the best of the comparision tools in the world, but UltraCompare shows the result clearer to users than ExamDiff Pro in this case.
The binary comparision is a very important function to some users, perhaps you can consider to make it better.
Re: Binary comparision not correct
So in this case, how would you show the differences?
psguru
PrestoSoft
PrestoSoft
Re: Binary comparision not correct
I think the binary (fast) mode of UltraCompare shows a better way in this case. You can download it to try from http://www.ultraedit.com/downloads/ultr ... nload.html
I think this feature is very useful to compare firmware files. The firmware files are usually the same size, so the binary (fast) mode is easier to understand to users.
Maybe, you can provide two ways to show the binary comparison in the future, one is the original way and another is the binary (fast) mode to look like UltraCompare.
It is also a good idea to show offset in left side of the window. It can let users read the result easily.
I think this feature is very useful to compare firmware files. The firmware files are usually the same size, so the binary (fast) mode is easier to understand to users.
Maybe, you can provide two ways to show the binary comparison in the future, one is the original way and another is the binary (fast) mode to look like UltraCompare.
It is also a good idea to show offset in left side of the window. It can let users read the result easily.
- Attachments
-
- UltraCompare2.PNG (82.92 KiB) Viewed 25851 times
Re: Binary comparision not correct
So it looks like their Fast Mode is just a byte-by-byte comparison. We don't have this; perhaps it's a good idea to have something like this.
See http://www.prestosoft.com/edp_faq.asp#17It is also a good idea to show offset in left side of the window. It can let users read the result easily.
psguru
PrestoSoft
PrestoSoft
Re: Binary comparision not correct
I second the request.So it looks like their Fast Mode is just a byte-by-byte comparison. We don't have this; perhaps it's a good idea to have something like this.
Sometime that method is simpler to understand.
Of course it should be available as an option in addition to the present mode.
Re: Binary comparision not correct
I'd like to refer also to this thread:
viewtopic.php?t=421
which more or less mentions the same problem.
In case of text comparing we already have the needed functionality (discussed here: viewtopic.php?p=1492). But the same mechanism (using similarity 0) could also be applied to the binary comparison, in which case the described problem would be solved.
And more, provided such case is enabled by user, we can also display hex offsets in the panels (because there will be no comparison gaps already). I tend to agree that such display would add readability.
viewtopic.php?t=421
which more or less mentions the same problem.
In case of text comparing we already have the needed functionality (discussed here: viewtopic.php?p=1492). But the same mechanism (using similarity 0) could also be applied to the binary comparison, in which case the described problem would be solved.
And more, provided such case is enabled by user, we can also display hex offsets in the panels (because there will be no comparison gaps already). I tend to agree that such display would add readability.
Re: Binary comparision not correct
The following suggestion has been added to ExamDiff Pro wish list, with all of your votes:
Implement fast binary comparison mode (byte-by-byte comparison)
Implement fast binary comparison mode (byte-by-byte comparison)
psguru
PrestoSoft
PrestoSoft