[WISH] differences for big files

[Dark-Star]

(I hope it's okay to bring that one up again... I found a similar thread on the forum but it was quite some time ago...)

I would really like a smarter "compare by content" function, which is able to show the real differences for big files. The problem is that I often have to handle those large files while developing, and from time to time I need to know, for example, which bits/bytes changed from one file to another. A simple "the files differ" is not enough in that case.

would it be possible to implement something like that? It would perfectly okay if it would only work in binary mode but not in text mode.
Then you could, for example, read a 64k block around the differing file position and display that. I imagine that for text mode comparisons, you'd have to start parsing backwards/forwards for linefeeds, which probably isn't very straightforward to do

[sqa_wizard]

for example, which bits/bytes changed from one file to another

That is exactly what the internal compare tool currently does !

A simple "the files differ" is not enough in that case.

It looks like you are using the new "External Compare Tool" feature ...

To override this mode temporarily, you have to use the command "cm_IntCompareFilesByContent" (as button or shortcut) which shows the binary differences.

[Hacker]

sqa_wizard,

That is exactly what the internal compare tool currently does !

Not if the files are too large.

Roman

[petermad]

2Dark-Star

Then you could, for example, read a 64k block around the differing file position and display that

But what if there are hundreds or thousands of different file positions scattered all over the files?

[Dark-Star]

sqa_wizard,

That is exactly what the internal compare tool currently does !

Not if the files are too large.

Exactly. It works fine for small files, but not for large ones. I haven't (yet) found out what the exact size is where it stops working, though.

But what if there are hundreds or thousands of different file positions scattered all over the files?

Note that this already works for small files (for example, try comparing two entirely different 1meg files). So having "thousands of differences" in two files is definitely not a problem.