Revolver bullets tend to creep forward due to recoil. Even with a big roll crimp, a bullet may still go forward a few thousandths until the back edge of a canelure or crimp groove gets pushed firmly against the crimp. And, if the recoil is really hard, it can go a little farther than that without actually "jumping crimp" and starting to walk farther out the cylinder with each shot. So, it is best to leave a little space between the bullet's nose and the end of the cylinder. Some folks recommend at least 0.03", while others push it a little farther out to accomodate some favorite bullet. If you are pushing it, you should check for bullet movement after each shot until you get the mouth tension and crimp right to hold the bullet frimly enough. Otherwise, you may get a bullet so far out that you can't rotate the cylinder, and it will jam-up your gun.
If you need to get the bullet a little farther into the cylinder, you can always trim a few cases to less than the "minimum" so that they give the necessary overall length when crimped onto the bullet in an available groove. But, remember that this leaves less space for the powder than would be there if the bullet was crimped into the same goove in a case of normal length. So, if the DATA that you are using calls for a normal-length case crimped into that same groove, then you would need to reduce the maximum charge. (There are some Speer bullets that have had data published with several different case lengths, for instance). But, if you are actually still seating the bullet farther out than was used to create your data, say by crimping in a different groove in a slightly shortened case, then the published maximum would be safe and could even be increased a little.
There are calculations that can be made to adjust maximum data for different seating depths. If you need to do that, ask on this forum and some of us would be glad to help with the calculation if you aren't sure how to do it.
SL1