@BenLoosli and the others who mentioned random and meaningless filenames, I have a rant for you 🙂
Purely random and forced sequential filenames are stupid. Modern Databases should be able to auto create filenames and propagate attributes based off basic programming (even excel can do this!). E.g. if I want certain parts to be in certain part ranges, why is that so hard for modern software? The inability to do this seems terribly outdated and bad/lazy programming. With very low effort I could get excel to do this, and I am can barely do VBA. If I want a filename to start with a year code, project number, model number, etc, I should be able to do that. If I can even program the naming logic in excel, then a real database should be able to do this easily without the every growing scary "C" word (Customization...ahhhhh! 
).
To be able to look at a P/N and immediately understand useful information about it is powerful, and should be easy to auto number based off of basic criteria. If your database cannot do this without doing that scary "C", stuff then your database is garbage.
End Rant. 🙂
It seems logical to keep filenames short and meaningless.
And I assume this is also how PTC intends it, as they provide means for an auto generated number for the file name, and also expect the file name to be unique:
What I do not understand however, is that they then use the (meaningless) file name as the primary identifier in all other dialog boxes and info pages, e.g.:
File - open dialog:
How is one supposed to know that 0381212444.asm is the actual assembly to be opened? It's a meaningless name.
In this case, one can argue that the correct procedure is to search in Windchill instead, and select "Open in Creo".
But the same goes e.g. for the "assemble" dialog box, which is identical to the "File - open" dialog. I don't know an alternative for this straight from Windchill...
Also e.g. in Windchill:
The Common Name is shown under "CAD Document Attributes", but all the other fields refer to the file names, e.g. also in the structure view. I know these views can be customized, but it really wonders me why on the one hand it seems to be intended that file names are meaningless, auto generated numbers, and on the other hand in the standard Windchill layouts, the file names are used as the main identifiers...
Also the model tree shows these "meaningless" file names. I know the Common Name can be added as an extra column, but this takes up valuable screen space, and is IMHO less readable (e.g. not in the leftmost column, not indented, etc):
In a company I worked for previously, this was 'solved' by giving a readable file name to each 3D component, instead of a "meaningless" number.
They had a customized 'file - new' dialog, which checked for file name uniqueness, and added some random characters to the end of the file name to ensure this uniqueness (hereby limiting the nr of characters to even less than 31...).
I think this is the way we will also go, albeit without the customisation. Only real drawback I see is that, without this customisation, you only know if the chosen filename effectively is unique upon check in. Which probably also will be annoying but I don't see a better solution currently...
Regards,
Johan Rutgeerts
At minimum for Legacy purposes, and for things like emails stored in WC, the length should not be limited (WC should just match Windows and other supported OSs), because when you are importing something it increases the likeliness of a poorly thought out migration where users are forced to truncate the file names and essentially lose information that was stored in the file name for many years prior to migrating to WC.
@bsingleton and all others reading, do yourselves and your customers a favor and do the following:
- Migrate the long filename as another field (name, description, or better yet create a non-required field/attribute such as Legacy_filename)
- Get your company to decide on a standard naming scheme for the filename/number meaningful and not random (smart sequential is better), and then use that going forward.
- Get your company to use your company created attributes, or create some, to capture essential information that you would want to search by later (e.g. product name, category, etc).
- Update your standard.
This way you might actually be able to narrow your search when trying to find the right document later. This is advise that I wish we received and followed during a past migration because it would have put us in a much better place now...rather than doing a search and getting a hundred files that no longer have a descriptive name and you have to open them all to find the right one...every time you do a search!
Little pain upfront, but a lot less pain and much more benefit later on (read: increased productivity & reduced user frustration).
Hopefully for you this will not be great advise too late (like Mom yelling "careful" after she hears you get hurt...or the kaboom! 
)
