Macworld Forums: How I name files on my Mac - Macworld Forums

Post your comments for How I name files on my Mac here

For the truly anal-retentive, Adobe Bridge has all sorts of file-naming and other tagging functions that are presumably a godsend.

I also date my files, but at the end of the name, because what's in a file is more important than when it was created. Just saying.

I agree with this article.

I've had observers tell me that including the date is redundant, as date information is embedded in the file with regards to creation, modification, and the like. But those things can change, and using the date in the title working from largest unit to smallest (year, month, THEN date) makes sure the data is retained, and sorting assured.

Especially in a professional setting, taking the time to create naming conventions that you adhere to--including abbreviations that keep file names from getting TOO long--are as important to preventing loss as is backing up the data itself.

aestival, on 07 June 2012 - 06:08 AM, said:

I agree that contents are more important than date, especially when relative date is stored within the file headers itself.

BUT, placing the date anywhere other than the beginning of the tile name makes standard, alphabetical sorting based on that criteria impossible. While you could continue to sort using Finder's own date sorting tools, as the author mentions, the file creation or modification information in the headers does not correspond to the actual date relevant to the contained information. For this reason, using standard and uniform numerical date in the beginning of the file name is a superior method.

An aside for *some* of us—web designers—on something you mentioned: the importance of all lowercase. I made the stoopid mistake a few times, before I learned. Some hosts are case-sensitive. Thus, if I place an image on my host by the name Greatpic.jpg and link to it as img src="greatpic.jpg" (notice the G) it won't show up. Best—as you said—all lower case!

This is a great suggestion. Personally, I've come up with my own naming "system". I usually use this structure: ProjectName - Description (yyyy.mm.dd - version).ext

For example, a department store charge card statement in PDF would be "Sears - 2012.05 (2012.06.05).pdf" This way, I can easily search with keyword "Sears", and get a list of all my statements indexed by year and month of the statement, and also see the date that I actually received it.

For a project, I'd name the file something like "NOB - WBS and Programme (2011.05.30 - v03).xlsx". Once again, I can easly search for files with keywords "NOB" and "WBS", and I'd get in return a list with such keywords, and also be able to not only see real quick what the file is about, but also see the date in which I created it, and the version of the file to not only see the latest revision, but also be able to review a previous one.

Cumbersome? Maybe. But so far it's proven quite effective.

what about using tags! filenames and trees and directories are "onedirectional". What if I'm looking for water bills and water consumption

Where I work, I work at an edu institution where I have processed thousands of graphics (~80,000) for all the various programs taught there. I constantly repurpose content and update it so consistent file naming is essential in finding existing graphics or in producing new ones. The file naming scheme I adopted is a hierarchical one. The biggest main division is considered first in sorting the graphic files, (which in my case, is Health sciences and Industrial technical). Then the particular program is identified in the file name using an acryonm of 2 or 3 letters; then the actual course name is identified, followed by the learning outcome, the figure number and a one or two word descriptor. Each file is different from each other in at least two characters so the possibility of deleting one accidentially due to it being misnamed is almost nil. I restrict the file name size to 27 characters or 31 when the extension is added. Metadata key words are also applied to each file.

120607 has always worked for me...i have been scanning photos for years and this is how it all started...back then, ( ) SPACE wasn't that cool either so it (still) looks like this:

120607_MrX_Field

or whatever (date_person_location)

when 10.7 came out ALL MY FILES seems stupid (1 out of 9,996 selected) but as one got accustomed to it..its a really nice and fast way to find files for lets say, the last three days or so....
(alas, i share 2 dropbox folders and whatever is done in those shows up as well, so its a little cluttered, but perfection is yet to be found...)

Folders would be the other half of file management. They take some time to set up and require a few steps when saving and so on, but make it easier to find files if used well.

Edit: Duplicate posting.

This post has been edited by choiceweb: 07 June 2012 - 06:59 AM

I use a similar approach with data files from work. I can quickly check my lab book and find the file I need. The other piece of advice is to have good headers within the files. You can be surprised about how important minor details are. I was looking at some older data I'd saved from my post-doc. It was pretty much just X-Y data and that lab book is lonnnng gone.

This approach isn't needed for my personal files. I don't really care when I saved a certain recipe, just need its name. Likewise, I have an Excel for handling all of our finances, another for freelance income, another to keep track of child care expenses. I won't have duplicates of these files, so a simple descriptive name is adequate.

BB

This article makes me understand why Steve Jobs wanted to do away with the (visible) file system. Such file naming conventions are antiquated. Metadata (what another poster referred to as tags) allows a nearly infinite number of ways to organize and locate information. Utilities like Default Folder X (www.stclairsoft.com) let you enter metadata/tags right in a Save dialog box rather than having to go to the Finder and use Get Info. That way, you can still have a filename that is readable in icon and column view (a shortcoming of Mr. Sparks' system) but is Spotlight-friendly.

I used to be big on file organization; I still do more than most in both the Finder and my ever-growing e-mail archive. But I am finding more and more that what I name something or where I file something is only important if Spotlight can't read its contents. I can see why a lawyer like Sparks might go to such lengths in a profession where documentation is everything, but I do believe anyone building a system right now would be better off using metadata than creating long filenames.

I simply have too many similar documents to do anything other than hierarchical folders.  For my teaching materials, for example, I have a folder for each school year, sub-folders for each course, then folders for each month, then folders for each day.  Each folder has similarly named items (Bellwork, Lesson Plan, etc).  If I included the date in the filename, I'd have to change every file name whenever a lesson was pushed back.  As it is now, I just change the folder name.

This also means that when I'm looking at something, I'm specifically looking at one class or date, instead of seeing a list of every "Lesson Plan" and "Homework" from 9 years of teaching.  I do the same thing with materials from every other organization I deal with.

The only way to avoid this would be some sort of database program which could assign different files to dates on a calendar.  Absent that, I'll keep my current method of organization, thank you very much.