Contents
A filename has a small but demanding job: identify something when the surrounding context has gone away. I would not expect it to describe the entire history of a project. I would expect it to help distinguish the draft I need from the three similar files beside it.
This example uses an invented community newsletter project. There are meeting notes, draft text and approved images. No real personal documents are involved. I will choose a naming pattern for this collection, explain its limits and leave other collections free to use a different pattern.
Imagine the search from the other end
The search I care about is “Find the approved introduction for the autumn issue.” That tells me the filename needs to distinguish the project, the content and its state. A camera timestamp might matter to an image collection, but it would not answer this editorial question by itself.
I would write down two or three likely searches. If one pattern cannot answer all of them, I would not keep adding words indefinitely. A folder, an index or application metadata may carry information that does not belong in every filename. Names are one part of retrieval, not the whole system.
Choose a readable pattern for this collection
For the demonstration, I would use an issue identifier, a short description and a clear state. Hyphens make the parts easy to scan here, but that is a local convention. The important point is that every collaborator understands what the parts mean and can produce a matching name without consulting a long manual.
The Library of Congress recommends meaningful filenames and a simple structure as part of digital preservation. I would apply that general principle to this specific project. It does not prescribe my exact syntax, promise permanent compatibility or establish that a complicated naming system is better than a modest one.
Write a small example before renaming anything
Here is a sample directory for the fictional newsletter. “Approved” is used only after the project's actual approval step; it is not a synonym for “I like this version.” The README-style index would explain the state terms and identify the person responsible for changing them.
I would try the pattern on copies of a few ordinary files. That shows whether names remain legible in the tools the group uses. It also reveals exceptions such as a long title or several related images. I would resolve those exceptions before applying the pattern to an entire collection.
newsletter-autumn/ 00-index.txt autumn-introduction-draft-01.docx autumn-introduction-approved.docx autumn-cover-image-approved.jpg 2026-11-10-editorial-meeting-notes.txt Index note: The date on meeting notes is the meeting date. “Approved” follows the agreed editorial review.
Keep dates and versions from changing meaning
A date can mean when a document was created, when a meeting happened or when an image was taken. I would choose the useful meaning for this collection and state it in the index. Otherwise, a neat chronological sort can put unrelated kinds of dates into a misleading sequence.
I would also decide how working versions become an approved file. If the tool already supplies a version history, a filename containing every revision can become awkward. If files travel by email, a visible draft identifier may be valuable. The right choice depends on how the group actually exchanges material.
Check connections before renaming
A name change can affect links and references. Obsidian documents internal links and their supported forms; other applications have their own behavior. I would not assume a copied filename change updates every document, shared link or export that refers to it. A representative test matters more than a reassuring-looking folder.
Before a larger change, I would make a separate backup and record the old-to-new mapping when references need preserving. I would also consider people who have already received a file. A small collection can be renamed in a controlled pass; an active shared project may need a coordinated change instead.
Keep private details out of the name
Filenames often appear in screenshots, attachments and shared links. I would avoid putting unnecessary personal identifiers in them. The invented newsletter does not need a volunteer's phone number in a meeting-note filename. A useful identifier should help retrieval without revealing more than the task requires.
I would revisit the pattern when the project changes, not every time I see a prettier system. If the current names let the group find the right file and understand its state, the pattern is doing its job. A convention that remains understandable is more useful than one that is endlessly redesigned.
Walk through the newsletter folder as its next editor
I would begin by imagining someone joining the newsletter project after the original discussion. They need the approved introduction, the image selected for the cover and the notes explaining the choice. A folder full of filenames that made sense to the previous editor may not answer those questions. I would write a short index identifying the issue, the purpose of the folder and the meaning of draft and approved. That small record can carry repeated context without forcing it into every name.
Next I would choose one working file and trace how it becomes an approved artifact. In the invented example, the introduction begins as draft-01. A group review may produce another working version, but approval is a separate event. I would not rename the file approved simply because it is the most recent one. The naming convention should reflect a real process and identify the current authorized copy. If the group has not defined that process, a better filename cannot supply the missing agreement.
The approved file should remain distinguishable from later experiments. Suppose someone tries a shorter opening after approval. That can be a new draft with a clear relationship to the approved version, rather than a silent replacement. The project index can say which file is ready for the intended use. I would not ask collaborators to infer that from a modification time, which may change during copying or editing without representing a new editorial decision. A state word needs a shared meaning to be useful.
Finally, I would rehearse the handover. Can the new editor find the approved introduction without opening every draft? Can they identify the source of the cover image and the note about its permitted use? Can they tell whether a meeting date belongs to the discussion or the release? Those are concrete questions. If the convention cannot answer them, I would improve the index or names before expanding the folder. The goal is a reliable choice at the next step, not a directory that merely looks orderly in a screenshot.
Choose what belongs in a filename and what belongs elsewhere
A filename should contain the information needed to distinguish the file in its likely context. For the newsletter, an issue identifier, short content description and state may be enough. A detailed explanation of the approval belongs in the decision record. A creator credit and license may belong in an image register. I would keep those records connected, rather than attempt to encode the full provenance and every editorial comment into one increasingly long name.
I would test the file outside its parent folder. An attachment called introduction-approved may be understandable inside a clearly named issue directory but ambiguous in an email download folder. Adding the issue identifier can help. That does not imply every file needs the organization's complete name and a date repeated several times. I would choose the smallest identifying context that survives the way people actually exchange these files. The intended journey matters more than a universal preference for short or descriptive names.
For images, I would preserve a connection to the original and to any edited copies. A crop prepared for a cover can have a role-based name while the index retains the original identifier and source information. I would not overwrite the only original merely because the cropped copy has the cleaner name. The convention should help distinguish the transformation, not erase the history needed to understand what the project is using. That distinction is useful even in a small personal collection.
If an explanation is too nuanced for a filename, I would move it into prose. For example, “approved for this issue, with caption still awaiting confirmation” is not comfortably represented by approved alone. The file can carry a provisional state and the index can identify the remaining condition. Names should aid interpretation; they should not make a conditional decision appear final just because the filesystem offers no special field for the nuance. A simple text record is often enough to retain it.
Plan a rename without breaking the working project
Before a larger rename, I would make a separate backup and list the files actually involved. I would identify active links, scripts, shared references or documents that mention the current names. The aim is not to discover every possible dependency in the world, but to understand the ones this project uses. A folder that looks private may still supply a linked image to a shared document. Changing the name can affect that relationship even when the file's contents remain untouched.
I would then try the convention on a small copied sample, with one awkward case included. A long title, an unusual character or two similar attachments can reveal limitations that ordinary examples miss. I would open the sample in the tools that need to handle it, not just inspect the directory listing. The test should reproduce the intended use closely enough to answer a specific question about naming, while leaving the active collection available if the pattern needs revision.
The old-to-new mapping can be a short table. It becomes valuable when another person has already cited the original filename or when several versions have been distributed. I would keep the mapping with the project rather than rely on memory. A rename is a change to an identifier, and a brief record can preserve continuity. If the collection is large or actively shared, I would coordinate a limited change window instead of asking people to work around unexplained missing references.
After the rename, I would inspect representative paths: a link from the project index, an attachment in the working document and a file copied to the handover folder. If those paths work and the names remain clear, the change has served its purpose. I would not continue renaming unrelated historical material to make the entire storage system match. A convention can be introduced at a useful boundary, with older collections retaining their own documented logic until a real task justifies further work.
Treat an old archive differently from an active draft folder
An active newsletter folder supports current decisions. An inherited photo or correspondence archive may preserve an original order that carries meaning. I would not apply the active project's state words to that archive automatically. A name in an old collection can be evidence of how the material was arranged or received. Improving retrieval may require an index rather than replacing every identifier. I would inspect the collection's purpose before choosing the scale of change.
Dates need particular care. A scan date is not necessarily the date a photograph was taken, and a file's creation timestamp may reflect a copy operation rather than the original event. I would describe the relevant date in the record and keep uncertainty visible. A chronological filename is useful only if the date's meaning is understood. Otherwise the tidy sort order can create a false account of when things happened, which is harder to correct once the files have been shared.
Privacy also travels with names. A file containing ordinary project notes may not need a person's full name, contact information or sensitive circumstance in its title. Names appear in screenshots, attachment lists and shared URLs. I would choose identifiers that support the task while keeping unnecessary detail in an appropriately controlled record. A convention should not turn the directory listing into an accidental summary of private information simply because descriptive names are generally helpful.
My final check would be to ask whether a future reader can identify the item, understand its role and find the relevant context. If the answer is yes, the naming system is adequate for its current job. I would leave room for exceptions with a brief explanation instead of distorting every file to fit a rule. A useful convention stays understandable through ordinary work; it does not require continuous redesign to preserve an appearance of perfection.
Try the filename after copying it out of context
I would copy a representative demonstration file into a separate handover folder and inspect its name without relying on the original directory. The issue identifier, content description and state should still make sense. If they do not, I would add the smallest missing context rather than attach the project's whole history to every file. A short index can retain the deeper explanation.
I would also open the copied document and compare its stated status with the filename. An approved label should not sit above a document still visibly awaiting an ordinary review step. If the status is conditional, the record should identify the condition. This simple comparison keeps the convention connected to actual editorial work rather than allowing a renamed file to create an appearance of approval.
The convention passes when the next editor can select the intended artifact, find its relevant context and distinguish it from a working experiment. It need not predict every future project. I would preserve a small explanation and leave older collections alone until a real task justifies change. That makes the naming pattern a useful local tool whose benefits can be inspected, instead of a universal cleanup rule applied without regard to how the material is used.
Frequently Asked Questions
Q: Is a date-first filename always best?
A: No. It suits collections where chronological retrieval matters. A project or subject-first pattern may be clearer elsewhere. Decide what the date represents before making it the leading field.
Q: Should I remove spaces from every filename?
A: Use names compatible with the systems that need to handle them. Hyphens are one readable option, but this article does not establish a universal technical requirement to remove spaces.
Q: Can I use “final” in a filename?
A: Only if your project has a clear meaning for it. If later edits are likely, an approval state or dated release may be easier to interpret than a chain of “final-final” names.
Q: Should I rename old archives to match?
A: Not automatically. Existing references, original order and provenance may matter. Start with a small copied sample, preserve a backup and change older material only for a clear retrieval benefit.
Q: What if filenames become too long?
A: Move repeated context into a folder or index, and keep the distinctive information in the filename. Check the requirements of your actual storage and sharing tools rather than assuming unlimited name length.
One thing to take forward
I would name the next file for one likely future search, then check that the result makes sense outside the current window. A short, explained convention can carry a project a long way without trying to encode its entire history in every name.
Sources & further reading
- Library of Congress: Preserve Your Digital Files, Papers, Photographs and Books ↗
- Obsidian: Internal links ↗
Something needs correcting? Write to the editors with the article link.



