-
So, pretend you got everything down.
-
JD; Areas, Categories, ID.
- Everything running smooth.
- Hummin’.
-
Zoom down to individual documents/drawings/files.
-
What is best practice for versioning?
- By best practice, I’m targeting;
- least amount of places you have to input/update (filename, inside file)
- possible to sort files in lexicograpical order
- implicit information in shown versioning
- By best practice, I’m targeting;
-
What I always do
- I always make a copy of the file.
- I always prepend date to the beginning of the filename, I use YYYYMMDD- I know it’s not ISO 8601 - but for me it’s faster.
-
What I normally do
- Bound by tradition, versions of a document/file/drawing, always starts as 0 with the initial version, and then increments to A, B, C and so on… (I don’t know why - I’ve inherited it from work)…
- I normally (would) put version in filename somewhere.
- Often directly after date, as in YYYYMMDD-A-rest-of-filename.
- This is however a big hassle when working with Sharepoint-based systems (Sharepoint should be killed with fire). Sharepoint only recognizes exact same filename if you are trying to update a file and upload a new version. So for Sharepoint, I need sub-optimal workarounds that keep the same filename as original.
-
What I see some smart people doing
-
They use internal revisions:
- A1, A2, A3, A4 are internal
- Before released/submitted/sent to the customer - they change to new main revision letter, in this case B.
- Then B1, B2, B3 would be internal again.
- C is the next release to customer.
- The internal revisions also works great in Excel-lists
- First column in spreadsheets contains “Version”
- All rows that are edited/changed gets a “B3” or “B4” or whatever.
- You can then filter to see what has changed since last time you checked the spreadsheet
- If the last version you reviewed was “B1”, you just filter “B2”, “B3” and “B4” to get updated on all changes.
- This does however require dicipline from everyone who are editing the spreadsheet to add this to all edited rows.
- First column in spreadsheets contains “Version”
-
Real world document handling
- Often, a document or drawing has an external document number/name and version (say the company or vendor that created the product).
- In addition to this, you might want to have an internal document number/name and version/approval record.
- This is often where things starts getting tricky, both on a people/dicipline and on a systems/workflow level.
-
Anyone have any thoughts/experience/best practices - that are relatable?
-
1 Like
For me, best practice is today, that I do not need to save documents in different versions, because OneDrive/SharePoint does it for me.
When I had a ‘proper job’ and distributed stuff to customers/clients, I tended to keep it simple. And actually I use the same system now.
The working copy of a document is just the document. No version number in the name. I know some people like the version number in the name so if that’s you, go for it.
Keep version numbers simple. You can’t lose with the date as yyyy-mm-dd. Because ‘v1’ ‘v2’ or whatever is meaningless. And just one more schema to try to keep track of: is this v1.1 or 2.0 or what and why?
When a file goes out, a copy of, say, the Word doc is dropped in /distributed along with the PDF that was actually distributed. At this point both of these versions contain the date in their name. Ideally at the start, but wherever so long as the entire folder sorts by date.
If the file names in this folder aren’t consistent, you’re fired.
This was all I ever needed. Keep it simple.
1 Like
-
The type of documents/drawings I was thinking about needs very down-in-the-details tracking of information changes.
-
Typically, changes are issued with revision bubbles/revision clouds to show those who review it what has changed since last revision:
Example:
Often, each page/sheet has it’s own revision history, as in
Example:
This information is shown in the title block on each page/sheet:
Example:
For lists, it could be a small change in a signal text on a signal in list of hundreds of signals, where it is important to keep track of revisions so that texts can be imported/exported correctly between systems and other documentation:
Example:
It’s possible to go for large, complicated, (expensive) document management systems. But I would like to avoid that, and instead have good rules/syntax on files and folders - and apply a JD like system to keep track. 
- I think it would/could work in a small, motivated team.
- I am more sceptical about success if doing it on a larger scale/more people.
Ah wow! I see. This looks fun. I’ll watch from the sidelines. 