Careful with that, dropbox has been known to corrupt git repo’s like that! I’d recommend using KeyBase’s git storage. It also has distributed file storage like dropbox if you need it, but it has a specialized git area that keeps consistency and safety unlike the distributed file storage style that both keybase and dropbox use.
Ah you just keep files, well still be careful about offline synching potentially changing files versions in unexpected ways too, use git instead (github, gitlab, whatever)! ^.^
+1 I have so many slightly edited versions of so many files from programming 20-30 years ago… >.>
And I even used the CVS era (I still have my old SVN server running…).
Precisely, that is why you say a file at commit <hash> (or even just joe/master or so), or just a link to it in your git server now.
But yeah, this is inherently disconnected so will get out of sync, everything should stay in the VCS.
Hmm…
Many editors now can show precisely when a line was added based on git, the commit, the message of the commit, get details about the branch, commit, server, etc… I can literally just slide a scrollbar to go backward and forward in history.
Sure you can find when a line was added - but I want to know where the line came from
Suppose I browse to some web page - select some text and put it in the “cut” buffer
then I paste this text into a file that I’m editing - all the editor knows is that the
text came from the cut-and-paste buffer and NOT the web page that I cut-and-paste from.
After the event when I read (or somebody else reads) the text they might be left wondering “where the heck did this data come from” - I’d like to drill down and find out exactly where everything came from.
Editors just record keystrokes - but they don’t know or care or record where the data came from - so one of the most important questions “how the heck did the author of this text get the necessary information to write this text” is not answered.
Doesn’t the history show that? It shows branches and merges and comments and all?
Oh you mean something like that! I think that kind of thing would require interface changes through-out everything and as such is untenable at this time.
I personally usually put a comment with the webpage or so of where something came from if I copy something. ^.^
After all, it’s not like the program can read your brain and know that you aren’t typing it from your phone screen or something.
Yes I do mean “something like that” - It would be perfectly possible to record meta-data associated with each cut-and-paste information and save all of this,
Many word processors have a track-changes operation, but usually you cannot unroll all previous edits - only the latest - and if you pasted data from some other source you cannot dig down into the other source.
The ability to dig down into text to see where it came from is one of the ideas in Xanadu by Ted Nelson. Ted coined the word “hypertext” and had some great ideas about what was wrong with files.
Some of the ideas of tracking data sources can be found in
My point was that it doesn’t have to be that way. I’ve worked on projects where it is not done that way, everything is just checked into a single folder. Its pure insanity.
That is indeed pure insanity. FWIW, I’ve seen lots of projects that way on Windows–it seemed to be common practice. And historically there was a good reason for it. For a very long time, include paths in the predecessors of Visual Studio were comma-delimited in a single text field, they were not recursive, and the text field was limited to 255 characters. (Yes, that was INCREDIBLY stupid, but it was what it was.)
