This commit is contained in:
parent
3d25893744
commit
24674f7767
1 changed files with 18 additions and 0 deletions
|
@ -0,0 +1,18 @@
|
||||||
|
Hey,
|
||||||
|
|
||||||
|
I've found that git annex works great as a way to publish websites to a web server. I can edit my website on my computer, `git annex sync` my working directory to the VPS, and then `git annex get files-I-want-to-publish`. This works great. I can maintain my normal working directory structure on the VPS and I don't have to worry about people seeing files I DIDN'T want to publish, since the dead symlinks just show up as 404s.
|
||||||
|
|
||||||
|
There's one small problem.
|
||||||
|
|
||||||
|
Say:
|
||||||
|
|
||||||
|
1) I've already published a file using `git annex get file-to-update`
|
||||||
|
2) I update that file on my computer
|
||||||
|
3) I do `git annex sync`
|
||||||
|
4) I do `git annex get file-to-update`
|
||||||
|
|
||||||
|
Between steps 3 and 4, file-to-update goes from being an accessible web resource to being a dead symlink. It's not really a problem for me, as hardly anyone visits my site. But it would be nice if I could make `sync` leave the old symlink to the old file until I `get`ed the new one.
|
||||||
|
|
||||||
|
Is this possible?
|
||||||
|
|
||||||
|
PS: For those who might follow in my footsteps, remember that you probably don't want people reading the contents of your .git dir, so make a re-write rule for this!
|
Loading…
Add table
Add a link
Reference in a new issue