You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@subversion.apache.org by Karl Fogel <kf...@red-bean.com> on 2022/02/03 06:34:21 UTC

Re: pristinefulness and svn:needs-lock

On 31 Jan 2022, Daniel Shahaf wrote:
>The main use-case for pristineless files is large, undiffable,
>undeltifiable files checked out from a box in the next rack unit 
>over
>(= wide/cheap uplink and downlink).
>
>Undiffable files will likely have svn:needs-lock on them, won't 
>they?

FWIW, I don't know that network-nearness of the repository is an 
important element of the use cases -- it's certainly not part of 
ours anyway.

I also don't know that undiffable files generally tend to have 
'svn:needs-lock' anyway.  In our own use case, at least, the files 
don't have that property.

I don't really know how people use the locking feature in 
practice, but I guess I always assumed it was used equally on 
mergeable and non-mergeable files, because it's more about how 
teams communicate than it is about the files themselves. 

>Is there a correlation between "has svn:needs-lock" and "doesn't 
>need
>its pristine"?  Does either of them imply the other (at least in 
>most
>cases)?
>
>Should svn:needs-lock files that aren't locked have pristines (by
>default)?  This could be viewed as a backwards-incompatible 
>change, so
>assume the user has opted in to new behaviour.

My feeling is: svn:needs-lock can be treated as independent of 
pristinefulness for now.  There's no requirement that the two 
interact -- there's no bug that happens if we keep them 
independent.  Maybe in the future we'll see some reason to make a 
connection, but I don't see why we would make that connection 
right now.

Best regards,
-Karl