Re: ABI breakage

On Mar 22, 2005, at 4:33 PM, Philip Martin wrote:

> - I have a nasty feeling that it qualifies as an ABI change, so the
> rules require a new svn_wc_entry2_t. Yuck! I really don't like the
> size of the changes we would need to accomodate that.

Okay, ouch. I understand. I hadn't thought through this, and didn't
realize that when previously adding fields to the end of structures,
they were always private structures.

> Perhaps the lock data could be stored in a separate cache within the
> access baton, then no change to svn_wc_entry_t would be required.
> A separate cache could avoid using any significant memory if there are
> no locks.
>

Hm, this seems a whole lot less painful then revving svn_wc_entry_t.

Do I understand you correctly?

You're saying that .svn/entries gets parsed into potentially *two*
different structures: an svn_wc_entry_t, and an svn_lock_t. Both of
those structure types end up being cached in the adm_access_t.

Um... any heroic volunteers willing to take this on? Preferably
someone comfy with adm_access_t stuff?

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Wed Mar 23 01:09:56 2005