[sword-devel] Version stamp format
refdoc at gmx.net
Mon Oct 22 11:50:07 MST 2007
Karl Kleinpaste wrote:
> I understand (now) that spec restricts the version stamp that way, but I
> have yet to see any reasoning *why*. Is there code which performs
> arithmetic on them? In GS, it's treated just as a string, like any
> other key=value string in a *.conf; if our module manager finds
> differing values, without any consideration of a "greater than"
> semantic, we put up a refresh icon, as visual advice to the user.
That is curious. So if I replace a module 1.1 with a "new" module 1.0, I
would actually get a message that my installed module can be updated?
This appears to make the current numbering system a bit irrelevant if
simply _any_ change would result in an update message.
> As I think about this, imposing a "string greater than" semantic in GS
> would be easy, using strcmp() for more than just a != test. And I
> already see where that would benefit users (i.e. me) where there are
> modules made available in more than 1 repo, or where an updated module
> has been withdrawn for some reason.
This would make sense - it would also allow targetted downgrades -f
particular new features in experimental modules do not work of if I
simply do not like its features.
> I had never previously thought to check for any limitation on version
> stamp format. I faintly remember noticing around the time that I first
> created the GS manual module that bible.org's module version stamps were
> 2-point, and of course Sword itself and all the UIs have version stamps
> which are 2-point. The version name restriction seems arbitrary unless
> there is arithmetic being performed. (And if there is indeed a problem
> with this, then someone really does need to have a chat with bible.org.)
> sword-devel mailing list: sword-devel at crosswire.org
> Instructions to unsubscribe/change your settings at above page
More information about the sword-devel