On 08/05/2015 Amenel VOGLOZIN wrote:
Like Stephan Bergmann, I also think that the current test for newer version
cannot remain as it is.
I have adopted the version scheme based on the date as Andrea recommended. But
I am quite uncomfortable with:
1- effectively asking the entire extension developer community to adopt a
version scheme just so they can work;
2- having an "overlook"/mistake/comparison bug remaining in some code that does
not implement the intended action.
I'm not sure we have understood messages the same way.
I'm definitely not asking anyone to adopt a YYYY.MM.DD numbering scheme,
I'm just saying that I do so and that it works for me.
I think we all agree that any numbering scheme where alphabetical order
is equal to numbering order (e.g., version 1.000, version 1.001, ...,
version 1.009, version 1.010...) would work well with no surprises.
I though we had a bug with situations such as 0.9 -> 0.10; Stephan wrote
that the bug is either not there or not as bad as I thought and I trust
him more than my memories from years ago when I was investigating the
issue for one extension only.
I'll post other comments to
https://bz.apache.org/ooo/show_bug.cgi?id=126298 when I have time for
further checks, but I wanted to make the above clear before doing that.
Regards,
Andrea.
---------------------------------------------------------------------
To unsubscribe, e-mail: api-unsubscr...@openoffice.apache.org
For additional commands, e-mail: api-h...@openoffice.apache.org
