They are replaced and erased, no copies of old versions is kept. If you
ever need to upgrade a patch you should save a copy of the patch that you
applied. You will need it to upgrade your patch. In any case you can
always start from clean source applying the new patch.
You need the patch that you applied first, all you have to do is to
reverse the old patch and apply the new one. This means that you should
execute the following commands:
What is reversing a patch?
There may be lots of differences. For one thing, a patch for one version may
not be suitable for another version of Pine. This is because some changes in
the main source code of Pine make some hunks fail. However, the most important
difference is that some bugs in earlier versions may only be fixed in
later versions, so try to use the latest version of the patch if you can.
Another difference is that I keep adding features to some of the patches,
so eventually a newer version of Pine will have much more features than
old versions of the same patch.
If you see an image or text that says next to a patch title in
the main page, it means that the patch was updated within the last 15
days, otherwise the last line of the page that contains the description of
the patch says the last time the patch was updated. The update date there is
not the update of the page, but the one of the patch (this is always accurate,
since it is generated by a script that saves patches).
As an alternative now I keep a XML/RSS feed, which I update when an update has been made to a patch in this web page. This is a way of announcing updates for patches in addition to the ones indicated above. The feeds can be found at pine.xml.
You can not know this for all patches, but I try to keep a log of updates, at least for the most popular patches. If I keep a log of the changes of a patch, you can find it in the updates section. In case I do not keep a log of the change, it may mean that a bug was fixed or a feature was added, but unless you ask, there's no way to know it.