Hi Andrej. On 26.04.2018 18:52, Andrej Shadura wrote: > This doesn’t seem a wise move to me, since this is a significantly older > version of the code which is missing some features, bug fixes and tests, > and doesn’t bring a lot of fixes. I’m not sure what value this switch > had, to be honest.
The main reason for this is that the current version is *maintained* by upstream. He accepts bug fixes and tries to keep the software running. When I compare the popconn of tclxml (470) with the one of saods9 (433), which is the only reverse depency of tclxml, I think that saods9 is the major reason to keep tclxml in Debian. William Joye, the current upstream maintainer, also maintaines saods9, which makes me hope that he will continue to maintain tclxml for a while (saods9 is one of the most important programs in astrophysics, which is not going to die soon). If some functionality is missing, probably the best way would be to send a patch to upstream <https://github.com/wjoye/tclxml/pulls>. I would be quite sure that he accepts it. Otherwise, we could even include it as a patch in Debian. The tests however were always unused (also in the old package), so I don't see a big loss there. At the end, it is a question what weights more: a responsible upstream actually doing maintenance, or a few features that were rarely used. In my eyes its is the first. Feel however free to write bug reports ;-) Best regards Ole
