Hi I'm a lazy module maintainer and I would like to make d5 ports of my d6 modules. Would there be some way to convert deadwood into something that would downgrade modules, and what obstacles would I face? I imagine there are some things unique to d6+ that either wouldn't backport or would need to add in extra functions, but I am cool with that - as long as the module points this out in comments or something.
Anyone discussed this sort of thing before?

Comments

solotandem’s picture

solotandem commented
Status: Active » Closed (won't fix)

Reversing the process sounds like quite a bit of work. You would have to rewrite all of the search and replace routines. This is not something in my plans.

Going in reverse (i.e. creating "deadwood" from "goodwood") goes against the entire philosophy of this module. Why not bring all D5 sites up to D6 and be done?

danielb’s picture

danielb commented

Going in reverse (i.e. creating "deadwood" from "goodwood") goes against the entire philosophy of this module

Well I can't argue with a make-believe argument like that!

This isn't 'wont fix', this is 'wont try'.

jennycita’s picture

jennycita commented

Subscribing, just in case. I'm currently using an ad-hoc method to keep this straight: a 6.x (master) and 5.x git branch. The 5.x branch has a couple of extra commits necessary for the basic backport. Every time I develop something new, I rebase the 5.x branch using the 6.x branch as <upstream> and fix any conflicts. Then, of course, I have to commit everything into both cvs branches on drupal.org, all the time wondering if I've introduced any subtle bugs in the D5 version due to the changes.

Would love to hear how other people are handling this. I frequently see requests for backports of D6 modules in their issue queues and think a backport module makes some sense. Perhaps the n.x backporter could be removed from the available releases once Drupal n+2.x is released to keep things from getting out of hand. ;)