Originally Posted by cyberstorm
How about some kind of volunteer distribution system? On the top of my head - how about a torrent solution? Put all addons in a single torrent - it's at most a couple of GB.
|
*beats head on desk*
A lot of this was covered in many threads on the WoW forums, but there are many reasons why using a torrent solution would be inefficient (grossly so, in many cases) for add-on distribution.
Originally Posted by Elhana
I do not agree that temporary fixes should not be implemented at all - some addons may be broken for ages for different reasons, yet addon is still "maintained". Just search for "fan update" here and you'll see how many people actually putting their dirty hands into your cookies, even more if you look for hacks in comments. Actually WoWI shall implement ability to upload custom patches, that would help if someone wants to implement some minor fix without forking addon completely. Examples: Quartz hack for Deaden (BT RoS), recent NeedToKnow self debuffs tracking hacks.
No, that doesn't create any version issues - either updated beforehand and there is nothing for WM to fix or it is "temporary fix" getting updated on both official source and WM when new "official" version comes out.
|
Wholeheartedly disagree with this. What you're advocating is the ability to modify and redistribute without working with the actual author, which just causes even more support headaches. If user A has a problem with a silently-"fixed" v1.1 of author B's add-on, that author may waste a lot of time trying to debug code that they didn't write or may not know about (especially if the revision wasn't altered to reflect the modification). It's a catch-22 from a user perspective as well. If WM applies a "temporary fix" but doesn't increment the version number, the author now ends up trying to support a "fixed" version which behaves differently from what was originally written. If they do increment the revision number, the author is equally stumped, not having released that version themselves, and in the worst case, the "fixed revision number" is the same as a new upstream release's revision number....but with different code. In either case, you end up with a frustrated author
and frustrated users if problems crop up.
Ideally, all patches should be submitted to the author for consideration and inclusion, and let them determine how they want to handle it. If they decide to leave it broken until they do a rewrite/other implementation...oh well, they're well within their rights to do so. I've personally held back submitted "fixes" in large software projects because the visible bugs revealed a larger issue that I needed to resolve. Sure I could commit the cosmetic fix and release a minor update, but I'd just be doing the same thing we've already talked about
: hiding a real bug and pretending I've done something to fix it. In every case, it was better for me to just work on fixing the real problem or find a better solution.
Lastly, remember, the add-ons are effectively "owned" by their authors, not the users. Add-ons certainly don't have a life of their own either (meaning they aren't independent of their authors unless they're "freed" via licensing). If an author wants to quit the game and never relinquish their rights to their code, you may as well forget the add-on and write your own clone of it because you have very few options otherwise.