Hi Ariel, On Donnerstag, 2. September 2010, Ariel wrote: > Please reverse saytime 1.0-22 from testing, and put back 1.0-21. thats sadly not directly possible. What is possible is either removal from testing or reupload of a fixed version... > 1.0-22 has a grave bug #587124 which makes it totally unusable, plus, > although I don't wish to insult the maintainer, it has been rewritten very > poorly, and is not production code. > > 1.0-21 works OK (but not great, it only works with oss, not alsa). > > I would like to write a patch to make saytime work with alsa (I've offered > this to the maintainer, and he accepted), but if I do it'll be based on > 1.0-21, not the rewritten 1.0-22. the issue is "just" that 1.0-22 misses some bits of 1.0-21 as the -21 (and -20) uploads were done badly, and I missed that when uploading 1.0-22. In 20+21 the source was modified directly, without patches... (And no arguing needed that my -22 upload was also bad...) I have a mostly fixed -23 sitting here on my harddisk. > And if I don't get to it, at least something that sort of works will be in > squeeze. I'd still welcome your improved patch, as indicated in the BTS. (Just indicate against which version you base it...) I've looked at the existing patch and was unsure how to handle it, since a lot of logic was added to be able to choose between oss and alsa, which is unneeded as at least in the Debian context as oss is dead. So whats left is basically just a sox wrapper for alsa... I hope I'll finally have time for this this weekend... patches certainly welcome. cheers, Holger
Attachment:
signature.asc
Description: This is a digitally signed message part.