Forums » General
Yeah the 6k is a perpetual update. don't worry about it if stuff is working fine.
As for the mac version, we turned on gzip compression of the transfers so it's not counting the size of the files properly.
We'll turn off gzip as soon as we can.
As for the mac version, we turned on gzip compression of the transfers so it's not counting the size of the files properly.
We'll turn off gzip as soon as we can.
The 6k worth of data is intentional. It's the sum of a script patch, the version information, and the message of the day. And most of the time it's cached, so you just instantly "download" all 6k. So don't worry too much about it; the old updater basically just never showed this to you. The progress indicator doesn't work the same as it used to right now.
I'm pretty sure the Windows version works fine. That's what I've been testing my latest changes with; my latest changes probably break other platforms. Sorry about that.
Icarus, you had a corrupt vendetta.lfv, or it just had extra junk in it. I will fix that bug now.
Now I need to figure out the deal with the Mac version. I think that broke by me enabling gzip compression of those files on the web server; I'll disable it again and see if that helps. It could be some other bug though.
Once we get all the bugs worked out we'll roll my (now 7.9k and growing) updater patch into the compiled-in script.
update: server-side gzipping has been disabled for updater files.. see if this had anything to do with it
I'm pretty sure the Windows version works fine. That's what I've been testing my latest changes with; my latest changes probably break other platforms. Sorry about that.
Icarus, you had a corrupt vendetta.lfv, or it just had extra junk in it. I will fix that bug now.
Now I need to figure out the deal with the Mac version. I think that broke by me enabling gzip compression of those files on the web server; I'll disable it again and see if that helps. It could be some other bug though.
Once we get all the bugs worked out we'll roll my (now 7.9k and growing) updater patch into the compiled-in script.
update: server-side gzipping has been disabled for updater files.. see if this had anything to do with it
Nope. Now I'm getting 8/3KB with nothing else visible. D'oh!
Second try worked though. It's just being a little constipated.
Second try worked though. It's just being a little constipated.
(oops, I just messed something up.. fixed now. it was ray's idea :P)
BTW, we did determine that the server-side gzipping was causing problems for the mac version. We were using the content-length header from the response as the decompressed length, which was wrong.
BTW, we did determine that the server-side gzipping was causing problems for the mac version. We were using the content-length header from the response as the decompressed length, which was wrong.
/me slaps a1k0n
BAD Andy! BAD BAD Andy. RTFS!
(Yes, this is a JOKE!)
BAD Andy! BAD BAD Andy. RTFS!
(Yes, this is a JOKE!)
Ok, i'll ignore the 6k file thing.
now then, updating vendetta on my other box, an old win98 machine (i'll update the OS eventually, the CD is here somewhere) that i use a dialup connection for. (dont ask it's a long story)
I noticed a few things about the updater information display.
1. It is apparently confused by how long the downloads will take. It was telling me that a 500k file would take 1912hours 17:05 minutes. the time went UP as the file came in. (total download time was a couple of mins, not 1900 hours) Incidentally I hav eyet to see the time go down with the new updater, it just seems to go up.
2. the download progress bar seems to jump immediately to the "complete" stage, not moving slowly across the screen like it used to.
now then, updating vendetta on my other box, an old win98 machine (i'll update the OS eventually, the CD is here somewhere) that i use a dialup connection for. (dont ask it's a long story)
I noticed a few things about the updater information display.
1. It is apparently confused by how long the downloads will take. It was telling me that a 500k file would take 1912hours 17:05 minutes. the time went UP as the file came in. (total download time was a couple of mins, not 1900 hours) Incidentally I hav eyet to see the time go down with the new updater, it just seems to go up.
2. the download progress bar seems to jump immediately to the "complete" stage, not moving slowly across the screen like it used to.
For what it's worth, I'm still getting the 8/3KB and nothing else bug. And now it won't go through successfully on the second attempt.
=/
EDIT: Third time worked. But I was getting it after you guys reportedly fixed it.
=/
EDIT: Third time worked. But I was getting it after you guys reportedly fixed it.
Spellcast, yeah.. That's because when Ray ported the Windows version, he switched the 'already downloaded bytes' and 'total bytes to download' fields to the progress indicator callback. :P That's fixed. We'll roll that patch out Wednesday probably, and figure out what the deal with the Mac version is in the meantime.
Ys know, Andy...I've seen you shift a lot of blame to Ray lately. Is he really gumming up the works or are you deflecting, afraid to admit mistakes? Lay down here on my couch and tell me all about it...
DELETED!@
lol
I'm getting segfaults from the updater (Linux).
I got mad at it and decided not to stop hammering the update server until I got a strace of the darn thing.
Submitted a bug report about it, and there's also an email address there.
Drop me a line if you want the strace log.
I got mad at it and decided not to stop hammering the update server until I got a strace of the darn thing.
Submitted a bug report about it, and there's also an email address there.
Drop me a line if you want the strace log.
Same (linux), segfault with the same error msg as previously mentioned...
Interesting. Seems that there's a race condition somewhere. We will find it Soon™! We may already have a solution.
I'm not trying to deflect blame onto Ray.. it's just as much my fault for having unclear code for him to port. I just say it jokingly, probably too often, but I don't have anyone else to blame since we are woefully understaffed. In reality, Ray writes ten times more code than I do with the same time constraints for bug testing.
Certainly if the Linux version is crashing, there's nobody to blame but me. Or the author of libcurl... ;)
I'm not trying to deflect blame onto Ray.. it's just as much my fault for having unclear code for him to port. I just say it jokingly, probably too often, but I don't have anyone else to blame since we are woefully understaffed. In reality, Ray writes ten times more code than I do with the same time constraints for bug testing.
Certainly if the Linux version is crashing, there's nobody to blame but me. Or the author of libcurl... ;)
Blame Canada?
Andy...I still think you're blocking. Please, look Ray in the eyes and tell him exactly what you are feeling.