svn-1417 OS X feedback

[Anonymous]

All seems good except the false hanging issue on initial startup is still present. A simple stop and restart seems to fix it, and I’m sure most mac users have figured this out now 🙂

[rpedde]

@andyg wrote:

All seems good except the false hanging issue on initial startup is still present. A simple stop and restart seems to fix it, and I’m sure most mac users have figured this out now 🙂

That one really is a rendezvous bug, but I did find out that one cause of it is really long share names (but it should be checking for that now, and logging a better error if it happens).

At some point I want to integrate growl notifications for the events on mac, so I’ll be changing how the helper app talks to the server. Once that’s done, the startup stuff should go away.

— Ron

[marcv]

First of all: Praise! Stumbled into Firefly yesterday, and I love it; easy and geeky at the same time. and above all: really fast with low CPU-usage.

My problem is this: I don’t want to have Firefly use my iTunes library files, but instead have it scan my mp3’s (scattered over several drives/partitions) independently. (I’m using iTunes6, but this is even more handy with iTunes 7 because it can load different libraries; in that situation it’s better to have all music still available on the remote side).
So I make a folder, put symlinks in it to my folders containing mp3 etc, and point Firefly to it. I had the setting for follow symlinks on “yes”.
This worked in svn-1359 but not in svn-1417.

Thanks, marc

OS X 10.3.9 on Beige G3 (IDE drives)
svn-1400

[rpedde]

@marcv wrote:

So I make a folder, put symlinks in it to my folders containing mp3 etc, and point Firefly to it. I had the setting for follow symlinks on “yes”.
This worked in svn-1359 but not in svn-1417.

Gack. Something strange with stat on osx? Grumble.

I’ll have to look at that. Meanwhile, you can work around by adding multiple folders to your music path — you’ll want to do that through the web interface though.

Meanwhile, I’ll log a ticket for the symlink thing. Thanks.

— Ron

[marcv]

rpedde wrote

Gack.

🙄

The adding multiple folders from the webinterface didn’t seem to work;at first. Just installed Firefly on a 10.4 machine, noticed it worked on that one; probably a script-problem with Safari in 10.3.9, will try Firefox for that.
BTW: couldn’t share my iTunes 7 ===> iTunes 6 from iTunes7; another really good reason to run Firefly on the iTunes machine as well.
(I’m hooked!)

BeigeG3/10.3.9/iTunes6/
iMacG5/10.4.7/iTunes7
both svn-1400

[hm42]

When using Roku, firefly process sometimes starts to eat all the available CPU and slows down the system noticeably. Happened twice so far. In this state it also stops responding to clients. The process has to be killed, it won’t settle down by itself.

Also, it sometimes is not fast enough to “wake up”, ie when connecting to the server that has been idle for some time, it takes two connect and browse requests to wake it up – first attempts result timeouts on Roku.

Compilations handling still does not seem to work quite right; only some of the compilation stuff is under various artists.

[rpedde]

@marcv wrote:

rpedde wrote

Gack.

🙄

The adding multiple folders from the webinterface didn’t seem to work;at first. Just installed Firefly on a 10.4 machine, noticed it worked on that one; probably a script-problem with Safari in 10.3.9, will try Firefox for that.
BTW: couldn’t share my iTunes 7 ===> iTunes 6 from iTunes7; another really good reason to run Firefly on the iTunes machine as well.
(I’m hooked!)

BeigeG3/10.3.9/iTunes6/
iMacG5/10.4.7/iTunes7
both svn-1400

Found this, btw — I made a wrong default.. double-check your web interfacethat follow symlinks is YES, and then do a full scan from the status page. That really should do it. Does here on 10.4.8, anyway.

[rpedde]

@hm42 wrote:

When using Roku, firefly process sometimes starts to eat all the available CPU and slows down the system noticeably. Happened twice so far. In this state it also stops responding to clients. The process has to be killed, it won’t settle down by itself.

Also, it sometimes is not fast enough to “wake up”, ie when connecting to the server that has been idle for some time, it takes two connect and browse requests to wake it up – first attempts result timeouts on Roku.

Compilations handling still does not seem to work quite right; only some of the compilation stuff is under various artists.

Version? This sounds like a svn-1313 bug, fixed in later. Probably should be running 1359.

The wakeup thing I can’t really do anything about… if the thread does a db read and it takes a while for the spindle to spin up, I can’t program my way around that.

I’m of the opinion that the soundbridge times out too fast, but I’d guess that’s probably a tuning issue — how long you wait for spinup, versus how long you have to wait when you accidentally click on a rendezvous entry that isn’t on the network anymore.

If you aren’t running 1313, I’d really like to see some of your logs, though. Haven’t seen that before, would like to see what that is.

— Ron

[hm42]

@rpedde wrote:

Version? This sounds like a svn-1313 bug, fixed in later. Probably should be running 1359.

svn-1417 running on mac os x 10.4.8.

At the default debug log level only these errors in the system.log seem to have happened around the same time, don’t know if they are related:
crashdump[8379]: bfobserver crashed
launchd: com.apple.dnbobserverd: exited abnormally: Bus error

I’ll run the server with a more detailed log level for a while and wait for the issue to reappear.

– hannu

[rpedde]

@hm42 wrote:

@rpedde wrote:

Version? This sounds like a svn-1313 bug, fixed in later. Probably should be running 1359.

svn-1417 running on mac os x 10.4.8.

That’s a new one, then. I wouldn’t mind seeing logs of when that happened.

At the default debug log level only these errors in the system.log seem to have happened around the same time, don’t know if they are related:
crashdump[8379]: bfobserver crashed
launchd: com.apple.dnbobserverd: exited abnormally: Bus error

bfobserver is notorious for sucking up ram, and pushing machines into swap. Wonder if that’s part of the problem. When you noticed the high cpu, was it high user cpu, or high system cpu?

I’ll run the server with a more detailed log level for a while and wait for the issue to reappear.

– hannu

I’d appreciate any help you could provide in that direction. Thanks.

— Ron

[Author]

Posts