Software Version Testing

[Ron]

When a new version of the DJ / Karaoke software that you are using is released, how much testing do you perform on it, before using it live?
Like most of my gigs are 4 hours, I usually do not hammer at the software for that same amount of time when testing. I try all the scenarios I can think of (switching the order of songs in the queue, changing the song in the queue deck several times while the active deck is playing, etc...), but just not over a 4 hour period. But if the software has some bad code it could start eating up RAM or have memory leaks that after so much time causes it to crash.

 

[drzinc]

The program I use has an automix feature so I let it run for 24 hrs after a new release if it passes that it gos out with me on my next gig.

 

[dunlopj]

The program I use has an automix feature. So I let it run for 24 hrs after a new release. If it passes that it goes out with me on my next gig.

And that is exactly what I did with Rockit! Over a year now and not one glitch!

 

[Ron]

Yea, but just because it can run a playlist for 24 hours doesn't take into account the constant use of the software's other functions: looping, setting multiple cue points, constant changes to the playlist, pausing, loading a playlist while songs are playing, searching the library, playing video or mp3+g, etc...

 

[drzinc]

Yea, but just because it can run a playlist for 24 hours doesn't take into account the constant use of the software's other functions: looping, setting multiple cue points, constant changes to the playlist, pausing, loading a playlist while songs are playing, searching the library, playing video or mp3+g, etc...

While in this test period everything is hooked up so I do a mock set. This way if there is changes in the way the program iteracts with my controller its found out diring this time.

 

[DjDennis]

lol when they release 9/10 times there is always bugs,

want to really trust it then go for it

I dont untill I know 100% its stable

 

[Proformance]

The best test is a trail by fire. There are some critical applications like a corporate presentation, or a bride's first dance, for example, where I MUST use a proven performer but, there are many more less critical opportunities where a ready alternate or backup source makes this kind of live testing perfectly acceptable.

It is not my standard to be "glitch free" rather, it is simply that no one should notice the glitches.

 

[Ron]

This is the way I tend to test as well. I do run it through it's paces at home a little, but trial by fire still seems to be the best way to go.

 

[nextgen1]

I load it up and come back the next day. I figure running 24 hours should be a pretty decent test. And that's not just the software. If I have the software on different computers then they both get the same test. And I will usually wait about 2 weeks after the release before I add it to my systems. Just to see if any bugs appear. You can usually get that info from the softwares forums. My playlists always have video and audio songs so that it will see both and interact with both.

 

[Ed Witles]

When doing a major update, I set a 6 hr automix set and record the mixer output - then go back and listen for anything odd.

What I cannot to is automate a 5 hr event (load/jog/pitch/etc).

But always have a backup system at the ready a few versions back just in case.