We are currently testing something like this http://www.ventismedia.com/mantis/view.php?id=6280

[quote="agraham"]
That's true... how about STARTING_UP. Set it right at the top of main() and clear it again as soon as the startup is done.[/quote]
Has the same shortfalls unfortunately, MM can stall at start-up for a number of reasons other than iPod/iPhone related issues (bad plugin, failure to load DB, etc.)

[quote="nohitter151"]It's a nice suggestion but what about in instances where something else crashes MM, ie. some 3rd party add-on, system failure due to power loss, etc. It would be quite confusing the next time you start up MM why your iPod/iPhone isn't detected.[/quote]

That's true... how about STARTING_UP. Set it right at the top of main() and clear it again as soon as the startup is done.

It's a nice suggestion but what about in instances where something else crashes MM, ie. some 3rd party add-on, system failure due to power loss, etc. It would be quite confusing the next time you start up MM why your iPod/iPhone isn't detected.

[quote="nohitter151"][quote="dswicegood"]Appreciate your reply Lowlander, it just appeared that nohitter had some additional info or a different opinion about a fix - maybe not[/quote]
I was saying that there is a fix for MM crashing after iTunes 9.2 was installed. I don't see a valid reason for complaint when a fix is available.

There is no Apple sync SDK. They make it as hard as they can for third party developers because they only want people using iTunes. More iTunes users = more people buying iTunes music.[/quote]

MM crashes basically whenever iTunes changes. _Eventually_ a fix is always released. I think that is a valid complaint. We all know digging out functions in private DLLs and calling them is a dicey proposition. As the op says, the calling process is likely to get corrupted if the arguments or function offsets change. Since we know that is going to happen we should design for it, otherwise the program is not correct wouldn't you say? His idea of communicating on a different process that can crash without taking out MM sounds like a good one but of course would be a big change - MM would then have to communicate with this new process, manage it, monitor it to detect when it crashes, bring up various dialogs to explain what happened, etc etc. All that code would probably cause more problems than the problems we are trying to avoid.

But one thing MM _could_ do is write out a little boolean to the registry: EXITED_NORMALLY. Clear it when starting up and set it when quitting. If it starts up and that bool is not set it knows it crashed last time and it can offer to start up without accessing iTunes. Then at least people can can use the app until a fix comes out. Especially those that don't even use the iTunes functionality but just happen to have iTunes on their computers.

Maybe something for MM4 if not 3.2.2?

[quote="dswicegood"]Appreciate your reply Lowlander, it just appeared that nohitter had some additional info or a different opinion about a fix - maybe not[/quote]
I was saying that there is a fix for MM crashing after iTunes 9.2 was installed. I don't see a valid reason for complaint when a fix is available.

There is no Apple sync SDK. They make it as hard as they can for third party developers because they only want people using iTunes. More iTunes users = more people buying iTunes music.

No, he was wondering if the fix for iTunes 9.2 was not working for you.

Appreciate your reply Lowlander, it just appeared that nohitter had some additional info or a different opinion about a fix - maybe not

As said so many times there is no fix for iOS4, it's not supported. MediaMonkey 3.2.2 is slated to add support for iOS4, keep an eye on the beta forum for the first beta release.

Guess that I falsely assumed that there was a spec in the OS4 SDK (There isn't???) which has been out in beta for quite some time... Weird, they use SyncML for most everything else...

[b]You are saying that there is a fix for the sync issue? [/b] I thought that I had the most current version installed, and it just gives some 'firmware version/read only' message when I attempt to do anything with the phone in media monkey...

First, what exactly is your problem? There is already a fix released. Second, what "spec" are you referring to? Apple provides no specifications for third party developers to sync with iPods/iPhones. It's all done on reverse-engineering.

No dates are released.

The problem with Media Monkey crashing is with Media Monkey. If they are using the plug-in model, then architecturally speaking, no plug-in should cause their app to crash. One, they shouldn't be talking with plug-ins on the main UI thread. Two, they should have a try/catch (or their language equiv) around the call-out to the vendor specific dll that hosts the communication functionality.

Ok, they don't have the correct dialect for conversing with OS4, but the Media Monkey app crash because of this is *their* problem, not Apples. I have been writing software for quite a few years, this is a rookie mistake.

The whole app should not crash because of any plugin. I believe the design pattern is called 'inversion of control', and they do not have a proper implementation of said pattern.

Now, that aside - is there a date for the release of the fix? And btw, the spec is out there now....

The problem is with Apple not MediaMonkey. Apple does this every release where they change the code to block out third parties like MediaMonkey and don't publish how to get access. The crash was caused by iTunes and changes Apple made and not MediaMonkey. The new version support iOS4 will be released when ready.

And as the news item states, version 3.2.1 fixes the crashing issue and was quickly released, now only iOS4 support is missing.

Wow,

This seems totally unacceptable from a company that makes such a good product. So all of us users that spent the extra dollars on buying gold are unable to sync because of new sync details and we get an application crash (which could have been easily avoided by a simple try/catch block or perhaps not attempting communication on the main win UI thread).

So, now I have a non-functional product that I paid money for and have to revert to using something else... Mitigating the application crash should be just a few lines of code, without it there - it makes the software appear sub-par to an end user. It should be unacceptable for a commercial software vendor to have missed this during the negative-testing phase.

When is this expected to be fixed? And, what is the refund policy for licenses? I have blogged this product and suggested it to sooo many iPhone users that I am sure I will hear about the crash in the next few days.

Haven't they had the updated sync spec published for some time? I have to imagine that a large portion of your users use iPhone 3g or greater.... seems like it would have made the new feature triage - especially since it makes iPhone sync non-existant (a lot of us bought it for this reason)

Awaiting a reply,
Dave