I have no real authority to ask this, I'm just curious.

 

Please only respond with 'yes' if you meet the following criteria:

 

1) You have a test server or a death-wish for the data on your production server

2) You have the time to put into testing betas, let's say a few hours per week

3) You have the time to give LimeTech useful feedback on what you find

 

If you respond with 'maybe, more info needed' please specify what type of info you need.

I put maybe because it would depend on how much work is involved. I know the current Beta testing I'm in with other devices takes up a fair amount of time to do the testing properly.

Me too. I don't know if I can commit to a time requirement. I may be more active some weeks and less others.

After 30 years working in the IT industry, I have learned that there is nothing like the randomness/unpredictability of live use to find the bugs in a system.  Using beta software solely in a test environment is likely to be more planned and prescriptive - while this should weed out the 'obvious' problems, it will miss a great deal.  It also requires a great deal of effort in order to be performed effectively.

 

[Yes, over the years, I have spent countless hours producing 'one-off' data repair utilities!]

 

I would argue that experienced and intelligent users who are prepared to install the beta software into their live systems should provide valuable 'pre-release' exposure.  I am willing to do just that - provided that I'm allowed to choose which s/w builds I'm happy to subject my live data to, based on reports from other users (in test systems?), and knowledge of what has been changed, based on the contents of the release notes.

I'd be up for it, i've collected a fair bit of spare gear over the years that I can use for testing purposes.

After 30 years working in the IT industry, I have learned that there is nothing like the randomness/unpredictability of live use to find the bugs in a system.  Using beta software solely in a test environment is likely to be more planned and prescriptive - while this should weed out the 'obvious' problems, it will miss a great deal.  It also requires a great deal of effort in order to be performed effectively.

 

[Yes, over the years, I have spent countless hours producing 'one-off' data repair utilities!]

 

I would argue that experienced and intelligent users who are prepared to install the beta software into their live systems should provide valuable 'pre-release' exposure.  I am willing to do just that - provided that I'm allowed to choose which s/w builds I'm happy to subject my live data to, based on reports from other users (in test systems?), and knowledge of what has been changed, based on the contents of the release notes.

 

Keep in mind that there's another step after beta: rc (release candidate).  I believe the current proposed scheme is that betas would be available only to people on the beta testing squad.  The squad's primary goal would be to find glaring, obvious bugs that crop up during normal unRAID operation, and perhaps some less obvious bugs that occur when add-ons are used.  Once a beta passes the squad's scrutiny, then it would become a rc and be available to the entire unRAID community.  It is at this stage that wide-spread hardware testing would take place all the quirky random bugs would appear.  Finally, an rc that survives long enough would become a stable release.

 

So I'm saying that it is fully expected that a closed beta will 'miss a great deal'.  However, with any luck it will catch the fundamental bugs that can lead to data loss.  I'm not saying people still won't lose data with rc's, but hopefully the prevalence of that will be drastically reduced.

Without voting (and without having read the thread either  ), my cheap opinion is that the beta should be kept closed... but easy to sign up for.  I know it's a bit of a management headache for LimeTech but rather than freely download the betas you should have to register for each minor version (e.g. for 5.0, 5.1, etc., which gets you access to -beta1, -beta2, etc)... which would just be the process of setting up a login account.  The registration process would include the appropriate disclaimers.

 

This makes it easily available to anyone who wants to test, but takes the heat off of LimeTech wrt stability problems and still allows LimeTech to target their testers (e.g. private forums, etc).

 

Personally, I played with -beta2 on VMWare but other than to check out the UI I didn't do much else.