jbrinkman wrote
More people with access to betas = more support issues. Been there done that, have the thousands of garbage gemini posts to prove it (just ask Alex and Shawn). Instead we have chosen to move to limited betas, with a manageable number of testers, who each have a strong business interest to ensure that DNN continues to work well and be relatively bug-free release after release. These are not people who have a passing interest in seeing the latest DNN version, but people who have a vested financial interest in ensuring the success of DNN. I will take a couple hundred really dedicated beta testers over a 100,000 interested users any day, because the former are likely to give me better feedback about real issues and not have us wasting time educating users about basic DNN concepts. We get enough posts in the install-it forums to tell us that an open "beta" will only add confusion to what needs to be a very methodical process.
Joe... Believe me I understand COMPLETELY. If you think that Beta issues get nutsy on DNN try working on "AAA" games, I guarantee you its "flip out" time :)
PC Entertainment software (for the most part) did away with wide-beta's years back. But, a wide-beta might work for DNN. In other words, simply dont call the RC an RC, call it a Wide-Beta. Now all the people who have basically no code background or commercial experience in the "mass market" software world will be jumping for joy. "I get to be a Beta!". I know some folks who went through this with the Postal Service new release of their software for shipping, overjoyed and enthusiastic about being a beta. The software actually to me is so bad we coded our own in house of the USPS API after fighting with them for 1/4 of a year to get full API access. Welcome to politics.
I dont now know what the actual installation numbers are for DNN 4+. Downloads of course do not equal permanent installations, in fact, in most cases of open sourceware apps 1 in 50+ actually decide to go forward with the stuff which is why so much OpenWare turns into "never worked on again ware" as the community does not exist to entice the engineer(s) to continue onward.
If what the other guy says is true, about 15 real testers exist then thats a problem indeed. DNN should have at least 40-50 engineer based Beta leads. These Beta leads should then have 5 Beta subordinate webmasters under their charge. The JOBS of those sub's are to locate bugs, document them as well as they can (using a standardized form, format, repeatable on and on, standard sommercial industry crap) as well as bring forth new ideas. The Beta leads work to find/solve engineering issues and pass the solution(s) to the core. If they cannot solve them, then the core need be notified of that. ANYONE who opt's to be a Beta ***MUST*** allow both their Beta Lead or anyone from the Core ACCESS to their portal install, so, bugs can be looked at DIRECTLY (an advantage DNN has over non web apps).
Some form of Beta Sub standard need be in place as well. For example in games they like a very diverse range of hardware (though not as diverse as it used to be years back). So if I'm a Beta Lead and 2 of my Beta Subs are on DiscountASP.Net, thats not what I want.
Further... the Beta's work on Core Framework's and the Core Projects as far as installs. With DNN Beta's cannot fart round with Joe/sephine Johnsons 3P module Add-on's. This simply screws the pooch as far trying to locate core bugs/core project module bugs. Beta's are EXPECTED to not only POUND on the software but also come forward with IDEAS, documented ideas. Again... All materials standardized formats, not slip-shod this/that.
After the core is confident that the framework and modules are in good working order.... Wide-Beta.
Now all the forum peep's and such can use it and if anomolies occur can be dealt with. I am sure its frustrating to the Core Team and Core Module Leads/Subs to read these forums and see people seemingly with inflamed testicles. I know exactly how this feels to Core Team people, frustration, anger, why am I doing this? on and on. Core team members TIME is better spent (as you noted) working the code/app not chasing around people in the forums trying to defend what many cannot even understand.
Thats not to *Flame* them, the end customer (be they paid of free) is the end customer. They want things to work. So does the core.
I have dug into the DNN code (boy I wish it were better documented :) )... Its excellent. I am still attempting to "grok" it. I'm considering just crash coursing C# (I know C++) as VB.NET 2005 appears to be a butchered C# sorta and the C# community appears far more prolific in .NET World than VB is.
The core team may wish to consider what I have stated here. The "RC" comes after Wide Beta. Wide Beta keeps all the webmasters peppy! Meanwhile a real Beta Lead Team properly organized each charged with dealing with some webmasters works the problems. Offloads work on the servers (you and others LOL).
The core (as I noted in your blog) needs to consider who is most important. I would state its the customers, those with commercial interests or not is immaterial as they are using the same ware. In releasing a RC and sites running into woes is a BIG turnoff to customers or prospective customers. Instead, whatever the stable release is, thats the release. New releases are ALWAYS wide-beta, "Use at your own risk" and people will... Look at Brian, for example. Itching to get his hands on it.
While Brian may come across as somewhat brazen believe it or not, thats good. This is exactly they type of Beta tester that fits nice under Beta Leads. He's not in for popularity, he's excited, he wants to use it, he wants to beat the snot out of the software. He has ideas as to how this or that might be better. Maybe he's wrong in some, right on other things. Thats a perfect beta tester. However, the "core team" never need be exposed to it. A Beta Lead is exposed to it and said lead makes sure things look like a real bug, try and figure it out, make sure thigns are documented and ideas forwarded or expanded upon.
You said it yourself Joe, the "Community" is DNN's strength. If you have as was stated "15" Platinum Benefactors then slap 5 Beta's under em', now you have 75 installations where the software can get pounded on and ideas can stream forth from. Get some of thsoe Beta Leads to deal with some of this forum jazz as time allows. Perhaps those (since we are talking all developers) can get the reward of "DNN Core Approved" developers slapped on em'. Perhaps allow them to peddle some modules at the DNN site off some link as appreciation for what they are doing.
Heck, Mike Washington is my new hero as he has taken the time to create some simple module development that can actually be followed. I have the DNN book and well... there should be three books. Book "A": DNN WEBMASTERS - Book "B": DNN DEVELOPMENT FRAMEWORK A TUTORIAL APPROACH - Book "C": THE DNN CORE SOURCEBOOK .... I think the names describe what I am saying. Heck, you all could make 5 books out of it taking Book's B & C into one each for VB and C# respectively.
Some solutions to continual problems might also get addresses as long as I am at it. Connections strings 101 :) ... Perhaps it would be a good idea at part of the DNN site to have a Connections Strings database. So as users come in they get asked, "Please take the time to submit your connection string if your host provider is not listed in the DB" and flag the users accounts as they do so they are not asked on next login. Simple.
While we are at it, how about a small section of good places to host. I know that DNN has whatever affiliate deals they are with varied hosts for referrals. Thats fine. Feature them and slap the rest in some lesser visible point on the page.
First and foremost I would hope taking away from my rather long post is that DNN realize that your financial interests (and that of others) depends on the end consumers of the product. The key to continual success and real movement is those customers feel happy/empowered and supported. Wide Beta does just that, "Here ya go, go play... but understand, its a Beta..." The core need not respond to "Issues with the Beta" other than if finding someone who seems to know what they are talking about become a Sub or Beta Lead.
Responses to Wide Beta Beef's are : "The stable release candidate is version n.this.that. Wide Beta x.where.what is an experimental version of the software and thus may have unpredictable results in usage. We are constantly working on improving DotNetNuke and our Wide Beta service helps us steer the direction of DotNetNuke as well as improving the performance of the software". Thats it.
No guarantee that the Wide Beta will ever be an RC. However, if it does come to pass the Beta Leads and Subs have pounded on it so when it does become a RC the folks out there are happy, your happy, everyone's happy and we can work on the next thing... solving Global Warming. :)