I'm posting this issue because it's the conclusion I've reached after several attempts to use AI 7+.
It's apparent now that the Enhanced UI feature is a pretty disruptive code change/addition to AI. Each successive attempt I've made to use the feature has resulted in show-stopping issues for me. Here are all the issues I've hit since AI 7 release:
- http://www.advancedinstaller.com/forums ... f=2&t=9105
- http://www.advancedinstaller.com/forums ... f=2&t=9111
- http://www.advancedinstaller.com/forums ... f=2&t=9371
- http://www.advancedinstaller.com/forums ... f=2&t=9405
- http://www.advancedinstaller.com/forums ... f=2&t=9406
- http://www.advancedinstaller.com/forums ... f=2&t=9408
- http://www.advancedinstaller.com/forums ... f=2&t=9409
AI is obviously a complex product, with many permutable, interrelated features and options. That is AI's strength, but also its vulnerability with respect to product quality and reliability.
Nonetheless, I think that most or all of the issues I've hit could have been caught before release, if the following were taking place at Caphyon:
- a product philosophy that dictates quality is more important than features
- robust test suites including a large set of setup projects that exercise all features and reasonably representative combinations thereof
- a commitment to full regression testing when disruptive features are added/changed
- a test philosophy that assumes anything could go wrong, and will
- explicitly detailed test cases that confirm data and UI outcomes
- beta testing with willing customers
I say this because, as a customer, I feel that I have been an unwitting tester of several unbaked releases. Although the occasional bug is expected, several glaring failures have left me with the impression that Caphyon's test (and possibly design) procedures are currently falling below standard.
To state the obvious, I expect the product to work.
The set of issues I've hit in AI 7+ have reduced my faith in Caphyon's ability to deliver the features it aspires to. My attitude toward incorporating new releases has gone from "confident and willing" to "suspicious and skeptical". Indeed, I fear that even if each of my outstanding issues are fixed, there will still be significant new ones lurking in the next release. This is obviously a problem for a customer such as myself who pays for an AI maintenance plan.
Caphyon, please slow down and revise your quality standards and test procedures before releasing another version.
Once again, reverting to AI 6.9.1.
Suggested Articles
Re: AI 7.1.1 not well tested
Hi,
I am sorry to hear that. The very reason we are making releases so often is to discover problems like these as early as possible and fix them. If you are not satisfied, we can refund your license at any time. Just let us know.
We are working very hard on making AI as robust as possible but indeed sometimes things slip through and we can only try to fix them as soon as they are discovered.
Best regards,
Ciprian
I am sorry to hear that. The very reason we are making releases so often is to discover problems like these as early as possible and fix them. If you are not satisfied, we can refund your license at any time. Just let us know.
We are working very hard on making AI as robust as possible but indeed sometimes things slip through and we can only try to fix them as soon as they are discovered.
Best regards,
Ciprian