nicksan wrote in post #9903936
How can you fix something that behaves a certain way by design?
In other words, if Canon doesn't think it's broken, there's really nothing to fix. ...
I don't own the lens (yet), and this is probably just me sorting through semantics and being verbose ... I have a little ancient history working within Quality Assurance and Quality Control processes in an unrelated industry. Highly engineered and tightly spec'd QA processes and QC procedures are 'living' processes and procedures in the sense that they are connected and very reliant on product performance and customer service reporting. At least, that is, when they are run well.
Once a frequently occurring problem emerges, internal testing steps determine whether it is a field use issue, a distribution issue, a manufacturing issue, or a design issue (those are my general terms). The proper tendency is to rule out each possibility from the way a customer uses the product in the field in a reverse process direction--as far back as it takes to identify the source or sources that cause the problem. Validating the R&D and design is certainly active, too, but it never makes sense to throw away product development and tooling expenses until you know you actually, factually have a design problem.
That's a long way to say that when we hear a lens is "not broken" it really only means that they haven't identified any problems with the actual design. Of course it could be anything else in manufacturing or distribution and they could correct the issue. Theoretically the lens (in this case) would begin to perform better as if there were "unofficial changes." It only means they might now be doing something better than they were doing before, but the CAD drawings and key specs haven't changed at all. I hope that's the case. I have my eye on this one after I choke down my tax liability next week. 