Re:2 more Bank / Bag suggestions

There are some theories of varying amounts of accuracy (or inaccuracy) about how our code is structured in here, but yes, it is a very risky change to make. In fact, people have attempted to re-write it several times. Each time the result was MANY unanticipated consequences so each time the effort was scrapped.

This has been on my radar since I started on the project (if nothing else due to the challenge level). But as lead I have to make choices about not only what is good for the game (which this would obviously be) but about risk as well.

The risk isn't really about corrupting files, it's more about introducing crashes and especially exploits. Over the years as hackers have made various intrusions into the system, the little holes here and there have been patched. So the inventory handling code (where MANY exploits have occurred) is literally a patchwork of fixes, etc. Collecting all of those lessons leanred into a gleaming new re-written inventory system is virtually impossible to do in rigorous fashion. So lacking a set of unit tests for every one of those exploits you are left falling back on testing. And that means you either have to employ hackers as testers, or rely on code reviews, etc.

It's a bit naive to assume that enough QA can solve all issues of this nature. QA (and subsequently the test server) catches an amazing number of problems. And of course I would always love to have more QA resources, but relying on QA to catch subtle issues of this nature is asking for long days and sleepless nights.

That's not to say it'll never happen. It's still one of those issues that is irritating to me that we haven't been able to address.

So if you want someone to blame (other than the original coders and original tight scheduling). Blame the hackers.





More...