I agree with both points .
This comes back to the user developer divide , and this is an important point .
The wallet.dat all being called the same file is incorrect if the intention into have the end user use a essentially beta / experimental piece of software .
The wallet.dat is correct as it is if this software is still in open source development , and there is another essentially linking peice of code that is needed to bridge the Developer User divide .
Both are very important , obviously if you are a developer and you don't listen to investors you may as well start investing your time into making a cool little program that notifys you each day that " you fucked it all up"
However there could be an argument that as this is open source development in progress then , if you choose to use the Qt design then learn how to back up folders .
But the topic is useful and correct in that there is an end user issue there .
In a " multi wallet" something that I see as a scenario in the future there would be use to have individual named wallet.dat as when one switched between currencies there would be less chance of a bug or error , but I foresee that being a computer generated code name .
In other words the user shouldn't be accessing the wallet.dat and the fact that they have to means the software is flawed .
Exactly, yes. The user should not be seeing any of that stuff, think more like how a website is done, the user never sees inside, even having them see extensions on the page names like .php or .shtml or whatever is extraneous code-leak. Going back even farther actually even just making the users learn geeky crap like dot com, dot net, dot org was a very very hard sell, what they actually wanted was more like AOL keywords or even the google search engine input widget, having to even have heard of "http://" was and maybe still is totally insane geekspeak leak, that is why browsers rapidly defaulted to appending dot com on the end and assuming http if not specifically told to use some other protocol such as gopher or ftp or mailto or bitcoin or any other prefix any user or group of users might choose to teach their browsers to fire up some cool user-side handler for.
Ignorance is desperately desired on the user end, heck now we have computers reading and writing should soon be as out-dated and useless and obscure as knapping flint to make stone age implements, users should just cry and say "baby want, mommy get!" and their needs automatically be taken care of... sigh, being in this primitive pre-holodeck age is such a trying life, isn't it?
Ultimately though maybe either users learn to teach their computers, meaning in effect users become programmers, or computers end up having to teach their users, hopefully eventually teach them to teach their computers...
Hey we are back full circle to open source: "if you want it programmed properly program it yourself!"
-MarkM-