The State of the GUI

[Joakim]

On Sunday, 28 October 2018 at 07:19:11 UTC, rikki cattermole 
wrote:
> On 28/10/2018 8:12 PM, Joakim wrote:
>> On Sunday, 28 October 2018 at 07:08:33 UTC, rikki cattermole 
>> wrote:
>>> On 28/10/2018 8:03 PM, Joakim wrote:
>>>> Building your own GUI toolkit is very much an '80s or '90s 
>>>> mindset, back when that was paramount. It isn't anymore, and 
>>>> would be a waste of time to try with D.
>>>
>>> WIMP dates back to the 60's in a fully functional state.
>>>
>>> It isn't going anywhere.
>> 
>> Oh, I never said it is: I explicitly said it will always be 
>> around like COBOL. When's the last time you wrote any COBOL or 
>> interviewed for a COBOL job? Exactly. :)
>
> The last time I worked with COBOL was 4 years ago when I wrote 
> a parser for it.

Then you're the exception, as I suspect most programmers are like 
me and have never even seen it.

> There is still a very large market for COBOL developers: 
> https://www.cobol-it.com/

I agree that it's still a surprisingly large market, as COBOL is 
still part of key infrastructure, particularly in finance and 
government:

"In the US, around 80 percent of in-person transactions and 95 
percent of ATM swipes are based on programs written in COBOL. The 
problem is there’s not enough people to maintain the current 
COBOL-based systems.

According to Reuters, around three trillion dollars in daily 
commerce flow through COBOL systems. Many major financial 
corporations and some parts of the federal government have built 
their entire infrastructure on COBOL bases from the 70s and 80s."
https://thenextweb.com/finance/2017/04/10/ancient-programming-language-cobol-can-make-you-bank-literally/

Similarly, I think WIMP will be around for decades for certain 
kinds of software, but it will simply be dwarfed by sales of 
software that's driven by voice, just as mobile device sales now 
dwarf PC sales:

https://twitter.com/lukew/status/842397687420923904

So like Adam said, should D focus on that shrinking GUI market or 
the growing text/voice mobile market? Obviously the latter 
deserves much more emphasis.

That doesn't stop Adam or anyone who wants to do GUIs or COBOL 
from going their own way, I'm just cautioning that it wouldn't be 
a good direction for the D community to pour much resources into, 
ie the 5 paid developers he references.