[Info-vax] User Interface Design, Implementation

[Arne Vajhøj]

On 6/9/2019 9:46 AM, Richard Maher wrote:
> On 9/06/2019 7:50 am, Arne Vajhøj wrote:
>> On 6/7/2019 9:56 PM, Richard Maher wrote:
>>> On 8/06/2019 7:37 am, Arne Vajhøj wrote:
>>>> I can observe that the majority of client side developers use
>>>> those frameworks.
>>>
>>> Yes, and every job advertised wants them 'cos Gartner or some other 
>>> dick lied to them that they were "modern" javascript Frameworks.
>>>
>>> Long before PWAs and Custom Elements.
>>>
>>> Facebook doesn't even use AngularJS on Apps that count and then 
>>> blames the web for being too slow.
>>>
>>> Like jQuery Angular/React's time has come and gone.
>>
>> When the vast majority pick a given solution and stick
>> to it consistently over the years, then I start to think that
>> maybe the solution do provide some value.
> 
> You mean like COBOL, Java, ASP, jQuery, McDonalds?

Yes. I believe that they all provided something that was needed.

COBOL: in the 1960's when you had to pick a language for
some business computing then COBOL was a no brainer compared
to Fortran II, Lisp and ALGOL, in the 1970's the COBOL
was still a pretty good choice even though PL/I and some Basic
dialects had arrived, and even in the early 1980's it was a
strong contender even though C and Pascal had arrived, in the 1990's
and later was probably not so obvious but then it had 30 years
of code base.

Java: when Java showed up OO was languages like C++ and Object-Pascal
suffering from lack of portability, memory leaks etc. - Java was a
pretty obvious choice, sure EJB's was a bit complex but they were still
way easier than CORBA, if Java had been invente dtoday then Java would
not have looked like Java but more looked like Kotlin as things have
progressed since then.

ASP: when ASP showed up the common way to create dynamic web content was
CGI scripts in Perl or C or something else - the template with embedded
code and the COM model was a huge improvement, PHP and JSP and ASP.NET
and all the rest also went that route and today ASP look old and
primitive.

jQuery: there was a need for a standardized way of doing this stuff
instead of having every developer come up with his own - jQuery filled
that gap.

> And here was I thinking Computing was a science and not based on hearsay 
> and superstition :-(

Some parts of computing are math that can be proved correct.

Other parts are way too complex to be proven or even analyzed
theoretically and instead rely on the "best practice" concept.

Arne