Has D failed? ( unpopular opinion but I think yes )

[Laeeth Isharc]

On Sunday, 14 April 2019 at 05:35:41 UTC, silentwatcher wrote:
> On Sunday, 14 April 2019 at 04:48:00 UTC, Walter Bright wrote:
>
> i think there are some other things you don't understand yet.
> go back in this thread and read what people think about D and 
> why they think it should be forked.
> you guys made lab experiment of it instead of finding out what 
> the real world programming needs. why are people dissatisfied?
> stop finding all those mediocre things/solutions like help here 
> or do that. i admire the gdc project but i would not use it.
> my experience is, that programmers (at work) look at D and are 
> afraid of using it. to large, to many features that are nice 
> but nobody needs, web like syntax, the GC and after reading in 
> the forum, they are even more afraid to use an unstructured 
> experiment.
> then there no tools, libraries etc., just visuald and that is 
> the only reason why some try D for a short time.
> well now will tell us again - write the tools/libraries, but in 
> a business you don't have the time to do that. all those things 
> come with c++, c# or java. why should i invest time and money 
> into D?
> no time to experiment.
> i like D, followed it for a long, long time, but it turned into 
> a moloch. some of us still use D1 for little tasks. i hope 
> somebody takes D1, strips the gc, adds templates, cleans up 
> this mess and forgets about the rest of the bloat.

Maintenance constitutes the largest part of the total cost of 
ownership of code.  In other words when the code is 'finished' 
then you have made only a down-payment.

If you don't want to invest your own time into working with D, 
then I don't think you should.  I wouldn't dream of trying to 
change your mind.  But there's clearly some ambivalence there 
because why else would you be posting.

The most important aspects of D might be the values and good 
taste.  When I decided to learn D, there was nothing calculating 
about it - I just wanted to find a language I could stand to 
program in because otherwise I wouldn't be doing much programming 
as life is too short and it's not my only interest.  One thing 
led to another and working from my home office led to consulting 
and then being in charge of technology and part of research for a 
150 odd person firm.  I don't know, and it's a longer story but 
probably that wouldn't have happened if I hadn't been drawn to D.

Turns out also that the values that characterise the language and 
the community are peculiarly suitable for the commercial context 
of what's still an early stage firm that has grown very quickly.

It's true that there's a high barrier to entry in the beginning 
with D - much better now, but if you are put off by discomfort 
and needing to figure things out for yourself then it's going to 
be quite a tough experience.  D doesn't place a high value on 
being accessible and if that's what is important then maybe 
somewhere else would be a better fit.  On the other hand it's 
also a moat that constitutes a filter - as it happens 
commercially people who are okay with being uncomfortable, with 
having to figure stuff out, are quite a good fit for our 
business, especially at the stage we are at.

It's a really very big world, not like it was when I started to 
learn to program in 1983.  Most software is developed by private 
organisations that don't talk much about what they are doing.

I never thought that D would be considered a  substitute for Ada 
and Pascal but that's been the case in one quite interesting 
commercial application for ship design.  Similarly for us D - and 
a little functional domain specific language written in D - has 
been a replacement for Excel spreadsheets and SQL reports.

Brutal upfront sometimes, but the upfront emotional experience is 
the least important part of the cost of developing and 
maintaining software.

The productivity experience has a different profile too - you 
spend time at different points on different things, but looking 
at our own experience from a commercially perspective I really 
don't think it would have been possible to be as productive 
taking a different route.

There are quite a lot of libraries now - 1500 last I checked on 
code.dlang.org

Calling python libraries is easy and calling C libraries using 
DPP is also easy.  DPP keeps improving and in time it will handle 
much more of C++ headers.

Yes - the tolerance for experimentation is an important question. 
  In firms where there isn't much of a tolerance for experimenting 
and for some of those experiments to fail then it's probably not 
the right environment to use D.  But I think the more interesting 
places to work are quite different.