Where will D sit in the web service space?

Laeeth Isharc via Digitalmars-d digitalmars-d at puremagic.com
Thu Jul 23 10:03:30 PDT 2015


On Wednesday, 22 July 2015 at 21:38:14 UTC, Joakim wrote:
> The finance guys seem to be coming on board, the Dconf '15 talk 
> by the fund guy, Smith, probably helps.

Yes - that's my impression, too, and I am doing my little bit to 
encourage people to consider the benefits of D.  People are 
sometimes more open to hearing something from a practitioner and 
business person who knows a little technology than from a pure 
tech guy.  Finance is a funny area because the stakes are high, 
but people are much more open-minded in some parts of it to 
trying things (it's not always as enterprisey as more stodgey 
firms in this respect).

There's a crisis in the use of technology in finance.  Citadel 
(large US hedge fund where I worked in London) used to tell 
people that the firm was almost more of a technology firm than a 
finance firm, and I am not going to comment on that.  Similarly, 
the banks also conceive of themselves as leading in technology.  
The reality (in the general case, and without wishing to comment 
on specific firms) is that things are a mess - fragmented, siloed 
operations with significant costs from legacy approaches that no 
longer fit the new needs of the business.  I agree with Dicebot 
that I haven't found an alternative to D that has high 
productivity, decent speed and power, and suitability for 
multiple problem domains.  Innovation requires the ability to 
iterate rapidly, and to get an answer to questions quickly.  In 
theory when cloud power is free, efficiency doesn't matter, but 
it's not that easy in practice.  Despite the theoretically big 
budget, being able to produce a workable proof of concept quickly 
without needing budget approval has tremendous value.

Thanks for the rest of your post - I agree very much with what 
you write, and it carries weight as you know the problem domain 
well.

> And as I've said before, focusing on a domain means you 
> optimize for it, which inevitably means you become less general.

Quite!  Optimization creates brittleness to things you don't 
optimize for, and nobody can know at this stage what the 
important things might ultimately be.  Central planning has its 
downsides, even if the way that planning occurs through 
polycentric activity in a community seems less rational and pure. 
  Hayek had something to say about this.

> But some of us think general-purpose, native languages are 
> coming back,

Yes.  Now why do you think this is the case?  I tried to 
articulate it as best I could for now, but Ola has all these 
_reasons_ why this isn't the case, which may mean he is right, 
but might not.




More information about the Digitalmars-d mailing list