Why I'm hesitating to switch to D

[Nick Sabalausky]

"James Fisher" <jameshfisher at gmail.com> wrote in message 
news:mailman.1259.1309290936.14074.digitalmars-d at puremagic.com...
>
> *# Package management*
>
> This one's *really* important.  The front page of
> digitalmars.org/d/describes it as having "... the programmer
> productivity of modern languages
> like Ruby and Python".  The language, maybe, but this statement is
> absolutely not true while there is no associated package manager.
>

Yea, that's why we're working on it. There's been a few rather big 
discussions on it very recently and a number of people have been working on 
actual code.

> I've followed the "two standard libraries" debacle with some confusion 
> (I'm
> still not clear what's being done about it).

For D2, the std lib is Phobos. Period. For D1, the de-facto standard is 
Tango (which was created because, at the time, Phobos wasn't very good and 
didn't have much manpower). This hasn't been an issue in ages.

> The closest thing to a central repository seems to be
> http://www.dsource.org/.  But this is all wrong.  People don't want to use 
> a
> language-specific site to host their project under SVN.  They want to use
> GitHub/Bitbucket/Google code/etc. Dsource.org should just maintain a list
> of packages for an installer -- say, like http://search.npmjs.org/.

DSource predates the widespread popularity of GitHub/Bitbucket/Google 
code/etc. I think it actually predates Google Code, period. Also, I'd 
caution against blanket stantements about "what people want." I, for one, 
like DSource. There are some things about it I like *better* than the 
others. And I'm happy using DSource for a number of things.

> That D
> projects are hosted on a D-exclusive site tells me that D has a closed
> community.
>

What this tells me is that you enjoy making assumptions and jumping to 
conclusions. ;)

> *# Community fragmentation*
>
> Where is the D community?  I see:
>
> http://www.digitalmars.com/d/ -- seems to be official
> http://www.d-programming-language.org/ -- also seems to be official!
>

You caught us right in the middle of the transition from 
http://www.digitalmars.com/d/ to http://www.d-programming-language.org/ 
That's why it seems confusing.

>
> http://www.dprogramming.com/

I don't see what's so wrong with that. It never claims to be anything 
official. I'm sure other languages have user-created sites, too. Big deal.

> http://www.dsource.org/
> etc.
>

That's for project hosting. Not the same.

> *# It's unsearchable!*
>
> This one is really trivial. There's an important reason that other
> languages have squiffy names: searchability.  Googling for "d <query>" is
> useless, and "d language <query>" is still awful.  Other languages that
> suffer from the same affliction have the convention of appending "lang" as 
> a
> suffix -- "golang", for example -- and this works well.  It seems that
> "dlang" has not caught on.  I see there's a site at http://dlang.org/ 
> (yes,
> yet another one!).  Whois says it's owned by http://oscarbrynolf.com/. 
> The
> (seemingly recent?) move to GitHub and new website would have been a 
> chance
> to get this right.  Prepending "d programming language" to every search I
> make is still absolutely horrible.
>

"d programming" works fine for me. And you're right, it is only a trivial 
matter.

>
> *# No marketing or brand awareness*
>
> OK, I can live with this.  But make no mistake: it *does* seriously cut 
> down
> on the people migrating to it.  Take a look at the "free images" offered 
> at
> http://digitalmars.com/d/dlinks.html -- this is a marketers nightmare! 
> That
> list *literally* makes me shudder.  I'm not saying that D requires another
> generic Web 2.0 HTML5 look with gradients and rounded corners that one 
> sees
> on the latest fashionable projects.  I *am* saying that it needs something
> consistent and clean, and currently it, seemingly willfully, has neither.
>

We've been doing a lot of marketing and promoting. As for the images, that 
seems to be an incredibly petty "issue".

>
> The point I want to get across here is: the problem with the D programming
> language *is not that there are problems with the D programming language.*
> The language and compiler (from what I know) have been world-class for 
> some
> time.  I've seen a few conversations where people using D get quite
> indignant that people are interested in Go, Rust, etc.  D is not getting 
> new
> users because it doesn't look like it *wants* new users: it dresses
> sloppily, it doesn't make itself findable easily, and it doesn't have
> infrastructure for new users to share and benefit from sharing.
>

The main problem with D is that there's too many people with their lists 
about what's wrong with D, and not enough willing to actually jump in and 
help out. I'm sorry if I come across a bit harsh, but we seriously get 
another one of these "What's wrong with D" lists every few weeks (and each 
one seems more out-of-date than the last) and hardly any of the people 
writing them have ever actually made any real contributions besides just 
complaints. If you're going to help out, then fantastic, and welcome aboard! 
But if not, then please understand that we already have more volunteer 
supervisors than we need, so that sort of thing gets very frustrating for 
those of us dedicating our own time and effort for free to actually 
accomplish all the goals.