On Feb 19, 2006, at 11:17, Glenn Smith wrote:

> David - we can start a flame war, all rather pointless and doesn't do 
> you or
> I or anybody else any favours.  Or we can simply agree to shake hands 
> and
> have a "virtual" beer.

Or you could ignore him. *I* certainly found comments like "It takes at 
most five minutes over a slow connection of clicking around to figure 
that out on the ruby-lang website, hopefully much less after
the revamp, and at most 30 seconds for anyone with mediocre google 
skills to get the basic points right." to be not only annoyingly 
condescending, but quite opposite of my experience, which seems to 
involve spending a lot of time finding documentation misplaced, 
dreadfully obscure, or just plain missing, and discovering new and 
different ways for those oh-so-easy installers to fail.

I think he's probably wrong in believing that "You don't represent any 
significant majority of Ruby users on any platform in my opinion." You 
certainly are representing my opinions pretty accurately. On the other 
hand, if he's right, then Ruby might well be going down the 
incoherent-and-eventually-irrelevant path. I see there being a danger 
of that, but I don't think it's at all inevitable, yet, thank goodness.

> I think you've taken by initial post on this thread to be a real moan, 
> when
> what I was really trying to do was simply put forward one user's 
> perspective
> (ie. mine) of how Ruby was being presented to potential users.  For 
> what
> it's worth, I think your point "Programming languages and tools are not
> end-user software." is quite wrong.

I agree. I have and use Ruby because I want to build programs and tools 
for myself...in Ruby. The time I spend having to poop around with 
recompiling Ruby, reinstalling Ruby, re-downloading source for Ruby, 
debugging Ruby's installers, is wasted time. I wasted something like 
three or four hours trying to get readline support working with irb, 
IIRC.

The only installer tool I've used for Ruby or Ruby-related material 
that has NOT errored out or installed something incorrectly 
is...Apple's standard OSX installer. Unfortunately, as far as I can 
tell, right now, only Ruby 1.8.2 is available in that installer. OTOH, 
it includes Readline, Gems, Rails, TclTk, and the RI documentation (and 
not just all using the same installer, but in the same package! Woo 
hoo!), so I may just revert backwards to it, since Gem/Rails has not 
yet installed correctly in three tries (one of those tries was on a 
brand-new freshly installed OSX 10.3, no less), I have to install Ruby 
and Readline by downloading and compiling source (the instructions for 
using CVS for that failed on the second command), Rails with Gem, and 
and I still have no idea why RDoc  has completely failed to document 
any of the core material. I've read over the stuff that came with the 
source code, and done what it told me, and still, trying to get 
information on, say, "Array" just gets me some useless chit-chat about 
something called YARV.

I'm sure if I spent more time not trying to actually get work done, I 
could get that fixed, by asking questions here. But I have a copy of 
Pickaxe, so I just use that. If I have to spend time working on my car 
instead of driving to work, then my car isn't very good. If I have to 
spend time working on my computer instead of using it to get work done, 
then my computer's not very good. And if I have to spend time sending 
messages to Ruby-Talk trying to find out how to get Ruby to work 
instead of programming, then Ruby's not very good.

I'll tell friends who are programmers about Ruby, but I haven't yet 
recommended it to anybody. It's too unstable, too undocumented, too 
hard to use. It's too young. I've completely shelved any Rails 
development because hours of searching, and an inquiry posted here, 
have revealed the absence of critical documentation for database 
design. I just don't have the time to join a whole new mailing list and 
see if I can coax somebody to document exactly what *all* of Rails' 
assumptions about the underlying database are. I've already been much 
the same thing with RubyCocoa, and now THAT is actually working as 
expected, and I'm getting things out that work, so I'm just going to 
stick with the system I've got that's running, and put off projects 
that want Rails as long as possible, and hope it's more mature when I 
come back around to looking at it.

I think the biggest difference between Ruby and Groovy is that there's 
somebody who "owns" Ruby and is still actively (and effectively!) 
involved: Matz. I do wish Ruby were a bit more specific. Parentheses 
are sometimes but not always optional, and the like. (shudder) But it 
seems to be headed in the right direction, and hopefully it won't be 
too long before the amount of time one must spend working ON Ruby 
instead of working WITH Ruby drops to nearly nothing.

I also agree with you, Glenn, in that I don't think these problems are 
fundamental problems with Ruby itself, or its tools, or its community. 
The problem is with people who don't recognize or admit that these ARE 
problems, and would try to deny or excuse them. These issues can be, 
and are being, resolved, but only as long as they're recognized as 
issues that NEED to be resolved.

Pointing out weaknesses in something to its fan base doesn't always 
make you friends, but that doesn't mean it isn't worth doing. Thanks, 
Glenn; hopefully your observations will help Ruby grow even stronger.