The Third Bit

I remember when Java books were selling like hotcakes. How ironic to see DHH and the railers point to book sales as evidence of long-term shift.

By Bookman on Apr 22, 2006

I agree with Bookman’s comment above. It seems to me like Rails is being overhyped. Sure, it’s a pity that Python is not getting more hype, but Python is definitely not going away, and getting more attention on dynamic languages from the big public is definitely a Good Thing (TM).

By Manuzhai on Apr 22, 2006

How many new major python books have been published this year compared to new major ruby books?

By Tim Parkin on Apr 22, 2006

I remember when Java books were selling like hotcakes, too. Three years later, most enterprise web applications were being developed in Java…

By Greg Wilson on Apr 22, 2006

> How do you read these statistics? Python book sales look up 43% to me.

fundamental point for me : the majority (and most best sales) of the Python books are not in O’Reilly catalog…

By Sébastien Douche on Apr 23, 2006

Hm, this blog entry appears to have changed and my comment is gone. It’d be nice to have some indication of updates that are major semantic shifts.

Anyway:

* As far as I understand, these are figures for beyond just the O’Reilly catalog, so they are interesting beyond just what O’Reilly sells.

* Even if a unified response to Ruby on Rails had been performed by the Python world, it would’ve been too late already. Rails was already in hype-mode, and nothing the Python world can do in the web space could be engineered to match that. Arguing for a unified response now is a) a hopeless cause; it just won’t happen b) even if it would happen it wouldn’t help popularity in a large way.

* It’s easier for Ruby to grow as they came from far below. It’s impressive how they’ve been able to catch up with Python in a short time - what hype does for you.

* It’s hard to engineer *any* platform (besides the web, which is already taken hype-wise) that would receive the same amount of hype for Python. I’m not sure it matters. There is plenty of good software written in Python. Python seems to be on a curve of steadily growing and gaining more acceptance, and has been on this curve for years, and Ruby’s popularity only helps this transition in the industry.

* Python’s maturity and broad support in a wide range of software areas is a strength of Python; you can’t hype it much, but it is a good ingredient for continued growth.

By Martijn Faassen on Apr 24, 2006

Python 3000 will drive developers away from the language, as it breaks all their software and forces them to port to essentially a new language while maintaining two separate branches of their project until the installed base of Python 2.x is completely replaced.

Why not just take the opportunity to switch to a better language with a better runtime and better backwards compatibility?

By Jean-Paul on Apr 24, 2006

Regarding hype, the argument that there’s already plenty of good software written in Python, and the like: I heard a lot of people saying the same things about Objective C back in the day, and about Linda, and several other systems that were (in my opinion) technically superior to the ones that displaced them. I think the message we send by not responding to RoR in a coherent fashion is that we don’t care; I don’t think that’s a message we ought to send.

By Greg Wilson on Apr 24, 2006

The wxPython book that was recently published has been on my wish list so long that now that it is out I am a little underwelmed. I think I will get it any way. I hope a book on Django or TurboGears is out before the next paradigm shift comes to the development world.

By roger on Apr 24, 2006

Lies, da*ed lies and statistics. Ruby is newer. I do not know the exact numbers but it is much easier to grow seven times from 2 to 14 books than to grow 50% from 200 to 300 books. And if they catch up with Python more power to them. I like Ruby. If they used indentation for scope and had a extensive standard library I would not mind switching.

As far as Python-3000 is concerned many people have wrong expectations. If the changes from version 1.x to version 2.x are any indication, version 3.x will change little for most people (and that is good). Most changes will happen “under the hood”. The “user interface” of the language will stay the same. So other than newer editions of reference books with phrases like “covers up to 3.0″ on the cover, the impact of P3k should have little impact on those curves.

By nestr on Apr 24, 2006

nestr: The version number bump from 1.x to 2.0 was more about licensing than anything else. The transition from 2.x to 3.0 will be quite a different one, and it’s been stated that it will break some things.

Having said that, Jean-Paul is just peddling FUD.

By Richard Jones on Apr 24, 2006

From the PEP,

“Python 3000 will break backwards compatibility. There is no requirement that Python 2.9 code will run unmodified on Python 3.0.”

I’m just repeating the official plans. Perhaps it’s FUD, but it’s PSF FUD.

By Jean-Paul on Apr 25, 2006

this argument is ridiculous in so many ways. first of all, percentage increase of book sales is a dubious figure of popularity (or lack thereof). If ROR sold 10 books last year, and 70 this year (7 fold increase), and python sold 1000 books last year, and 1430 books this year, python is still ‘more popular’. as for the increase, of course ROR is selling more books (compared to themselves), its brand new, python has been around for a while now, people arent going to re-buy ‘learning python’ every year. think before you post.

By Dan F on Apr 25, 2006

It’s good to see Ruby doing well, even though I use Python. The main reason for that is that I’m not interested in web programming, and Python has far more breadth in it’s library and framework support. As a Pythonista, this doesn’t concern me at all.

Python book sales have been growing at about the same rate for many years now, and the appearance of Ruby on the scene does not appear to have had the effect of slowing Python’s rate of growth, despite undoubtedly drawin some people away from Python. I think it’s fair to assume that rails is actualy growing the market for these sorts of languages more than it is eating Python’s lunch.

Simon Hibbs

By Simon Hibbs on Apr 25, 2006

You know what though? Outside of RoR I don’t hear anything at all about Ruby. I hear about Python all over the place and I don’t even use it (I use Tcl or Perl). It is easy for Ruby books to skyrocket since nobody was buying them before.

By Robert on Apr 25, 2006

Here is Guido’s take on Python 3000:

http://www.python.org/doc/essays/ppt/accu2006/Py3kACCU.ppt

By Robert on Apr 25, 2006

Greg; I think comparing Python with Objective C or Linda concerning library support is way off. Python has an enormous breadth of libraries far beyond web programming; I suspect only Perl, Java and C may have a wider support, but the competition is close as Python has traditionally been great at exposing C libraries, where Perl and Java are somewhat weaker.

By Martijn Faassen on Apr 26, 2006

“I remember when Java books were selling like hotcakes, too. Three years later, most enterprise web applications were being developed in Java…”

Take heart, at least one person is not ignoring your substantial point.

By AvoidingAnotherRubyEvaluation on May 6, 2006

I suspect all current open source languages must ultimately accept that none of them will be the programming lingua franca of programming by the time literacy skills include the ability to manipulate information by computer. Python is a step along one path, Perl is another, Ruby a third and so on. Choice is a great thing: just as not all people speak the same language, so they need not program in the same language. All should “play nice”. Is there a Ruby for .NET/Mono? How about a domain-specific language toolkit? Maybe the new AST compilation techniques will make it easier to use the Python interpreter as a DSL toolkit. For those who couldn’t see Guido’s OSCON presentation there’s http://video.google.com/videoplay?docid=-6459339159268485356

By Steve on Aug 19, 2006