Python development time is faster.

**Fredrik Lundh** · Nov 13 '06, 01:35 PM

Re: Python development time is faster.

Chris Brat wrote:

I've seen a few posts, columns and articles which state that one of the
advantages of Python is that code can be developed x times faster than
languages such as <<Insert popular language name here>>.
>
Does anyone have any comments on that statement from personal
experience?

have you tried writing something in Python, or are you just asking around to
see if it's worth the effort to download it and play with it a little?

</F>

**Paddy** · Nov 13 '06, 01:55 PM

Re: Python development time is faster.

Chris Brat wrote:

I've seen a few posts, columns and articles which state that one of the
advantages of Python is that code can be developed x times faster than
languages such as <<Insert popular language name here>>.
>
Does anyone have any comments on that statement from personal
experience?
How is this comparison measured?
>
>
Thanks
Chris

Here's a start:

http://www.tcl.tk/doc/scripting.html

Go Googling. There's a paper out their that compares error rates and
time to program, lines of code, for several languages and is often
cited in defence of scripting languages. (Scripting languages have
however moved on and now like to be called dynamic languages).

- Pad.

**Chris Brat** · Nov 13 '06, 02:05 PM

Re: Python development time is faster.

I work full time with Java, but downloaded python about a year ago and
started playing.

I've used it quite a few times in my working environment.

**Harry George** · Nov 13 '06, 03:05 PM

Re: Python development time is faster.

"Chris Brat" <chrisBrat@gmai l.comwrites:

I've seen a few posts, columns and articles which state that one of the
advantages of Python is that code can be developed x times faster than
languages such as <<Insert popular language name here>>.
>
Does anyone have any comments on that statement from personal
experience?
How is this comparison measured?
>
>
Thanks
Chris
>

Personal experience takes two forms.

1. Broad experience with languages, such that doing a single project
in a new language gives a sense of the relative power and
ease-of-use. Everyone I know who is a strong Python supporter took
that route. There was an Ahh-Ha experience part way into the first
project. This includes folks who could charitably be called
curmudgeons, and people who are truely fluent in, say, C++ or Lisp.

For these people the main success factor is that "it just works".
You spend time on new functionality, or experimenting with
alternative algorithms, not on debugging. Of course, we work in a
regression-test-driven world, so we don't pile up a lot of untested
code and then hope for the best. Python facilitates that
test-early-test-often approach with its modularity and fast
edit-run cycle.

2. Write the same thing in 2 or more languages. Due to machine
migrations and project redirections I have done that with
perl-and-python, java-and-python, modula3-and-python,
lisp-and-python. In all cases, python was the second language, so
there is some learning curve to be adjusted for (i.e., I understood
the semantics better). However, since I've done some
perl-and-perl, and lisp-and-lisp, I can maybe make that adjustment.

The result was that python was relatively faster-to-develop. I
can't give a specific speedup factor, but I sure can say Python is
now my default language. The success factors were:

a) Once you get the hang of the language (a weekend?), you can
write without reference to the manuals. Or if you do reference, it
is a quick lookup. No struggling to figure out how to code
something. Or to decypher what a line of code actually does.

b) The project doesn't bog down as you add features. The language
can accomodate new paradigms, patterns, or major functionality. If
you do need to refactor, that is easy too.

c) Peer code reviews are easy -- both you and the reviewers can
understand the code's intent at a glance.

--
Harry George
PLM Engineering Architecture

**Steven Bethard** · Nov 13 '06, 03:05 PM

Re: Python development time is faster.

Chris Brat wrote:

I've seen a few posts, columns and articles which state that one of the
advantages of Python is that code can be developed x times faster than
languages such as <<Insert popular language name here>>.
>
Does anyone have any comments on that statement from personal
experience?

I had to work at a laboratory a few years ago which used Java
exclusively. I was coming from several years as a graduate student
using Python almost exclusively for my own work. (But I used to teach
introductory Java classes at my previous university, so I had plenty of
Java experience.)

My own work and the work that I did for the lab were quite similar,
mainly focused on training machine learning models on natural language
processing tasks. I estimated that the Java code took me about 5x as
long. Part of this is the verbosity of Java, e.g. where you have to
write an anonymous inner class instead of using a function or a class
object directly. But probably a larger part of this was using the Java
libraries, which tend to be way over-engineered, and more complicated to
use than they need to be.

A simple example from document indexing. Using Java Lucene to index
some documents, you'd write code something like::

Analyzer analyzer = new StandardAnalyze r()
IndexWriter writer = new IndexWriter(sto re_dir, analyzer, true)
for (Value value: values) {
Document document = Document()
Field title = new Field("title", value.title,
Field.Store.YES ,
Field.Index.TOK ENIZED)
Field text = new Field("text", value.text,
Field.Store.YES ,
Field.Index.TOK ENIZED)
document.add(ti tle)
document.add(te xt)
}

Why is this code so verbose? Because the Lucene Java APIs don't like
useful defaults. So for example, even though StandardAnalyze r is
supposedly *Standard*, there's no IndexWriter constructor that includes
it automatically. Similarly, if you create a Field with a string name
and value (as above), you must specify both a Field.Store and a
Field.Index - there's no way to let them default to something reasonable.

Compare this to Python code. Unfortunately, PyLucene wraps the Lucene
APIs pretty directly, but I've wrapped PyLucene with my own wrapper that
adds useful defaults (and takes advantages of things like Python's
**kwargs). Here's what the same code looks like with my Python wrapper
to Lucene::

writer = IndexWriter(sto re_dir)
for value in values:
document = Document(title= value.title, text=value.text )
writer.addDocum ent(document)
writer.close()

Gee, and I wonder why it took me so much longer to write things in Java. ;-)

STeVe

**George Sakkis** · Nov 13 '06, 03:35 PM

Re: Python development time is faster.

Steven Bethard wrote:

A simple example from document indexing. Using Java Lucene to index
some documents, you'd write code something like::
>
Analyzer analyzer = new StandardAnalyze r()
IndexWriter writer = new IndexWriter(sto re_dir, analyzer, true)
for (Value value: values) {
Document document = Document()
Field title = new Field("title", value.title,
Field.Store.YES ,
Field.Index.TOK ENIZED)
Field text = new Field("text", value.text,
Field.Store.YES ,
Field.Index.TOK ENIZED)
document.add(ti tle)
document.add(te xt)
}
>
Why is this code so verbose? Because the Lucene Java APIs don't like
useful defaults. So for example, even though StandardAnalyze r is
supposedly *Standard*, there's no IndexWriter constructor that includes
it automatically. Similarly, if you create a Field with a string name
and value (as above), you must specify both a Field.Store and a
Field.Index - there's no way to let them default to something reasonable.
>
Compare this to Python code. Unfortunately, PyLucene wraps the Lucene
APIs pretty directly, but I've wrapped PyLucene with my own wrapper that
adds useful defaults (and takes advantages of things like Python's
**kwargs). Here's what the same code looks like with my Python wrapper
to Lucene::
>
writer = IndexWriter(sto re_dir)
for value in values:
document = Document(title= value.title, text=value.text )out two
writer.addDocum ent(document)
writer.close()
>
Gee, and I wonder why it took me so much longer to write things in Java. ;-)

Oh, the memories... I went down the same road about two years ago,
though I didn't know about PyLucene at the time and wrapped in jython
the parts of Lucene I used... never bothered to deal with java's
verbosity after that. It's a pity that jython resembles abandon-ware
these days, when jRuby showed up pretty recently and is gaining in
penetration with the java crowd. It will be a non-trivial loss for
python if it is left behind in the JVM world (at least if the latter is
not swallowed by the .NET dark forces, which doesn't seem to happen any
time soon ;-).

George

**Diez B. Roggisch** · Nov 13 '06, 03:45 PM

Re: Python development time is faster.

Oh, the memories... I went down the same road about two years ago,

though I didn't know about PyLucene at the time and wrapped in jython
the parts of Lucene I used... never bothered to deal with java's
verbosity after that. It's a pity that jython resembles abandon-ware
these days, when jRuby showed up pretty recently and is gaining in
penetration with the java crowd. It will be a non-trivial loss for
python if it is left behind in the JVM world (at least if the latter is
not swallowed by the .NET dark forces, which doesn't seem to happen any
time soon ;-).

I wouldn't consider jython abandonware. It is under active development, and
I'm using a 2.2 alpha successful for quite a while now - which usually
serves my needs.

The problem is/was that new-style classes were a major hurdle to take, and
this now seems to be conquered. So lets hope (or contribute code....:P)
that jython will see a 2.4 version ASAP.

diez

**Egon Frerich** · Nov 13 '06, 04:55 PM

Re: Python development time is faster.

Chris Brat schrieb:

I've seen a few posts, columns and articles which state that one of the
advantages of Python is that code can be developed x times faster than
languages such as <<Insert popular language name here>>.
>
Does anyone have any comments on that statement from personal
experience?
How is this comparison measured?
>
>
Thanks
Chris
>
>

Have a lookl at

http://wwwipd.ira.uka.de/EIR/otherwork/index.html

**flit** · Nov 13 '06, 04:55 PM

Re: Python development time is faster.

One thing I really like, is making "prototypes " on python.
Just to test some algorithm or procedure.
It is very fast and easy to debug.
So after, I make it in c++ (but not too much often, I leave it in
python today.)

Chris Brat wrote:

I've seen a few posts, columns and articles which state that one of the
advantages of Python is that code can be developed x times faster than
languages such as <<Insert popular language name here>>.
>
Does anyone have any comments on that statement from personal
experience?
How is this comparison measured?
>
>
Thanks
Chris

**Stefan Behnel** · Nov 13 '06, 05:15 PM

jython's future (was: Python development time is faster.)

Diez B. Roggisch wrote

I wouldn't consider jython abandonware. It is under active development, and
I'm using a 2.2 alpha successful for quite a while now - which usually
serves my needs.
>
The problem is/was that new-style classes were a major hurdle to take, and
this now seems to be conquered. So lets hope (or contribute code....:P)
that jython will see a 2.4 version ASAP.

The jython project currently deploys a pretty cathedral-ish development style,
pretty much hidden from broader audiences. FWIW, it seems to be stuck in
trying to get out a 2.2 compliant release first, then to provide a 2.3
compliant version (at least, there's a branch for it). Guess it's a question
of human resources, but personally, I wouldn't care too much about
compatibility and just let people go and implement whatever feature they want
to be in on the way towards something as close to 2.5 (or maybe 2.4, why not)
as possible. Honestly, how many important Python modules do still run on 2.2?
Most switched to at least 2.3 quite a while ago.

After all, it's the compatible *software* that makes the difference, not the
platform. It's easier to say: "hey, this feature is lacking to make my program
run" than: "let's see, what else is there in that outdated specification to
implement?". And it's a lot more encouraging, too.

Just my little rant on this one ...

Stefan

**Carsten Haese** · Nov 13 '06, 05:35 PM

Re: jython's future (was: Python development time is faster.)

On Mon, 2006-11-13 at 18:34 +0100, Stefan Behnel wrote:

Honestly, how many important Python modules do still run on 2.2?

InformixDB still compiles on 2.2 except when I accidentally temporarily
break backwards compatibility.

Of course it's a matter of opinion whether it qualifies as an important
module. ;-)

-Carsten

**Steve Holden** · Nov 13 '06, 05:55 PM

Re: jython's future

Carsten Haese wrote:

On Mon, 2006-11-13 at 18:34 +0100, Stefan Behnel wrote:

>Honestly, how many important Python modules do still run on 2.2?

>
InformixDB still compiles on 2.2 except when I accidentally temporarily
break backwards compatibility.
>
Of course it's a matter of opinion whether it qualifies as an important
module. ;-)
>

I think you will find that most of Fredrik Lundh's stuff will work on
1.5.2. He sets a fine example to us all.

regards
Steve
--
Steve Holden +44 150 684 7255 +1 800 494 3119
Holden Web LLC/Ltd http://www.holdenweb.com
Skype: holdenweb http://holdenweb.blogspot.com
Recent Ramblings http://del.icio.us/steve.holden

**Åukasz Langa** · Nov 13 '06, 06:15 PM

Re: jython's future

Steve Holden:

I think you will find that most of Fredrik Lundh's stuff will work on
1.5.2. He sets a fine example to us all.
>

1.5.2? Come on, it's 7 years old. Back then PHP was still at 3.0, Java
was at 1.2 (and compiling Hello World took over 5 minutes), Google was
nothing more than a search engine (still Beta at the time!).

What do we have all these cool new language features for if it's
considered bad behaviour to actually use them?

OK, I see the point in being compatible with Python 2.1 (well, maybe 2.2
with new-style classes and __descriptors__ ), just like it's still useful
to be Java 1.4 compatible and PHP 4 compatible.

But Python 1.5.2? From the official site: "*Do yourself a favor* and get
a more recent version!" ;-) It doesn't even have string methods. Today
Python 1.5.2 is being implemented on one-chip GSM modules (
http://www.telit.co.it/product.asp?productId=96 ). And yes, it actually
works quite well there :-)

Regards,
Åukasz Langa

**Steve Holden** · Nov 13 '06, 08:05 PM

Re: jython's future

Åukasz Langa wrote:

Steve Holden:

>I think you will find that most of Fredrik Lundh's stuff will work on
>1.5.2. He sets a fine example to us all.
>>

>
1.5.2? Come on, it's 7 years old. Back then PHP was still at 3.0, Java
was at 1.2 (and compiling Hello World took over 5 minutes), Google was
nothing more than a search engine (still Beta at the time!).
>
What do we have all these cool new language features for if it's
considered bad behaviour to actually use them?
>
OK, I see the point in being compatible with Python 2.1 (well, maybe 2.2
with new-style classes and __descriptors__ ), just like it's still useful
to be Java 1.4 compatible and PHP 4 compatible.
>
But Python 1.5.2? From the official site: "*Do yourself a favor* and get
a more recent version!" ;-) It doesn't even have string methods. Today
Python 1.5.2 is being implemented on one-chip GSM modules (
http://www.telit.co.it/product.asp?productId=96 ). And yes, it actually
works quite well there :-)
>

And thanks to Fredrik there's a chance you'll be able to use PIL there.

regards
Steve
--
Steve Holden +44 150 684 7255 +1 800 494 3119
Holden Web LLC/Ltd http://www.holdenweb.com
Skype: holdenweb http://holdenweb.blogspot.com
Recent Ramblings http://del.icio.us/steve.holden

Python development time is faster.

Python development time is faster.

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment