jugad2 - Vasudev Ram on software innovation

CloudJee from Pramati: Java apps in the cloud

2013-05-25T04:07:00.001+05:30

CloudJee | Mission Critical Java Apps in the cloud

CloudJee is a Java cloud apps company/product from Pramati, an Indian software company.

http://www.cloudjee.com/customers

http://pramati.com/about

http://m.prnewswire.com/news-releases/announcing-cloudjee-the-proven-cloud-platform-for-building-mission-critical-java-saas-applications-208471431.html

I remember Pramati from the early days of Java and J(2)EE. They were the first and maybe the only Indian company to create a J2EE (Java 2 Enterprise Edition - now called JEE) application server. Interesting to see that they have been existing all this while, and have come out with CloudJee.

CloudJee/Pramati also recently acquired WaveMaker, a tool for building Java apps, from VMware, and have integrated it into CloudJee.

http://gigaom.com/2013/05/01/vmware-garage-sale-continues-as-it-offloads-wavemaker-to-pramati/

- Vasudev Ram
dancingbison.com

Test post - ignore.

2013-05-24T19:34:00.001+05:30

By Vasudev Ram

Ignore.
- Vasudev Ram - Dancing Bison Enterprises

Share |

Inside Google X

2013-05-24T04:55:00.001+05:30

Inside Google's Secret Lab - Businessweek

Interesting stuff.

Zato, an open source ESB in Python

2013-05-22T03:50:00.001+05:30

Zato | home page

Zato is an ESB (Enterprise Service Bus) which is written in Python. It is free and open source (LGPL).

http://en.m.wikipedia.org/wiki/Enterprise_service_bus

Tibco was one of the first ESB products and Mule is an open source one.

According to the Zato site:

It supports HTTP, JSON, SOAP, Redis, JMS WebSphere MQ, ZeroMQ, FTP, SQL.

It has a web admin GUI, a CLI and an API.

Documentation and commercial support are available.

I got to know about Zato recently from the main developer, Dariusz Suchojad, who had earlier written to me regarding my blog post about PyMQI:

PyMQI, Python interface to IBM WebSphereMQ (formerly IBM MQSeries):

http://jugad2.blogspot.com/2013/02/pymqi-python-interface-to-ibm.html

Dariusz was a maintainer of PyMQI and also a developer on Spring Python, which is sort of a port of the Java Spring framework to Python.

Zato docs (quite detailed):

https://zato.io/docs/index.html

Part 1 of a basic Zato tutorial:

https://zato.io/docs/tutorial/01.html

I took a look at the tutorial. Broadly, it shows how to install Zato, create a simple Zato service in Python, that talks to PostgreSQL and Redis, and deploy it. Two servers get created, behind a load-balancer, and the service gets hot-deployed to the servers. Then curl is used to access the service. (This tutorial does not create a real client; curl is used to simulate one.)

Zato looks interesting and powerful (and somewhat complex, but that is to be expected for a product like an ESB).

I will check it out more and then report on my findings.

- Vasudev Ram
dancingbison.com

A partial crossword solver in Python

2013-05-21T01:33:00.001+05:30

A Cryptic Crossword Clue Solver ←

Saw this via Twitter.

It is a partial crossword solver, because it only helps solve a particular category of crossword clues - those in which the clue (which is usually a sentence or phrase) contains both a "definition" of the answer as well a hint of some kind that leads to the same answer. This solver tries to compute the answer using both the definition and the hint, and checks whether the results match. Ingenious.

I found it interesting because this is a somewhat difficult problem, and yet the author managed to create a solution (involving NLTK and parsing) that works in many, if not all cases.

Also, long ago, in college days, I had written another kind of partial crossword solver (in BASIC); it was much simpler, using a brute force method - what it did was help solve the kind of crossword clues in which the answer is a permutation of a substring of the characters comprising the clue sentence or phrase. The program would generate and display on the screen, all possible permutations of all possible substrings of the sentence, that were of the same length as the answer. Then you had to view those permutations and guess whether any of them was the right answer, based on the clue.

I wrote the permutation-generation code by hand, but saw recently that the Python itertools module has methods to generate permutations (as well as combinations) from sequences:

http://docs.python.org/2/library/itertools.html

http://en.m.wikipedia.org/wiki/Permutation

http://en.wikipedia.org/wiki/Crossword

- Vasudev Ram
dancingbison.com

Charge your phone in 20 seconds

2013-05-21T00:30:00.001+05:30

Teen's prize-winning invention may charge your phone in 20 seconds - CNN.com

Interesting ... She got a $50,000 award for it at an Intel Science competition for high school students.

HuffShell suggests aliases for your Unix commands

2013-05-20T03:38:00.001+05:30

paulmars/huffshell · GitHub

huffshell is a gem to suggest optimized aliases for your frequently used Unix commands. It looks at your command history to do that.

Seen via this Hacker News post which is interesting too:

What are your top 100 unix commands? (for science) :

https://news.ycombinator.com/item?id=5733426

- Vasudev Ram
dancingbison.com

Python's inspect module is powerful

2013-05-18T00:43:00.001+05:30

27.13. inspect — Inspect live objects — Python v2.7.5 documentation

The inspect module comes as part of the standard library of Python. It allows you to inspect live objects at run time by using its methods.

Here is an example:

import inspect
class Bar( object):
    def foo( self ):
        print "in Bar.foo 4"
        self .a = 1
        self .di = { 'b' : 2, 'c' : 3 }
        self .li = [ 4, 5, 6 ]

bar = Bar()
bar.foo()
for member in inspect.getmembers(bar):
print member

Running the above code gives this as (partial) output:

in Bar.foo 4

('__dict__', {'a': 1, 'li': [4, 5, 6], 'di': {'c': 3, 'b': 2}})
('__doc__', None)
('a', 1)
('di', {'c': 3, 'b': 2})
('foo', bound method Bar.foo of __main__.Bar object at 0x4035bfac)
('li', [4, 5, 6])

This shows both the bound methods and the member variables of the instance bar.

The inspect module can do a lot of other things too. Check the docs for it, linked at the top of this post.

You can also modify and run the above code snippet at this codepad.org URL:

http://codepad.org/dMLmkhQ6

It may not last there for too long, though, since they probably delete pastes to make room for new ones.

- Vasudev Ram
dancingbison.com

Reading WAV file info with Python

2013-05-16T02:04:00.001+05:30

The Python standard library comes with a module called wave.py. It allows you to read and write WAV format audio files.

Here is an example of reading WAV file info with Python:

import wave

wavf = wave.open('horse.wav', 'r')
print wavf.getnchannels()
print wavf.getsampwidth()
print wavf.getframerate()
print wavf.getnframes()
print wavf.getcompname()

The above code prints the number of channels (mono or stereo), the sampling width, the frame rate, the number of frames, and the name of the compression type (if any) of the WAV file horse.wav.

You can also write WAV files, but you need to know what data to write.

Wikipedia entry for WAV:

http://en.m.wikipedia.org/wiki/WAV

- Vasudev Ram
www.dancingbison.com

A blogging break for jugad2

2013-05-09T23:04:00.001+05:30

Dear readers,

I'm taking a blogging break for some days, due to less free time while I sort out some issues.

Will be back.

Inconvenience is regretted.

- Vasudev
dancingbison.com

PDF in a Bottle - creating PDF using xtopdf, ReportLab, Bottle and Python

2013-05-01T03:46:00.002+05:30

By Vasudev Ram

pdf_bottle.py is a program I wrote that allows you to create a PDF file from text, over the web, by entering your text into a form and submitting it.

Here is the program:

# pdf_bottle.py

# Description: Program to generate PDF from text, over the web,
# using xtopdf, ReportLab and the Bottle web framework.
# It can be used to create short, simple PDF e-books.
# Author: Vasudev Ram - http://dancingbison.com
# Copyright 2013 Vasudev Ram 
# Tested with Python 2.7.

# Version: 0.1

# Dependencies:
# xtopdf - https://bitbucket.org/vasudevram/xtopdf
# bottle - http://bottlepy.org
# ReportLab - http://www.reportlab.com/ftp/reportlab-1.21.zip
# Python - http://python.org

from PDFWriter import PDFWriter

from bottle import route, request, run

@route('/edit_book')
def edit_book():
    return '''
    <form action="/save_book" method="post">
    PDF file name: <input type="text" name="pdf_file_name" />

    Header: <input type="text" name="header" />

    Footer: <input type="text" name="footer" />

    Content:
    <textarea name="content" rows="15"   cols="50"></textarea>

    <input type="submit" value="Submit" />

    </form>
'''

@route('/save_book', method='POST')
def save_book():
    try:
        pdf_file_name = request.forms.get('pdf_file_name')
        header = request.forms.get('header')
        footer = request.forms.get('footer')
        content = request.forms.get('content')

        pw = PDFWriter(pdf_file_name)
        pw.setFont("Courier", 12)
        pw.setHeader(header)
        pw.setFooter(footer)

        lines = content.split('\n')
        for line in lines:
            pw.writeLine(line)

        pw.savePage()
        pw.close()
        return "Done"
    except Exception:
        return "Not done"

def main():
    run(host='localhost', port=9999)

if __name__ == "__main__":
    main()

To run it, you need to have Python, the open source version of ReportLab, my xtopdf toolkit and the Bottle Python web framework installed.

Here is a guide to installing and using xtopdf.

For help with installing the other products, consult their respective sites, linked above.

Then run the program with this command:

python pdf_bottle.py

Then, in a browser window, go to localhost:9999

Enter the details - PDF file name, header, footer and text content - in the form, then click Submit.

The PDF file will be generated in the same directory from where you ran the Python program.

This is the first version, and has been tested only a bit. It you find any issues, please mention them in the comments.

Various improvements are possible, including sending the generated PDF to the user's browser, or providing a link to download it (better), and I'll work on some of them over time.

P.S. Excerpt from the Bottle framework site:

[
Bottle is a fast, simple and lightweight WSGI micro web-framework for Python. It is distributed as a single file module and has no dependencies other than the Python Standard Library.

Routing: Requests to function-call mapping with support for clean and dynamic URLs.
Templates: Fast and pythonic built-in template engine and support for mako, jinja2 and cheetah templates.
Utilities: Convenient access to form data, file uploads, cookies, headers and other HTTP-related metadata.
Server: Built-in HTTP development server and support for paste, fapws3, bjoern, Google App Engine, cherrypy or any other WSGI capable HTTP server.
]

- Vasudev Ram - Dancing Bison Enterprises

Share |

Interesting interview of Gagan Biyani of Udemy

2013-04-30T01:10:00.001+05:30

By Vasudev Ram

Gagan Biyani, Udemy co-founder, interviewed on Mixergy.

And here is the story of Eren Bali, one of the other co-founders of Udemy.

From the Wikipedia page about Udemy:

"In February 2010, the Udemy co-founders decided to raise money to fund their idea.[6] They were rejected 30 times" ...

The Udemy Fundraising Story

- Vasudev Ram - Dancing Bison Enterprises

Share |

Hard-hitting presentation on Indian IT industry

2013-04-29T02:35:00.003+05:30

By Vasudev Ram

Seen via a link posted by @sharads on LinkedIn.

Whats Wrong with Indian IT Industry? from Rajesh "Raj" Varma

Got to say I agree with a lot of it. And I think most people in the know, who are honest, would agree too. Not sure how much is being done about it, on an industry-wide level, though I know of efforts here and there.

- Vasudev Ram - Dancing Bison Enterprises

Share |

PySynth, a pure Python music synthesizer

2013-04-28T03:18:00.002+05:30

By Vasudev Ram

PySynth is a music synthesizer library written in Python. (*)

Excerpt from the site:

[ There are three variants: PySynth A is faster, only needs Python itself, and sounds more like a cross between a flute and organ. PySynth B is more complex in sound and needs NumPy. It's supposed to be a little closer to a piano. (No competition for Pianoteq of course, but a reasonable fit for keyboard music.) Finally, PySynth S is more comparable to a guitar, banjo, or harpsichord, depending on note length and pitch.

The current release of the synthesizer is monophonic, i.e. it can only play one note at a time. (Although successive notes can overlap in PySynth B and S, but not A.) However, two output files can be mixed together as in the case of the stereo files below. ]

(*) Interestingly, the Changes section of the above linked PySynth page seems to indicate that PySynth uses both Pyglet and PyAudio, both of which I had blogged about some time ago:

Playing an MP3 with pyglet and Python is easy

PyAudio and PortAudio - like ODBC for sound

Here is PySynth on Github

PySynth supports ABC notation and can generate WAV audio files.

Here is an example tune:

D D C C B A G

and here is a PySynth program to play that tune as a WAV file:

# pysynth-ddccbag.py

import pysynth

test = ( ('d', 4), ('d', 4), ('c', 4), ('c', 4), ('b', 4), ('a', 4), ('g', 3) )
pysynth.make_wav(test, fn = "ddccbag.wav")

For some reason the last note seems to get partially cut off on my PC. Not sure whether that is a bug or a feature :-) or something to do with my hardware. Maybe the latter, since I'm using my laptop's built-in speakers.

Run that program as:

python pysynth-ddccbag.py

Then play the generated WAV file ddccbag.wav in a music player, such as VLC or some other one.

Here are a few better examples of sound synthesis by PySynth - the links are to MP4 files, which you can download and play without needing Python or PySynth:

The Sailor’s Hornpipe — PySynth S

Bach: Jesu, Joy of Man’s Desiring — PySynth S (treble) and B (bass)

Also check my recent post: Play the piano on your computer with Python.

If you want to try it out, note that the program has a couple of bugs, related to the frequencies of notes, since I am not musically trained; I just wrote it for fun and as an experiment. Though some of the post comments gave some background and suggested corrections, they did not seem to be sure, and corrected themselves, so I have not implemented any of those suggestions yet.

Enjoy.

- Vasudev Ram - Dancing Bison Enterprises

Share |

Copperhead: Data Parallel Python from NVIDIA Research

2013-04-26T02:03:00.000+05:30

By Vasudev Ram

News seen via PythonWeekly.

Copperhead, from NVIDIA Research, is a project to bring data parallelism to Python.

From the site (emphasis mine):

[ We define a small functional, data parallel subset of Python, which we then dynamically compile and execute on parallel platforms. Currently, we target NVIDIA GPUs, as well as multicore CPUs through OpenMP and Threading Building Blocks (TBB). ]

Soumds interesting.

Also see my recent related post:

Python-to-GPU compiler coming from Continuum Analytics and Nvidia

About GPUs and GPGPUs.

- Vasudev Ram - Dancing Bison Enterprises

Share |

Plot me some randomness

2013-04-24T01:34:00.001+05:30

By Vasudev Ram

In this recent post about plot.ly,

plot.ly, fast plotting in the browser,

I mentioned that plotting by pasting data worked, but plotting using Python had an issue, and that they said they were fixing it.

The plot.ly people got back to me saying that the Python issue is fixed.

I tried it out again with a couple of simple plotting examples involving random numbers. They worked.

Here are the examples:

from numpy import *
from random import randint

x = range(1, 10)
y = [ random.randint(1, 10) for i in x ]
plot(x, y)

from numpy import *
from random import randint

x = range(1, 20)
y = [ random.randint(1, 10) for i in x ]
plot(x, y)

(The only difference in the above two scripts is 10 vs 20.)

You can try out that code, or your own, in plot.ly.

Go to https://plot.ly/plot.

Click on the "tab" labeled "+" near the top left of the screen, then click "Script" from the drop-down menu.

Enter your script (no need to Save) and then click Run.

If your code has no errors, your plot will appear in a new tab (not a new browser tab, but a new "tab" within the plot.ly browser tab).

The plot may take 2 to 3 seconds to appear.

plot.ly is beta software, but looks useful. They have many more complex examples in their demos.

- Vasudev Ram - Dancing Bison Enterprises

Share |

pypng, pure Python module to encode/decode PNG

2013-04-22T05:28:00.001+05:30

By Vasudev Ram

pypng is a pure Python library to read/write PNG images.

Here are some code examples of using pypng. Note: the actual library is called png.py, not pypng.py. So import that, not this :-)

I tried out a few of the examples, and they worked.

The image above uses the transparency feature of PNG.

- Vasudev Ram - Dancing Bison Enterprises

Share |

Video: Neil Diamond - They're coming to America

2013-04-21T03:55:00.001+05:30

By Vasudev Ram

- Vasudev Ram - Dancing Bison Enterprises

Share |

Samuru, a new (and smarter?) search engine

2013-04-20T20:10:00.001+05:30

By Vasudev Ram - Dancing Bison Enterprises

Samuru is a new search engine from stremor.com

I tried it out a bit and it gave me some interesting results.

About stremor

About Samuru:

Update in a while, I'm on my mobile at present.

Hacker News thread about Samuru is interesting

- Vasudev Ram

With pleasure develop - JetBrains PyCharm blog

2013-04-20T08:52:00.001+05:30

PyCharm 2.7.2 released | JetBrains PyCharm Blog

you.get(it) ? :-)

sqlite3dbm, an SQLite-backed dbm module

2013-04-19T02:51:00.001+05:30

By Vasudev Ram

Saw this today. It seems to be on the Github account of Yelp.com.

They created it as a tool to help with Hadoop work on Amazon EMR (Elastic Map Reduce).

sqlite3dbm provides a SQLite-backed dictionary conforming to the dbm interface, along with a shelve class that wraps the dict and provides serialization for it.

I tried it out, and it worked as advertised.

How to use sqlite3dbm:

Import the module, use its open() method to create an SQLite database, getting back a handle to it, let's call it "db", then use Python dict syntax on db to store data.

Then, either in the same or another program later, you can again fetch and/or modify that data, with dict syntax.

Interesting idea. dbm modules, which implement key-value stores, are less powerful than relational databases (SQL), and were probably developed earlier (think ISAM, etc.), so it looks a bit backwards to implement a dbm-type store on top of SQLite. But the sqlite3dbm project page gives at least some justification for that:

[ This module was born to provide random-access extra data for Hadoop jobs on Amazon’s Elastic Map Reduce (EMR) cluster. We used to use bsddb for this because of its dead-simple dict interface. Unfortunately, bsddb is deprecated for removal from the standard library and also has inter-version compatability problems that make it not work on EMR. sqlite3 is the obvious alternative for a persistent store, but its powerful SQL interface can be too complex when you just want a dict. Thus, sqlite3dbm was born to provide a simple dictionary API on top of the ubiquitous and easily available sqlite3.

This module requres no setup or configuration once installed. Its goal is a stupid-simple solution whenever a persistent dictionary is desired. ]

- Vasudev Ram - Dancing Bison Enterprises

Share |

Now, Google+ comments on your Blogger blog

2013-04-19T00:45:00.000+05:30

By Vasudev Ram

Just saw the news via an email from Google.

Bringing Google+ Comments to Blogger

Looks like an interesting idea.

Some of the benefits mentioned for this feature:

- View your blog and Google+ comments, all in one place

- Help readers comment and connect with their circles

I enabled the feature for my blog - this one, jugad2.blogspot.com .

Let's see how it works out. There may be a few issues initially, but over time those should get fixed.

- Vasudev Ram - Dancing Bison Enterprises

Share |

Common DB developer mistake? Using natural instead of surrogate keys

2013-04-17T03:13:00.002+05:30

By Vasudev Ram

Just saw this StackOverflow article:

Database development mistakes made by application developers [closed]

via this Twitter conversation between me and @coding_insights.

The article had this as one of the answers:

"3. Using natural rather than surrogate (technical) primary keys"

followed by an explanation of why that person thinks so.

I tend to agree with that comment, based on my own experience working on many database-backed projects.

Though it seems counter-intuitive, using surrogate keys seems to be a better and more robust approach.

The commenter gave many reasons for using surrogate keys instead of natural keys, (and as I said, I've experienced at least some of those reasons), so I won't repeat them here; go take a look at that article.

- Vasudev Ram - Dancing Bison Enterprises

Share |

"Free is only a tactic, though, not a business model." - GigaOm

2013-04-17T01:39:00.000+05:30

By Vasudev Ram

I saw an interesting post by Fred Wilson:

Tenacity And Persistence Pays Off.

Fred's post also had a comment by Philip Sugar, to which I replied, asking for more info, and he replied, with a link in his comment to this post: The Penny Gap. That post had an interesting quote from an article on GigaOm:

[ To be fair to these VCs, they’re not advocating doing everything without pay. They’re suggesting free as a tactic towards getting paid in other ways: through advertising, or by premium services (as in a freemium model), or maybe even through being acquired by a company with a large wallet. Free is only a tactic, though, not a business model.

Conflating the two misleads web application developers into thinking they don’t need to do the hard work of figuring out what’s really of value to users before they build and launch their online service. ]

The GigaOm post is here: Free: a Tactic, not a Business Model

- Vasudev Ram - Dancing Bison Enterprises

Share |

The Python Brochure Project - preview available

2013-04-15T04:17:00.003+05:30

By Vasudev Ram

The Python Brochure Project, an initiative by the Python Software Foundation, has released a preview of the brochure. The purpose of the Python Brochure is to make the Python language and its benefits and uses more widely known, among other things.

The contents of the Python Brochure are pretty interesting, IMO, and show how widely and diversely Python is being used worldwide by companies, governments and other organizations.

Python may even go to Mars in a few years! Not kidding - see the brochure.

Excerpt from the site:

[ The brochure showcases Python projects with great impact, fascinating aspect and interesting twist from all application spaces covering use cases in business, science, industry, education, media, government, public sector and charity. ]

You can get the Python Brochure preview (as a PDF file). Scroll a bit down the page if you need to (depending on your monitor size), to see the download link.

I just downloaded it and scanned through it, and it looks quite good.

A minor nitpick is that I wish that they would have made it a single-column PDF, but that is my personal preference.

Update: Just realized that the two-column format is more suitable for printing (if the columns come on opposing pages), and brochures are printed ... Still wish there was a single-column version for computer viewing, though.

- Vasudev Ram - Dancing Bison Enterprises

Share |