• Mars: A MapReduce Framework on Graphics Processors: http://www.cse.ust.hk/gpuqp/Mars.html
• NVIDIA’s CUDA support: http://developer.nvidia.com/what-cuda
• NVIDIA’s support for OpenCL: http://developer.nvidia.com/opencl
• WebCL: http://www.khronos.org/webcl/
• OpenCL podcasts at MacResearch: http://macresearch.org/opencl. This rather interesting video from MacResearch made me spend some time on their podcasts. I really liked them
  
• An IBM research focusing on OpenCL: https://researcher.ibm.com/researcher/view_page.php?id=1835. I read a couple of their publications, but wasn’t able to find a public link. Rather interesting read.
• Finally, looking for a Java based API for OpenCL, I came across JOCL

The promise of a GPU powered speed-up in processing is a rather seductive one. There are ofcourse a bunch of disadvantages going with the GPU (refer to one of the Khronos, or MacResearch links for details).

This is an area I will be looking at in some detail going forward.

Some experiments later, we posted a forum topic, and the sMash development team like the idea.

With the latest release of sMash this support is available for all sMash developers and users. Karthik, and Thiru did a great job of making this happen.

Smarter Planet Icons

A bunch of things that I have been noticing about IBM for sometime now, but haven’t gotten around to actually writing it down.

The Smarter planet page – I really like the icons that represent the IBM Smarter Planet, and also the way the typography is laid out. Very nice.

One of the things that I have been spending time on has been the Smarter Education – the application of technology and systems to improve the effectiveness of the education system.

Claro Theme in DojoToolkit

I recently came across the Claro Theme for the Dojo toolkit. And it was doubly more interesting because IBM has contributed this theme to the community. Maybe a combination of JQuery with Dojo UI could be a killer combination?

Barn owls, from David Whistlecraft on flickr

Last evening, the screeching started again, once the sun set, and I rushed to my window to catch a glimpse of the creature that was making this noise.
As luck would have it, I happened to notice a pair of birds perched on the shaft-sill (which encloses apartment drain pipes). I was overjoyed – more so because I think I sighted Barn owls, even though in the late evening light I really couldn’t make out the details. Am rather confident they are barn owls – watching Animal planet and the like helps I guess.

In any case I noticed them again today, just soon after sun set. Need to get binoculars to observe them better. Looks like they are nesting. A quick web search pointed me to the Bangalore Barn Owl conservation group. So shot off a note to Dr. Subramanya, who runs that group, about this sighting.

Am overjoyed; love owls, and didn’t expect to sight Barn owls in a city setting. This news item provides some more information on why this is beginning to happen: “What a hoot! Barn owls in apartments“

In essence, these are the steps for that – first install the software, then configure it, and finally try the XML/HTTP interface to interact with it.

Install

• Download Tally. I tried the “Educational Version”.
• Double-click the installable to install it.
• Once you start Tally, create a default “Company” in it. Also enable the Payroll module, so that you can perform CRUD operations on Employee information.
• The documentation or a web search can help you with these steps, in case you falter.

Configure

• To initiate XML interaction over HTTP with Tally, the server port has to be enabled in Tally. Click on “Configurations” or F3-> “Advanced Configurations”, and then select the following options: I set the “Tally is acting as” to “Both”, and “Port” to “9002″. Save, and exit. Restart Tally.
• Open a browser, and enter http://localhost:9002 in the address bar. You should see a message “<RESPONSE>Tally.ERP 9 Server is Running</RESPONSE>”. Now you are good to go. Else, have a look at the documentation.

Integration using XML over HTTP

• Tally requires the client (the one which is integrating with Tally) to HTTP-POST XML documents. The XML documents contain the action to be performed, and details of the Tally Data on which this action needs to be performed.
• I searched for some sample XMLs with which I could experiment with this interaction. I came across some useful ones on the RTSLink site: http://www.rtslink.com/download-tally-xml-tags.html.
• Used cURL command line tool to try this out:

curl -X POST localhost:9002 --data @all_inventory_master.xml

• The structure of all_inventory_master.xml is as follows:

<ENVELOPE>
<HEADER>
<TALLYREQUEST>Export Data</TALLYREQUEST>
</HEADER>
<BODY>
<EXPORTDATA>
<REQUESTDESC>
<REPORTNAME>List of Accounts</REPORTNAME>
<STATICVARIABLES>
<SVEXPORTFORMAT>$$SysName:XML</SVEXPORTFORMAT>
<ACCOUNTTYPE>All Inventory Masters</ACCOUNTTYPE>
</STATICVARIABLES>
</REQUESTDESC>
</EXPORTDATA>
</BODY>
</ENVELOPE>

• The next option was to insert an Employee’s information into Tally.
• curl -X POST localhost:9002 –data @payrollEnvelope.xml
• The structure of payrollEnvelope.xml is as follows:

<ENVELOPE>
 <HEADER>
 <TALLYREQUEST>Import Data</TALLYREQUEST>
 </HEADER>
 <BODY>
 <IMPORTDATA>
 <REQUESTDESC>
 <REPORTNAME>All Masters</REPORTNAME>
 <STATICVARIABLES>
 <SVCURRENTCOMPANY>MyCompany</SVCURRENTCOMPANY>
 </STATICVARIABLES>
 </REQUESTDESC>
 <REQUESTDATA>
 <TALLYMESSAGE xmlns:UDF="TallyUDF">
 <COSTCENTRE NAME="ImportPSOFT" RESERVEDNAME="">
 <ADDRESS.LIST TYPE="String">
 <ADDRESS>Bangalore</ADDRESS>
 </ADDRESS.LIST>
 <MAILINGNAME.LIST TYPE="String">
 <MAILINGNAME>10</MAILINGNAME>
 </MAILINGNAME.LIST>
 <CATEGORY>Primary Cost Category</CATEGORY>
 <LOCATION>Bangalore</LOCATION>
 <DESIGNATION>Software Engineer</DESIGNATION>
 <BLOODGROUP>O Positive</BLOODGROUP>
 <AFFECTSSTOCK>No</AFFECTSSTOCK>
 <FORPAYROLL>Yes</FORPAYROLL>
 <FORJOBCOSTING>No</FORJOBCOSTING>
 <ISEMPLOYEEGROUP>No</ISEMPLOYEEGROUP>
 <SORTPOSITION> 1000</SORTPOSITION>
 <DEFAULTLANGUAGE>0</DEFAULTLANGUAGE>
 <LANGUAGENAME.LIST>
 <NAME.LIST TYPE="String">
 <NAME>Kiran Subbaraman</NAME>
 </NAME.LIST>
 <LANGUAGEID> 1033</LANGUAGEID>
 </LANGUAGENAME.LIST>
 </COSTCENTRE>
 </TALLYMESSAGE>
 </REQUESTDATA>
 </IMPORTDATA>
 </BODY>
</ENVELOPE>

• This should insert a new employee record, in Tally.

References

Use this Tally Integration Guide (pdf) to better understand what goes into the various integration options that Tally supports.

Good luck, and let me know if this works for you.

Alphonso mangoes, image from Shayer Jaleel http://www.flickr.com/photos/shayer/

he presented some of us the choicest mangoes from Ratnagiri. The Ratnagiri Alphonso overwhelms the senses, in the most pleasurable way. The fruity light scent, the golden hue of its skin, and the melt-in-your-mouth-explosion of sweetness taste makes a grown up man weep, in sheer joy.

Ok. Soap box done.

But seriously, the Wikipedia content on mangoes, and their Indian varieties, sure needs a lot more content. Wonder where I can source facts from; I would like to edit those pages with some useful information. What exists now is good, but then they don’t do justice to the “king of fruit” (Mangoes, that is, in case you are wondering).

Wikipedia content on mango cultivars: http://en.wikipedia.org/wiki/List_of_mango_cultivars, and specifically Alphonso: http://en.wikipedia.org/wiki/Alphonso_%28mango%29

Now, I have to return to the mangoes. Their mouth-watering scent has filled up my house.

The book is available for purchase on Amazon -  “Govind Narayan’s Mumbai: An Urban Biography from 1863“. I have posted a review of the book there.

There were some minor errors/nitpicks that I noticed in the book:

• Some words in the text are italicized but do not have notes entries for them. This inconsistency is a bit of an annoyance, because the notes are situated at the end of the book, and do not appear as a footnote on the pages that contain the italicized words. Pages 204 and 205 for example contain such examples.
• Some words are not italicized at all… for instance “Bharat Khand” on pg 203. I would have expected this to have some sort of a note, because the non-Indian reader would not have a clue as to what this means.
• Some of the italicized units of measure could have been converted to some metric unit of measure, in the notes section, so that the reader could estimate or visualize what the term means. For instance, “man”, “ser”, etc.
• Some inconsistency in names of places – Pune and Poona are both used in the text.

I haven’t noted down all the nitpicks/minor-errors in the book. Will note them down in this blog entry, the next time I browse through the book.

One of the things I noticed is that the benchmarks numbers are way lower, through this Java client, than if I were to run the ./redis-benchmark. These are the numbers that I see, haven’t figured out what to make of it.

./redis-benchmark numbers

~/redis-beta-8$ ./redis-benchmark -q
PING: 49266.01 requests per second
SET: 56191.01 requests per second
GET: 53191.49 requests per second
INCR: 29697.33 requests per second
LPUSH: 43125.00 requests per second
LPOP: 40249.00 requests per second

Java client benchmark numbers

SET rates: 13027.44 requests per second.
GET rates: 13412.41 requests per second.
SETNX rates: 14118.09 requests per second.
PING rates: 16025.10 requests per second.
INCR rates: 13786.44 requests per second.
LPUSH rates: 15134.22 requests per second.
LPOP rates: 13687.28 requests per second.

The main things we attempted to do:

Upgrade the current laptop

• Current R40e felt sluggish, and some trouble-shooting determined that the Harddrive was the cause. Changed the HDD to a spanking new 160GB IDE one.Current config is 256MB RAM. Something we can live with for the moment.

Install Ubuntu 8-04 LTS version

• Downloaded the ISO image, burnt it on a CD, and tried installing. Several times. Got stuck at the 15% install hang problem. The Alternate “Text based” ISO was suggested as the workaround.
• Downloaded the Alternate ISO CD-image from the MIT Media lab mirror. ISO md5sum matched. Things seemed fine.
• Tried to burn the image onto the CD, but then it was being recognised as a read-only disk, and the cd-burn did not proceed. Gnashing teeth. Brasero didn’t help either. Google did. K3b, with the cdrskin install was suggested as an option, so that whatever kinks exist in cdrecord are smoothened out. Installed it. CD was recognised as a write-able entity. Created the image on the CD.
• The install didn’t start off, as expected, once I booted from the install CD. Tried this several times, with increasing levels of frustration. And then, after a 5 minute break to distract my mind, realized that I had copied the CD-image as a DVD image (k3b provides the option to burn as CD or DVD ISO image. Damn!). Re-wrote the CD-Image as a CD ISO image, and then booted the machine. Install started off.
• The installation didn’t proceed after a point, and complained that the *NVIDIA*deb was corrupt. Bleddy hell!
• Seriously considered moving this back to Windows XP.Took a welcome break. Fortified with some piping hot chai, tried this again.
• Downloaded the Alternate ISO image from the easy-news mirror. ISO md5sum matched. Made a CD image. Kept fingers crossed, and the installation seemed to proceed along just fine. Let it run, and voila, Ubuntu was successfully installed.
• Restarted the machine, and let it boot into the native ubuntu OS. It informed me that I had 200MB of updates. Did that, without a glitch.

System response was rather good. Except when loading it with a whole bunch of resource-hungry applications – runnig youtube videos, and loading Open-office. A bit unresponse, while starting up these applications. but once started, things are ok. During heavy load, the disk seems to be constantly churning – based on the fact that the disk activity indicator light was constantly on. But what else do you expect with a 256MB RAM machine. I was impressed with the performance, nevertheless.

Additional Software

The next step was to install or atleast suggest alternate software, from the laundry list of things that Chandra needed:

1. http://www.inkscape.org/ – apt-get install inkscape
2. www.scribus.net/ – apt-get install scribus
3. hotpot.uvic.ca/ – run it with Wine: http://hotpot.uvic.ca/wine_hotpot62.zip
4. www.irfanview.com/ – not needed, you could use the f-spot image viewer.
5. mozbackup.jasnapaka.com/ – Run this with the Wine option.
6. any linux equivalent of MS Clear Type font tuner – I think the mstt* package will do. Again apt-get install it.
7. S/w to download pics from Canon EOS 350D – f-spot should help there.
8. Mu torrent – loads of options. Ubuntu has a default torrent client.
9. GOM Player – totem media player will help here. Its there by default. else apt-get install it.
10. Google Earth – available for linux
11. Google Talk – use pidgin, or available for linux
12. CD Burner – install cdrskin and k3b
13. PDF writer (one independent is still useful; i know openoffice has inbuilt pdf creator) – comes with a pdf writer option, while using the print-to-file option
14. Skype – available for linux
15. Yahoo Messenger – use pidgin, or use the yahoo messenger available for linux
16. Acrobat reader – not needed, Evince is a default pdf and post-script reader
17. Qtime – is this the apple .mov player? I think there is a decoder available for totem to play this one.
18. RealPlayer – hate this one. but again, there are decoder available for playing this in linux
19. some kind of player for WMV files (i THINK GOM can handle this, but not sure) – totem player.
20. Drivers for HP Business Inkjet 1000, HP Inkjet 3550, UMAX ASTRA 3600 scanner – drivers available to work with this. No worries.
21. install java to run online crosswords – sun-java will help here.

Using Ubuntu: first steps

Need to show Chandra the options to install, remove software from Ubuntu, and also to use some nice facilities like Gnome-Do to lauch software.

Basically she felt she needed to work on some usecases and problems in the area of SOA, and Semantic Web.

Rather than having a discussion straight away on these technology areas, I thought the important aspect was for her to understand these concepts better, and also to look at some other related concepts:

• Tagging / Folksonomies
• Ontology / OWL
• Delicious and Flickr to understand the way tagging helps in these sites, and also the concept of a tag-cluster and its power.
• ReST based webservices – to read Roy Fielding’s thesis on this.

This being the first step, the next is to keep the conversation going, and suggested we use Twitter to exchange notes, or any other medium other than email.

I am on Twitter as kirsn, and use the Twitterfox extension in Firefox browser to read and reply to tweets.