Home Rox's homepage
Integrity - Great Book
Sunday, 15 May 2011 23:29

I read all of it... Then I went back and re-read it and took notes.  Then I pondered my notes in relation to my life, my relationships, and my reactions.

This is one of those books that comes along at the right time, in the right way, to say what you need to hear when you need to hear it.  At least for me.

Dr. Henry Cloud has spent years advising CEO's on the aspects of quality leadership.  He has boiled that down into various aspects/qualities of an individual's behavior that indicate "integrity" - which is more than honor or honesty.  It is more than trustworthy.  Someone who has "it" you'd like to partner in .. in whatever they are doing because it will turn out GOOD.

I find there are parts of this book that make me uncomfortable.  I fall way short.  There are parts where I can see clients and co-workers have fallen way short.  I can see why there have been times where the team I was in just "sang"/"vibrated"/"clicked"... or "flowed" in his terminology.

So... if you want to improve your relationships in life.  No matter whether it's your children,  your spouse, your co-workers, your boss, your clients, or your business... yeah - go read this book and ponder.

Integrity: The Courage to Meet the Demands of Reality by Dr. Henry Cloud.

Ramblings on Trends
Tuesday, 15 March 2011 14:48

I was reading through the Communication of the ACM today.  I like skimming the articles.  Sometimes I learn something.  More often I see trends in thought, which then makes me think about where that will lead "us" in time...

Communications of the ACM.  March 2011 Volume 54 No. 3

In the letters to the editor Basudeb Gupta espouses that all software people (he used engineers) should be classified as engineers and be required to be certified, retrained, and controlled by licensing, much as Medical Doctors and Engineers are.  It's an argument I've heard off and on through my career, which isn't as long as his, but is sufficiently long to know why he's espousing his viewpoint.  Given the footnotes to articles published in the Communications of the ACM and their dates, this is a long standing argument for him.

I don't disagree on some levels, but to keep people from practicing without licensing, I disagree.  I think it's the equivalent of censuring software. You also run the risk of blotting out, ignoring, and pushing to the side valid approaches because those who control the process disagree with the value or validity. [look at difference in the philosophy of approach between Eastern and Western medicine - then look at the history of how and where they are practiced and/or controlled].

World Wide Web

Several years ago my daughter came home from a friend's house totally incensed.  She wanted us to sign a petition, she wanted us to broadcast this horrid event to our friends and get them to sign a petition, she wanted to have "something done" about this event... She and her friend had been told about a blog post where the person had written an article about an art show in Argentina.  Supposedly one of the artists had chained a dog in the corner of the exhibit and starved it to death... as a form of art.

So, her father and I spent some time researching "this event".  The blog post was written by an individual.  They weren't associated with a newspaper, they weren't associated with the museum in question.  They were "just an individual".  We looked further. We found a news article written for a reputable news agency (well as reputable as any of them get).  They hadn't gotten hold of the artist, but they had gotten hold of the museum director.  His story was... the artist had picked the dog up off the street - it was starving when he found it.  He did tie it in the museum - to raise awareness about the ignored populace of dogs in the city - that were starving.  The dog had been fed and watered, it gained weight while "on exhibit".  It had escaped one night, and so "disappeared" from display.

Who would you believe?

Logos forum

Logos converts printed material into researchable and cross-referenced electronic books.  They concentrate in the religious/historical sector. Someone made a request for a specific book to be added to the "library of research material" supported by Logos.  It is an English translation of a Syriac text.  My husband spent most of a morning pulling together a carefully constructed comment on the quality of the particular book's translation specifically in relation to *research* to be used by an unknown number of people with likely a very low level of vertical knowledge of the field. [Specifically Syriac language translation].  In short, his detail review of the material exposed some significant failures to conform with accepted interpretation of portions of the text.

How would they know whether the source was sound?

Communications of the ACM.  March 2011 Volume 54 No. 3

On pages 12 and 13 in the blog@ACM section there were two articles that dove tail into this discussion...

Mark Guzdial : What do Scientists and Engineers need to Know about Computer Science?

Chris Scaffidi, Mary Shaw, and rad Myers have estimated that, by 2012, there will be about three million professional software developers in the U.S, but there will also be about 13 million end-user programmers - people who program as part of the their work, but do not primarily develop software.
When they see "try-catch" in a piece of code that they're trying to understand, they don't know how to look up "exception handling", and they can easily spend hours reading about Java exception handling when they are actually working in JavaScript.

How close does this relate to my problem?

Greg Linden: Research in the Wild: Making Research Work in Industry

To take one example in search, without researchers who know the latest work, it would be hard for a company to build the thousands of classifiers that ferret out subtleties of query intent, document meaning, and spamminess, all of which is needed for a high-quality search experience.  Information retrieval is a field that benefits from a long history of past work, and researchers often are the ones that know the history and how to stand on giants' shoulders.

Who is an expert, what are their bias's?


As I understand it... Wikipedia uses a structured review process to validate article content while allowing virtually anyone to say anything about anything. They depend upon balanced review by interested parties to argue with the content and revise it towards some level of quality result.

Is this a valid approach, are the results valid?


Ebay instigated a feedback system and a arbitration process to allow the populace to police themselves.  If people generally have a good reaction to working with a specific person and they feel like telling people about it...

Are they trust-worthy...enough?  What are my guarantees?

Mankind has traditionally assigned "rank" to a person through birth or personal achievement which resulted in a visible standing. [material assets, wealth or titles]  Along with that comes some level of "trust" factor on the validity of their input about any specific topic.  But those too can be flawed.  Titles can be given for the wrong reason.  Assets can be acquired without the attendant skill to amass them [inherited wealth].

With the World Wide Web's low bar for entry... We have a world full of monkeys typing at a typewriter, and occasionally someone publishes a masterpiece.  But how do we find it and how do we know it's a masterpiece?

The question comes down to "How do we know whether to trust the content of what we are reading?"

Where does this lead?  While the search engines rank results in searches, they can't and don't apply any secondary validity rank on what is supplied. But then, should they?  "Who died and made you king?"

Information validity or source integrity, not only in the "wild woolly world of the web", but internal to organizations is becoming of huge concern with the exponential growth of electronic noise available to us. How we capture it, how we clean it, how we classify it, how we search it, and how we connect it to other results will continue to be vital to finding those nuggets of value and identifying those who produce more nuggets vs noise.

Reputation.  In a small community, reputation determines the value of their input. Generally people care about their reputation, they nurture it.  When that is migrated into a large community, it gets diluted in strength for the majority of the people, and concentrated into the few who are visible.

Somehow, we need to be able to assign a "pecking order" to the articles/data/information that comes our way, based on the perceived quality of the source.


Customer Support can make or break a sale...and a project
Friday, 09 July 2010 00:00

Funny how life moves on.  Raising two children who are 12 years apart is an interesting process, and it tends to take up all my spare time.  I read blogs from people who are passionate about different technologies,  I find I usually don't have the spare time.  I understand them, but it hasn't been me at this time.  My children are my passion - they have to be.  That being said - my work is where my analytical skills most often come into play.

My current client is attempting to create a generic product which will integrate data and provide monitoring and analysis information for social programs.  One of the technologies we had to identify was a way to migrate data from operational systems to a Data Warehouse.  I've done migrations from time to time, usually as one time efforts.  Most of those ended up custom SQL/perl/bat/batch scripts.  Because this is an ongoing migration, we looked at Extract, Transform, and Load (ETL) tools.  Because this is a smallish corporation, we looked at "open source" solutions.  The big players, Ab Initio, Informatica are too expensive. Oracle Warehouse Builder wasn't considered.  They don't use Oracle.

There is a plethora of open source ETL related packages on Sourceforge.  The biggest players in that arena are Pentaho, Talend, and CloverETL.  All of these products are commercial for their enterprise solution and provide some level of "open source".  Pentaho's Kettle is integrated into their overall Business Information Suite.   Given data security considerations on the project for the BI portion which would force customization of Pentaho to support them with a lack of in house knowledge, their Kettle solution was excluded from consideration.  That left Talend and CloverETL.

Talend provides a free Community Edition Designer for their ETL.  They provide a commercial package upgrade for people who have more in-depth needs.  For proof-of-concept scenarios, they provide customized (paid) consulting.  Having had a short delivery timeline for doing the comparative analysis and no financial budget to pay for their proof-of-concept approach, Talend lost to CloverETL through shear licensing approach.

CloverETL will be releasing a Community Edition version, but currently provides a free 30 day trial of their commercial product.  The initial screening for products took place 60 days before the actual effort to do a comparison.  Our trial had run out.  I dropped a line to customer support asking them if they'd be willing to extend our license, or whether we'd have to go with Talend by default.  Customer support passed the request on to Sales.

We were not only given an extension, and access to their high end product, the Server, but provided direct consultation with one of their top technologists to help identify how CloverETL could best meet our needs.  They then followed through with answering questions, demonstrations, and technical support (via the forum) to help us get up and operational with a proof-of-concept.

Javelin Inc. is a breath of fresh air in the world of high tech consulting.  They price their products to make a profit, but commit themselves to their client's success with the product first.  We all have to make a living, we can feed off one another as vultures, or we can build a pyramid by lifting each other up.

My client is moving forward with their efforts, and CloverETL has become part of their solution - all because Customer Support passed a request on to Sales, and Sales is committed to helping each and every customer succeed with their product.

Tax Season
Friday, 12 March 2010 04:19

We've run our own company for over 15 years.  I'd done my own taxes for over 10 years before that.  Multi-state, Schedule K's, investment income/losses, multiple jobs, dependents or none.  I've learned to just read through tax booklets and pay attention.  We shut down our old S-Corporation and started a new one for timing and cost reasons. [We couldn't get the old one up and operational in a new state fast enough to suit us for a reasonable price tag, but we could close and open a new one instead].  So now I find myself doing two sets of Corporate books with a few interesting activities that I've never had to deal with before [closing a corporation with a loss due to a non-business investment].  I went looking for a little bit of auto-magic guidance in the form of a tax program.  Not too many tax programs deal with S-Corporation forms.  Given the field, the amount of money I really didn't want to spend, and well, general impression of reputation - I went with H & R Block's Business application.

It will generate the forms that I need and I'll be able to do my personal taxes and electronically file them.  In the end, 2 corporate federals, 2 corporate state, and 1 personal federal/state filings.

BUT what a piece of junk.  I'm not taking time to identify the source for any of my issues, but the screens are totally messed up with radio buttons overlaying or totally offset from the text they are associated with.  At least it's consistent, so I can deal with it but having to count down the radio buttons and find the one associated with the text I want is annoying.

Once you have numbers in the system, if you go through the review, there are warnings and/or comments in there that, yeah ok, I can see how you got that, but it doesn't apply to me [I'm not using a professional preparer, so the fact that all those fields are empty is just fine].  There isn't a way to tell the program to ignore those.  I can ignore them and the program will print forms with the warnings and errors in effect, but still....

A couple of the errors make no sense, aren't clear where in the form process they apply to, and certainly don't give any detail level of discussion on what the error is, what it means, and what you might like to do about it.

Once you go to look at your returns and you decide that a specific number in a specific field isn't where you need it to be, it's hit or miss to find the interview question that resulted in the figure.  You have to hunt and peck and guess at which set of screens set the figure in order to change it.

Interview screens can result in new schedules or attachments being added to the overall packet as you answer questions one way or another.  You can choose to delete those if later you decide you don't really need that form.  Earlier today I was doing just that - moving numbers from one form to another because after reading the IRS docs I decided that I didn't need that specific form.  Well, I'm not sure now exactly what I was on, but I managed to delete the main 1120S form with all the corporate information, but leave the 2nd - 4th pages of that form along with the schedules and other forms.  Now I get an error that you really shouldn't open a Schedule K-1 w/o an S-Corporation.  Well - duh.  The delete page wasn't any more or less informational or warning than any other - but suddenly my company is gone.  OH, and to boot, when I quit in order to get back to the last saved (by me) version - what the program has stored is... the now corrupt file.  I get to start over.

I don't know what kind of help I thought I  might get from a program for less than $75 sold to literally thousands of people every year.  However I've found that I'm using the IRS pubs more than any help within the program and having to do what I was trying to avoid doing - read through all those docs to find the right place to put the information into the program to get the right amounts in the right fields on the right forms.

Frankly doing the forms by hand and then typing them in would be faster and a lot less frustrating.  If you have a choice in tax programs - I'd move on to another brand.

For just your personal taxes - H&R online is actually pretty good too - used that in 2007 or 2008.  I used olt.com last year.  It did take a little to get used to, but in thinking back to olt and using H&R this year - based on this system I'd use olt again.  But whatever you do, don't buy the H&R desktop version for business.  As a software engineer with over 26 years of experience - I'm not the least impressed with their Quality Assurance department or their UI design team.

Monday, 15 February 2010 16:39

So DC has had a record setting snowfall this winter.  During this last storm we had about 30 inches of snow dropped ontop of us.  We lost power for 4 days.  We heated the house with wood, boiled snow for water, and used the camp stove to cook on (and yes I know it puts out carbon monoxide).

Taken for posterity:

The wood pile

The Rabbit trail from snow 1