Feedback

Software Development

First, I would like to say that I find your magazine very informative and look forward to reading it every month. I am writing in response to your article in the January 2006 issue titled "Global Success Stories."

I am a 26-year-old software engineer currently employed full time. I have a B.S. and M.S. in software engineering and live near Pittsburgh.

I have to admit I found the article very disturbing. The amount of savings U.S. companies can capture by offshoring their projects is frightening. Yes, frightening. It took me three years after completing my M.S. to find a full-time job, and numerous friends from my graduating class still don't have jobs.

All I have to say to those companies is that they should think further ahead than the tip of their noses. They might be saving money today, but to whom will they sell their software five, 10 or 20 years from now when U.S. consumers are no longer able to afford it due to layoffs or bankruptcies? To India, the leading consumer of pirated software? Also, who will sponsor/pay for your magazine when U.S. customers can no longer afford it?

Please prove me wrong so that I may start sleeping at night again.

Akos Nagy
Software Engineer
Zelienople, Pa.

Coordinating Katrina

Regarding Alexandra Weber Morales's "IT to the Rescue" and "Dredging the Data" (Comment, Nov. and Dec. 2005), I have an interesting story on the "old" and "new" approaches in Katrina evacuee data management.

Over the Labor Day weekend, I catalyzed a project that mobilized 3,000 volunteers to build a "central database" of evacuees. You can see the basic project recap on my blog at http://social source.blogspot.com/2005/11/Katrina- peoplefinder-project-metrics.html.

In "Dredging the Data," volunteer Matthew Rolufs told Weber Morales, "You would think, after 9/11, that there would be a central Web system that would allow people to go in and register themselves as evacuees ..." That's what we thought, too, but in a Web 2.0 world, the key is not building a centralized database, but building a data standard and a network of databases. This is essentially what we did.

We built a data standard (http:// zesty.ca/pfif/). We hand-entered every single unstructured evacuee posting from Craigslist and virtually every other message board on the Web into the PFIF data standard. We scraped all the existing databases (familylinks and others) into the PFIF data standard. We threw it into a database (katrinalist .net). And we made a PFIF feed available to anyone who needed it—including Microsoft.

Ultimately, the vision was that everyone could maintain their own little fiefdoms (CNN wasn't going to give up the traffic their evacuee board generated), but there would be a central repository via a data standard and data exchange standards.

And it was all done by hackers and volunteers over a holiday weekend—without any support from the "major corporations."

David Geilhufe
Cofounder
Social Source Foundation
Beaumont, Tex.

It's a People Problem

In regard to Amit Asaravala's "Defending the Gantt Chart" (Viewpoint, Jan. 2006), yes, Gantt charts are very helpful. Now, if we could just get companies to hire project managers who are comfortable talking to and directing people. Many times I've worked for PMs who were so uncomfortable leading a team that they simply huddled behind their PCs, buried in Microsoft Project. It was comical, actually. These PMs never considered walking around, talking to the team members, asking them how things are going, or if there was anything they needed help with—much less actually getting to know their team!

Chuck Mobley
Support Analyst
Douglasville, Ga.

Give Them What They Want

The first couple versions of our project viewer, Steelray Project Viewer (www .steelray.com), were "ganttless" as well, and our customers demanded that feature. We listened, and when we released a Gantt chart, our sales took off.

Still, Gantt charts aren't perfect. In my opinion, they suffer from two serious problems, and no vendors (including us) have addressed either one. First, a Gantt chart isn't meant to be viewed page by page, or screen by screen. Your mind is focused on finding the right page or scrolling, and both activities are distracting. Second, although a Gantt shows a project as a timeline, it's still only a snapshot of the project at that point in time. You can't see how the project plan itself has changed over time, and that information has a great deal of value. For example, did the project come in on time because the managers deleted 80 percent of the remaining tasks?

Brian Leach
Managing Director
Steelray Software
Atlanta, Ga.

BG (Before Gantt)

What about CPM or PERT diagrams? Before Gantt charts became popular, these worked just fine. I do agree that some form of visual task/activity display is preferable to trying to figure out what's going on from the detailed task lists—a picture is worth more than a thousand words! Microsoft Project is like so many PM products: half-baked and half tested.

Richard Dymalski
Principal IT Consultant
Maricopa County, Office of the CIO
Phoenix, Ariz.

Amit Asaravala responds:

Good point! I'm picking on people who don't like Gantt charts because of the whole Basecamp vs. Microsoft Project battle that seems to have erupted. In general, however, I'd argue that any good project management tool ought to have some way to display information visually as well as textually.

Falling Up and Into

Thanks to Gary Evans for "Going It Alone" (Dec. 2005). I especially liked the magazine's front cover representing all the roles implied when working as a one-man team. You could have added documentation writer, SQL developer and, especially, designer and business analyst. Makes it challenging to target a résumé; something we have to do about every six months.

The reasons why people become independent contractors listed in the article failed to mention my situation. Remember in the movies when the rugged hero recites the line, "I'm not cut out for the corporate thing."? My story is a little less glamorous, but the basic premise holds.

If there is such a thing as charisma, there must be the equivalent of its opposite. Let's just say that I don't think I am personally overburdened with charisma. I also have a Steve McQueen-esque inability to show deference to people who feel they possess a job title that bequeaths upon them regal authority. For me, such status has to be demonstrated and earned.

Add to that my fiercely liberal political leaning and Dallas locale, and it should come as no surprise to learn that I tend to churn through jobs. Fortunately, because I'm intelligent and good at what I do, which happens to be computer software application development, I generally manage to "fall up." Each contract I get is usually more interesting and challenging, and tends to pay better than the one I lost.

So far I have been able to pull this off for going on 20 years, working through contract agencies. I've tried to bypass their overhead only to learn the big difference between submitting an invoice and getting paid. Working for an agency, I just fill out a timecard, and if the money isn't posted to my account on time, they hear about it. Getting the client to pay is not my problem. And I won't work 1099. Who can allocate and fill out quarterly FICA statements? I like W-2s, and usually get four to 10 at the end of every year.

The headhunting firms seem to have a greater personal turnover rate than even we programmers, so there are always lots of well-dressed people eager to place me with their fabulous new client. I've been able to "burn bridges" all over town and never suffer from a shortage of calls. For me, the greatest invention has been Monster, Dice, HotJobs and so on. I remember having to look at the Sunday classified ads and having to send faxes and actual stamped envelopes—but then I remember sequential access tape and green bar printers. But nowadays I can get a new job in about three hours.

I'm writing this to suggest that certain people might fall into independent contracting for reasons not entirely of their own choosing.

Name Withheld

Not So Literal

In his article "Going It Alone," Gary Evans writes, "My first client literally fell into my lap." To quote from The Princess Bride: "I don't think that word means what you think it means."

Clay Shannon
Senior Delphi Programmer
Dimension 4 Software
Oconomowoc, Wisc.

Proliferating Podcasts

I've enjoyed the Software Development podcasts. I particularly appreciated the focused subject matter and short duration. Too many podcasts seem to ramble, but these are nice bite-sized chunks.

I found them while checking out some referrers of a podcast I've set up at www.qapodcast.com. We've done eight episodes so far, each one a 20- to 25-minute conversation about some technical or business aspect of QA. I thought you might be interested, as it features similar topics.

Darren Barefoot
Partner
Capulet Communications
Vancouver, B.C.
Canada<

Mike Riley responds:

Thanks for the kind words, as well as bringing qapodcast.com to my attention. I've subscribed to the RSS feed, downloaded a few of the shows and queued them up on my MP3 player. They should make for some good listening during the morning commute.

Science Trumps Dogma

In his letter "Flaming Pulpits" (Feedback, Dec. 2005), Harry Rockefeller suggests that the U.S. tech industry can overcome its slide into mediocrity simply by embracing the ideology of Christian fundamentalism. I disagree. The foundation on which our tech industry lies is science. However, the dogma of Christian fundamentalism demands that we replace sound scientific theory with such foolhardy notions as creationism and intelligent design. Such action would not only fail to slow our tech industry's slide, it would accelerate it.

Tim Buchowski
Supervising Programmer
Austin, Tex.

Resorting to Religion

In his response to my letter regarding this country's future ("Decaying into Mediocrity," Feedback, Oct. 2005), Mr. Rockefeller takes exception to my pessimism but may not have realized that it's less the opinion of a misanthrope and more the consequence of awareness of fact, including standard of living, educational testing scores, life expectancy, infant mortality, GDP per capita, numbers of engineers graduated annually compared to numbers of lawyers, and numerous other hard statistical data.

More to the point, he may wish to consider that resorting to religion is typical of failing cultures. When reality becomes grim and disappointing, the rosy promises of the flaming pulpit attract adherents. Few bother looking beyond its vague and comforting tales. His suggestion that only his religion has the answers is fundamentally offensive in a pluralistic and nontheocratic state.

If I am wrongly pessimistic, sir, please show me incontrovertible evidence that all reasonable persons will accept as objective. George Santayana answered your question, "Does that mean there's no hope for us?" when he wrote, "Those who cannot remember the past are condemned to repeat it."

David Fisher
President
Fisher Aircraft Corporation
Rochester, N.Y.

UML Smackdown

Carl Dreher's rather uncivil comments ("A Classic Case of Denial," Feedback, Dec. 2005) regarding "The Coding Conservatives" (Deadline, Oct. 2005) compelled me to dig out the October issue to see for myself what could have stimulated such a visceral response.

What I found was well-considered comment on the deficiencies of the current state of the art of software development practices. Dr. Dreher's ignorance of the fundamental benefits of modeling is clearly demonstrated by his direct slap at UML.

UML is not the point. The point that Mr. Selic clearly made is that software development practices haven't progressed because software developers don't want them to. Until this barrier is overcome, no formal methods will be universally successful.

Mark Eisenberg
Principal Consultant
Eisenberg Consulting
Acton, Mass.