" /> CTDATA: October 2000 Archives

« September 2000 | Main | November 2000 »

October 31, 2000

Lottadot.org Makes Slashcode Mailing List Archive Searchable

On the Slashcode Mailing List, Shane pointed out that he is hosting a searchable archive of the mailing list. It is located at http://search.lottadot.org/cgi-bin/texis/slashcode/search.

This is a really spartan page, but it is a valuable resource for developers who are building Slash-based Web Sites. We added it to our list of "Slashcode Support Resources" in the "Quick Links" Slashbox on our home page.

October 30, 2000

How and Why We Built Our Own Slash Implementation

Last week, David Bernard Cuisinier sent the following message, "I've been following your site for a while now, and the progress being made concerning Slash. But I'm still in the "newbie" category. Have you ever thought of sharing your experience with running CTDATA on Slash, through a FAQ or HOWTO that would be better than the puny existing one?"

We decided that we would try to address this by first explaining how CTDATA built its own version of Slash and why we chose this route to get our site up and running.

Where did the idea come from?

We think that a lot of Slashdot users have entertained the idea of building a Web Site that acts like Slashdot in some fashion. However, many people with this idea do not have previous experience building database-driven Web Sites. We feel fortunate that Dave Aiello had about two years of experience building Web Sites driven by databases on a consulting basis for financial services companies.

However, the real key to our successful implementation of a Slash-based system was that CTDATA's desire to build a Slash-based Web Site for itself corresponded with one of our corporate client's desire to build a Web-based knowledge management system that would have a Weblog-style interface, at least on its home page. This allowed us to use Slash as the basis for that client's Intranet site.

Without the alignment of these two sets of goals, we probably wouldn't have succeeded in implementing Slash in such a short time.

When did you start development?
When did your site(s) go live?

We started developing our Slash implementation in November 1999. It took until January before most of the core functions of Slashdot worked in our target environment. Our client's Slash-based knowledge management system officially went live in March 2000. We implemented ctdata.com as a Slash site on June 21, 2000. The Web Site for the Rensselaer Club of New Jersey, rcnj.org, was turned on September 5, 2000.

Why did you start with Slashcode 0.3?

Rob Malda did not announce the availability of Slashcode 0.9 until January 26, 2000. Prior to that, he and Jeff Bates were fairly adamant in their refusal to provide a road map to future releases of the Slash source.

So until January 2000, developers interested in doing Slash-based sites had to choose their own path. Some, like Kuro5hin, started their own Slash-like projects out of frustration. We decided to develop on top of the Slash distribution in existance at the time.

For the record, that tar ball we began with was called Slash 0.3.

What is the major difference between your version of Slash and the distribution that you started with?

We began with Slash 0.3, a distribution that only ran on Linux with mySQL, Apache, and mod_perl. Neither CTDATA nor its consulting clients used any of these tools when we began this project. Therefore, we were forced to port Slash 0.3 so that it was far more platform-neutral.

The current version of our code is known to run in the following environments:

  • Operating Systems: Solaris 2.6 and Windows NT 4.0
  • Databases: Sybase 11.x and Microsoft SQL Server 6.5
  • Web Server: Netscape Enterprise Server 3.5x and later
  • Web Server / Executable Program Interface: Common Gateway Interface (CGI)
Since we forked from Slash 0.3, we added a number of user interface features to our implementation. Some of the features have not been implemented elsewhere, but a number of them are implementation of features that came into Slashcode after the release of 0.90. (We hope to document some of these in future articles on this Web Site.)

Do you plan to maintain this fork indefinitely?
Is your intent to create a competing project?

Maintaining a fork from Slashcode is not in CTDATA's long term business interests. However, we cannot migrate off of our fork until we can develop the experience necessary to run a Linux server securely on the Internet.

Of course, we could strike out on our own again and implement Slashcode 1.09 or Bender on Windows NT or Solaris, utilizing the same porting techniques as we did in 1999. However, we feel that our lack of Linux expertise is an impediment to our long-term participation in the Slashcode project and a disservice to our current clients. The Slashcode project might benefit from our effort to make Slashcode less UNIX-dependent. But, from a practical standpoint, we must become experienced on the Linux platform so that we can adopt code changes applied to the base distribution quickly.

We support the effort of the leaders of the Slashcode project to implement the so-called Bender release because:

  1. it will make it easier for us to contribute our enhancements to the main Slashcode project, and
  2. it will allow us to migrate our clients to Apache / mod_perl-based Web Servers without requiring them to also change operating systems and databases.

What are you doing to develop expertise on Linux?

Since we have come to the conclusion that we should merge our efforts thus far into the main Slashcode project, we are working hard on building Linux expertise. So far, we have succeeded in installing Slash 1.09 on a small machine running RedHat 6.1. This machine will be our test environment until we can get a copy of Bender running on a larger server.

Can you share any of your experience installing Slash on Linux for the first time, since many inexperienced Linux users might benefit?

About a week ago, we realized that a number of people might benefit from notes on our first installation experience, since we are coming from a Windows NT-user perspective. We will try to distill any insight that we have into an article or two, and post them here in the future.

Feel free to let us know if you have any comments, questions, or suggestions. Send an email to Dave Aiello or leave a comment attached to this story.

October 29, 2000

AOL to Its Customers: Home Page? What Home Page?

According to CNET, America On-Line has permanently fixed the home page on the browser portion of its new AOL 6.0 client software to its own Web Site.

What surprises us is that no one has picked up on this statement-- perhaps one of the dumbest statements made by an Internet executive in recent history:

"What we did was we looked at the
toolbar as a whole, and in that
redesign, in that streamlining, we moved some things around," said Jeff
Kimball, executive director at AOL. "And yes, {the home page button is} not there."

Forbes Says Sun Enterprise Servers Have Cache Memory Glitch

Forbes is reporting that Sun Mircrosystems has known about a glitch that has caused a number of its most expensive Enterprise Servers to crash repeatedly. These servers are deployed at large customers like Network Solutions and BellSouth.

According to the article, "A major telecommunications company endured repeated crashes on Sun servers. It
recently committed to buy its next batch of Unix servers from HP. While no one has
reported a consequential data loss from a crash, one Sun customer claims to have spent
$3 million trying to diagnose and fix the problem, according to Gartner Group."

It's amazing that information about this problem has been not come to light sooner. These servers are marketed as the gold standard of uptime, at least since server products like Tandem fell out of favor.

Another strange aspect of this story is that several of our largest clients are major customers of Sun. How could problems like this exist without IT people on Wall Street discussing it openly?

Finally, where are Web Sites like Slashdot on this issue? I haven't seen any geek-oriented Web Site report this yet.

October 28, 2000

Scripting News Readers' Dinner Held in NYC

Wednesday night, Dave Winer held a dinner for readers of his Scripting News Web Site. The dinner took place at Katz's Deli, a Jewish Deli with "atmosphere" on the Lower East Side of NYC.

Just as Scripting News is interesting its ecelectic mix of issues, the conversation at the dinner ran the gamut from the World Series to the RSS 1.0 standardization effort.

Summary of the major issues:

  • RSS 1.0: A good deal of the discussion at this dinner meeting revolved around RSS 1.0, a proposed extension to the RSS 0.9x content syndication standard. Dave Winer has been a vocal critic of that effort, and many of the participants agreed with him. Although we would like to do a more comprehensive analysis of RSS, a summary of CTDATA's position is that we feel that the RSS 1.0 proposal is more complex than we would like and was developed without the consensus of some of the major RSS users and developers.
  • Radio Userland: This is a music-oriented Weblog development and community building tool. It seems to be a hybrid of Weblog and Napster-like functionality. Dave Winer said he began developing the product because he believed that the U.S. Federal Courts would put Napster out of business this past Summer. A lot of the diners had worked with Radio Userland extensively, and it was clear that Dave Winer felt that this was one of the more interesting projects that he had worked on in his long career.
  • Adam Curry: Adam Curry was not an issue at the dinner, but he did attend the dinner and express some of his ideas about a personal server product that he would like to see. For those who do not remember, Adam Curry is known in the United States primarily for his work on MTV. However, he is an experienced business person with a good eye for technology. His work is worth following.
  • Creating Trading Markets in Bandwidth and Processing Power: One of the ideas that came up in the discussion was the idea of being able to harness the bandwidth or processing power of other Internet users who are not currently on-line. This was an idea brought up by Dave Winer, just to see if people felt that the idea had any potential beyond SETI@Home.

October 23, 2000

First Instance of Slashcode 1.x Running in Development at CTDATA

Dave Aiello wrote, "After nearly 2 weeks of tweaking and head scratching, I finally got Slashcode 1.0.9 running on my Dell Inspiron 7500 under RedHat. The work seemed really tedious, but that's probably because I do not know mySQL or Apache at all."

"In spite of the fact that it has come a long way from the 0.3 days, the installation process for Slashcode is still poorly documented. If it took someone like me (with 4 running instances of Slash 0.3) nearly a man-week to get my first instance of Slash 1.x running, imagine how long it would take a real newby."

Dave Aiello continues, "The good news is that I may start posting stories to this Web Site and to RCNJ.org now that I am past this milestone. I know all of you loyal readers have been waiting with bated breath...."

October 17, 2000

OSInvestor.com: An Investment-Oriented News Site about Open Source

We stumbled upon a Web Site called OSInvestor.com that provides news about the OpenSource-based business community from an investment perspective. This site has apparently been in existence since February, and the site credits indicate it is run by a sophomore in high school, Rob Radez, and his friend Todd Ostermeier from the University of Illinois.

Far be it from us to criticize a company started by a couple of kids. That was a good description of CTDATA 10 to 15 years ago. The question is: are they providing a valuable service and are they delivering it on a consistant basis. The answer, appears to be "yes."

No one is going to confuse OSInvestor.com with The Red Herring or The Industry Standard, but if you are looking for a site that is devoted to news about companies that are focused on OpenSource, this is the best one we've seen so far.

We found this site through its headline syndication placement on egrail.org. eGrail is an open source content management system being developed by the company called eGrail. (Note that eGrail's corporate Web Site is different from the eGrail Project Site.)

October 12, 2000

Headline Mailer Repaired on ctdata.com and rcnj.org

The headline mail services on ctdata.com and rcnj.org were inadvertantly broken during an upgrade that took place last Saturday, October 7.

The problem has been resolved, but few headline mailings have been made this week due to a reduced volume of stories on both sites.

What can we say? We've been busy.

October 10, 2000

Does Pearl Jam Have an Effective Strategy Against Napster?

A few days ago, USA Today reported that Pearl Jam continues to set industry records in the wake of their unprecedented concurrent release of 25 albums. They have now succeeded in putting five of those albums into the Billboard Top 200 in the first week that they were eligible. This is the first time that any recording artist has placed any more than two new releases or more than one live album on the Top 200 at the same time.

We wonder if these low cost, live albums are the future of the traditional recording industry. They certainly provide Pearl Jam fans with an appealing alternative to poor quality bootlegs distributed through Napster.

In spite of the fact that we don't generally agree with the band's political views, they are quite entertaining. And they certainly have used the normal distribution channels in a savvy way with these releases.

We already purchased the Katowice, Poland concert referred to in the USA Today article. We are likely to purchase others as well.

October 9, 2000

Clinton - Lazio Debate Includes Question about Email Hoax

ABCnews.com is reporting that the moderator of the second Hillary Clinton - Rick Lazio Debate stumped the candidates with a question about Bill 602P. Marcia Kramer, a correspondent for WCBS-TV in New York asked {a question apparently directed at Mrs. Clinton}, "I'd like to ask you how you stand on federal bill 602P... under the bill thatís
now before Congress, the U.S. Postal Service will bill
e-mail users 5 cents for each e-mail they send even though
the post office provides no service."

Neither candidate had any idea what Kramer was talking about, although both voiced some degree of opposition to the concept. The reason that they were stumped is that the viewer-submitted question is based on an email hoax that has been circulating for well over a year.

It seems that broadcasters cannot resist the temptation to include an on-line component in their programs, even when they run the risk of gaffes like this one. However, the performance of Marcia Kramer in this instance shows the extent of the pressure that journalists feel. Obviously, Kramer should never have asked the question if she:

  • didn't understand it, or
  • understood it but wasn't familiar with the issue.

We wonder how long it will be before a candidate is asked if he agrees that the case of imprisoned political activist Ba Ba Booey should be reviewed with an eye toward giving him an unconditional pardon?

October 7, 2000

mysocialsecurity.org Makes the Case for Privatization

The National Center for Policy Analysis has just launched mysocialsecurity.com. This Web Site presents itself as a calculator that attempts to show the difference in returns that a user would experience from a nationalized Social Security system and a privatized alternative system.

This site could prove to be an effective counter-argument to Vice President Al Gore's Social Security proposal if it is promoted properly.

CTDATA tries to limit the number of political positions it takes on this Web Site. Most of the stories we have run on political issues have been related to economic issues and legal issues that affect the Internet. As a company with a number of relatively young employees, our bias is to favor a partial or full privatization of the U.S. Social Security system.

However, the purpose of this article is simply to point out a potentially effective use of the Web to reach technologically optimistic people.

Wired Explains Impediment to US - Canada Poolings of Interest

Martin O'Donnell pointed out an article on Wired News that identifies a really obvious disadvantage of starting a new business in Canada. According to the article, "The most contentious of the tax laws concerns cross-border
merger restrictions and capital gains. When an American
company purchases a Canadian firm in an all-stock deal,
Revenue Canada taxes the Canadians on the capital gains
even though they have not cashed out their shares."

For some reason, Revenue Canada has decided to single out mergers between Canadian and non-Canadian companies for this tax treatment.

This is evidenced by another quote from the Wired News article, "In comparison, the capital gains tax does not have to be paid
until the shares are converted to cash if the deal is between
two Canadian companies or two American ones."

Rick Nathan of Brightspark is quoted as saying that, "he plans to incorporate 10 startup companies in
Delaware that... {his company plans} to incubate over the next

That makes sense, eh?

October 6, 2000

GuruNet Recommended for Research Tasks

Julie Aiello pointed out that Walter Mossberg reviewed GuruNet in his latest Personal Technology Column in the Wall Street Journal. Julie is a professional researcher at Gund Investment Corporation, and said of GuruNet, "so far it's been a terrific research tool - much quicker than logging into
Yahoo or some other search engine to look something up when I'm trying to find
information quickly."

In his review of the product, Walter Mossberg wrote, "It's not that hard to stump GuruNet. For instance, if you click on "DVD," the
acronym for the popular new video-disk format, it thinks you must mean
Dover Downs Entertainment, a racetrack company in Delaware whose
stock symbol is DVD. A click on the word "Cisco" yields information on
Cisco, Ga., not the technology firm." Some of us felt the same way about searches that we performed.

For instance, Julie Aiello said that she, "... tried typing in 'Dave Aiello' and sure
enough the first option was to go to {Dave Aiello's} resume on CTDATA.com." It should be pointed out that this result could not be duplicated on all machines. This leaves us wondering if the difference in performance has to do with the applications where the data resides that GuruNet uses to perform its search.

Regardless of these slight glitches, we agree that this product is novel and likely to improve over time. Therefore, we recommend it to our customers.

CNET Provides Radiation Statistics on Popular Mobile Phones

We stumbled across a promotion on CNET for a report that they published on mobile telephone radiation emissions. Some of our favorite phones, including the Nokia 8860 and the Ericsson T28 have among the highest radiation statistics.

This made us curious to find out what the radiation statistics meant. After all, why are the most feature-ladened, newer models apparently discharging more radiation than some of the old models we discarded because they were big and heavy?

The Cellular Telephone Industry Association through its World of Wireless Web Site, provides an explanation of the statistic used to measure radiation from mobile phones, the Specific Absorbtion Rate or SAR. You will note that SAR is measured in Watts per kilogram (W/kg). The SAR statistic for mobile phones is also carefully explained to be the partial body SAR.

The issue for us is whether the kilogram in SAR is a measure of the mass of the human body or the telephone. If it is a measure of the mass of the telephone, then it is perfectly understandable that the smallest telephones would have the highest SARs, irrespective of the phones' feature sets. If, on the other hand, the statistic refers to mass of human flesh, then it is easier to understand the relative exposure rates.

October 5, 2000

Dave Sims' Weblog Reports on Rob Malda's MIT Speech

Dave Sims posted a report from Joe Johnston about a recent speech that Rob Malda gave at the Massachusetts Institute of Technology. Rob Malda is the founder of the Slashdot Web Site.

This speech was briefly mentioned on Slashdot before it occurred, but this is the first account of Malda's performance that we have seen on the Web.

Slashcode Publishes Report about RCNJ.org

We note with pleasure that Slashcode.com, the official Web Site for the Slashcode Open Soure Project, reported on our relaunch of the Rensselaer Club of New Jersey Web Site, RCNJ.org.

Why do we gleefully point to every mention of our Web Sites on Slashcode? Because the people running the project have been very supportive of our efforts in spite of the fact that we aren't using the same code base at the moment. The "Slash is Slash" philosophy is one of the reasons that we are so loyal to the Slashcode project.

One day, we hope to contribute our enhancements to the main code base. But first, we need the Bender release of Slashcode to become a reality. We also need to migrate to Apache.

Of course, it would probably help if we ran a more UNIX-style operating system as well.

October 4, 2000

Prescription Errors May Be Reduced if Doctors' Handwriting Improves

CNN is reporting an AP story that says that hospitals are offering penmanship classes for doctors in an attempt to reduce the number of prescription errors. According to the article, "... up to 25 percent of medication errors may be related to illegible handwriting."

In his daily broadcast of The Osgood File, Charles Osgood of CBS suggested that errors would also be reduced if doctors used PDAs instead of prescription pads. We are looking for source material to document his statements. If we find it, we will post it as an addendum to this story.

Developer Calls Linux a "Failure", Cites "Freeloader" Ecosystem

Monty Manley writing on OS Opinion says that the hype surrounding Linux and the Open Source Movement has masked a harsh reality: thousands of buggy, half-finished software products that are not properly unit tested. This problem is allegedly made worse by the fact that many OpenSource projects are driven by volunteers and that the efforts of these contributors are second guessed by users who do not contribute to the project.

Manley says, "It's an odd fact that many of the most irrational and vociferous OS zealots are
not themselves programmers... Linux, more
than any other OS, suffers from a surfeit of testosterone-poisoned young men
who know little but speak much, and the whole community suffers from it. They
want things free simply because they don't want to pay. Saying that charging
for software is ethically wrong is only a dodge; they just don't want to part with
any dough. They are, in a word, punks. And Linux has far too many of them."

How many more "tragedy of the commons" editorials need to be written before people face facts? Read on if you are interested in our view on this subject.

The OpenSource movement is largely a volunteer effort. People who volunteer do so knowing that there will be people who use their software and don't contribute. Some users will even criticize the efforts of the project developers. Are there any differences between this volunteer organization and any other in America? Probably not.

It's hard to say whether Monty Manley's editorial attempts to cover too much ground, or if it simply rambles. He criticizes Linux users for demanding that commercial software products be GPLed. We are sure that he is not the first person to point this out. Yet, we always ask why the mere demand that something be released under an OpenSource license forces the developer to do so -- especially if the developer is not an individual but a real company?

Regarding the issue of unit testing and system testing, for years our major clients have applied all sorts of corporate viability tests before authorizing the use of an enterprise-grade product. From time to time, this has meant the exclusion of products from relatively large companies like Apple Computer and Sybase. Of course, these exclusions often had nothing to do with the number of bugs in the products. They had everything to do with declining market share or an unfavorable buzz around the companies in question.

So, why shouldn't we develop and apply similar viability tests to the OpenSource projects whose products we adopt? Here are a couple of good metrics: deployed copies of the project and diversity of the development group. Those are two of the reasons that we suggested that one client adopt Slashcode instead of Everything.

October 2, 2000

Electronic Toll Collection Finally a Reality on the NJ Turnpike

After a seemingly endless delay, the New Jersey Turnpike has implemented the electronic toll collection system called E-ZPass. This means that it is possible to travel from Boston to Buffalo and Albany to Wilmington along toll roads without stopping to pay cash.

Some of us who put E-ZPass on our cars two or three years ago never thought that we would live to see the New Jersey Turnpike and the Garden State Parkway fully implement electronic toll collection. Although we would prefer if tolls were totally removed from these roads, the implementation of E-ZPass is the next best thing. Motorists in the Northeast should rejoice in the same fashion as when our states finally accepted the 65 mile per hour speed limit.

Jakob Nielsen: End User Content Creation Key to Evolution of Internet

A lot of companies building Content Management tools have identified so-called P2P (peer to peer) tools as the next area of rapid development. Jakob Nielsen has touched upon that in his latest Alertbox. But, he has done so in the context of a larger question: How can we increase the number of people who contribute content to the Web?

Nielsen believes that a larger percentage of Internet users need to create content in order to maintain the current growth rate. However, a number of difficulties will need to be overcome in order to make this happen.

Jakob has gotten a lot of feedback on this article, both submitted directly to his Web Site and posted to other Web Sites like Scripting News. Read on for our comments.

Some of the comments that have been offered miss Jakob's point. Dave Winer's comment, ("What does the software do for you and how does it work, how accessible is it, and does it give you control?") is a great example. Users at the level that Nielsen is talking about are interested in services, not products.

This is why we are interested in discussion about and analysis of any application that can be used as a Content Management platform, whether we like the tool or not. But our primarily interest is in the deployment of available services on individual Web Sites, not the mere availability of a feature within the chosen framework.

When we look for new features to deploy to one of our Web Sites, we look at other Web Sites from the user's perspective. We hardly ever look at the underlying publishing application from the editor's perspective. Why should we? Products like Interwoven and Vignette are frameworks, not applications.

A lot of corporate applications developed on top of these platforms look like Client-Server applications, and not like best-of-breed Web Content Management Systems. So, the palette of features available inside these frameworks is not truly significant unless:

  1. these services are exposed to the user in highly usable ways
  2. the services needed by editors are deployed:

    1. rapidly,
    2. in ways that allow the underlying Web Site to scale, and
    3. in ways that enhance professional authors' and editors' productivity.