Letters

Letter Editor

Issue #100, August 2002

Readers Sound Off

Eyes Opened—Looking for Lurkers

I'd like to thank Charles Curley and (obviously) Linux Journal for writing the “Emacs: the Free Software IDE” article in the June 2002 issue. I've been using XEmacs for four years for my programming projects, and Charles' article opened my eyes to a bunch of options I wasn't even aware of. I guess I'll be spending the next couple of weeks (or months) going through the Emacs documentation to see what else is lurking in the 20MB of source code. Thank you LJ for publishing such an informative (and well-written) article.

—Robert James Kaes

Humor in Advertising

It's great that you can afford to spend an entire page on humor, even though it's not the April issue. And it's even that dry, sarcastic humor that you usually find only in British publications. I'm referring, of course, to the Compaq ad parody on page 51 of your June 2002 issue, where they brag about being such an early adopter of Linux when “their” employee, a Mr. Jon “maddog” Hall, ported Linux to “their” Alpha system. And how “they” funded and otherwise supported Linus as early as 1994.

As a matter of fact, wasn't Mr. Hall employed by DEC in 1994, when he ported Linux and helped to get funding for Linus to support the Alpha architecture? In 1994, wasn't Compaq, as well as Dell and all the other major OEMs, claiming that no one wanted a Linux system?

I realize that Compaq now owns all the assets of DEC and that marketing people have no shame, but if this ad is believable, perhaps I should be able to buy a Picasso and then claim that I painted it.

—Bill Peterson

Jon “maddog” Hall replies: Linux Journal was kind enough to allow me to answer your Letter to the Editor.

No one felt worse about the demise of Digital Equipment Corporation and its purchase by Compaq than I did. I had used DEC products since 1969, and it was on a PDP-8 through the use of a DEC training manual that I taught myself assembly language programming. When I was the Department Head of Data Processing at Hartford State Technical College, DEC repaired my PDP-11/70 for free because it had taken them so long to find the problem, and the school had no budget that year for the “time and materials” it would have taken to fix it. Later I learned UNIX and was system administrator to six VAX 11/780 machines at Bell Laboratories.

Through the years I got to know a lot about DEC culture and its commitment to its customers. As part of this commitment I was involved with DECUS, and from the DECUS members (Kurt Reisler, in particular) learned about Linux. Digital Equipment Corporation was more than a company, it was a family.

Yes, I was working for Digital at the time that the Alpha Project was started. Yes, it was several years before Compaq bought Digital. Yes, I felt a pang as Compaq took over Digital's role as the “first system vendor to join Linux International” (even before VA).

On the other hand, to deny Compaq the right to that claim denies the work that the people who still are employed by Compaq (or should I say HP?) did at that time. I know that Maurice Marks, who basically funded the Alpha Linux work in the first days is still at the “new HP”. Jay Estabrook (Hi Jay!) who did a lot of the low-level porting work is also still there, helping to continue support of the Alpha with Linux. The Alpha high-performance group, who started some of the first commercially available Beowulf clusters is still working away. The Digital compiler group, who ported their compiler technology to Alpha Linux is still (to my knowledge) working away. There were many, many people besides me who contributed their time and expertise, paid and unpaid, employee and customer, to the Alpha Linux Project. They did it because of the love of Linux and the love of Digital. They did it both with the support of their upper management and sometimes in spite of it.

To deny them their role in the support of Linux just because their company now has a new name is not fair either.

And if Compaq does not claim the right to say that they supported Linux first, who does? Digital Equipment Corporation is no more, and all the wishing to bring it back will not make a bit of difference. Soon I will expect to see Hewlett-Packard take up the mantle. And now, to set the record straight, a lot of other companies had people supporting Linux early on. Sun supported David Miller to do Linux porting work to Sparc. Compaq had people in Houston who helped to write device drivers and do porting. I know there were people at IBM and HP who were also doing Linux work. But for one reason or another the cultures of their companies did not allow them to have the visibility that Digital's culture allowed me.

In 1998, when a lot of companies held up their hands to say “they supported Linux” a lot of people were able to come “out of the closet”, but that does not mean those people were not active before. I had just come out a little sooner, and with a little more fanfare. Then again, I have never been a shrinking violet.

The world of computer company buyouts and mergers creates little tricks in time and space, and we should learn to live with them. A friend of mine, David Mosberger, who did a lot of work on Alpha Linux, recently wanted to return an Alpha system lent to him by Digital in 1995. But he did not know who to return it to, since Digital no longer existed. I told him to just hold on to it, and that things will be made right again. You see, for the last several years David has worked for Hewlett-Packard. His system is back home again.

The bottom line of this is that if I were writing a history book, I would mention the contributions of Wang, Prime and a host of other defunct computer companies like Digital Equipment Corporation. But I live in the here and now, so the current company that supported Linux in 1994 is Hewlett (née Compaq, née DEC) Packard.

Trivial Correction

Regarding the June 2002 issue, page 8, trivia question 1—please lay this myth to rest. Yes, Grace Hopper did find a moth in a relay. That log page is now in the Smithsonian Institute, I believe. But “bug” in the current sense has been around since Thomas Edison's days. See www.byte.com/art/9404/sec15/art1.htm and/or www.catb.org/~esr/jargon/html/entry/bug.html.

—Felix Finch

Suggestions Welcome?

Editor's note: The following Letter to the Editor came to us written by hand on a sheet of yellow legal pad paper.

From the Desks of Adrian and Mike: We understand that yours is a magazine that would like to appeal to Linux enthusiasts. Therefore, we have some suggestions regarding your fluffy layout:

  1. Use fixed-sized, monospaced font throughout the entire publication. All pictures must be represented as ASCII art or PostScript files, and there must be no color. You've got black; you've got white. What more do you need?

  2. The cover should not contain anything but the title and the date.

  3. The binding should be staples and not glue, which is bad for the environment and embarrassingly corporate.

  4. All type must be printed with high-impact printers to give each page a unique and profound texture.

  5. You should change the name of your magazine from Linux Journal to Linux Kernel and, under no circumstances, write about anything that is not a part of or that cannot be directly compiled into the kernel.

  6. There is no need to use English for all the articles. C and Bourne shell scripts are languages much more recognized by the global Linux community.

  7. We could go on for pages but feel strongly that if you just covered the first 6, just a half a dozen honest, down-to-earth and from-the-heart suggestions, your circulation would multiply by a factor of ten (we know we would by ten issues each month just cause they were that cool); your revenues would skyrocket, never mind the inherent irony about the relationship between your capitalist intentions and the brilliant and revolutionary open-source model that Linux embraces, and, not to mention, we'd buy you beer.

—Adrian and Mike

Social Destruction

On a whim I was reading Harriet Beecher Stowe's Uncle Tom's Cabin today. I was struck by how, at the time, abolitionists were considered socially destructive extremists—and the South's reflexive reaction was censorship. The essence of freedom is freedom of thought and creativity, bounded by essential (not submissive!) respect for others. I believe that, perhaps in 50 years, perhaps 2,000 years, but with the same certainty that humanity will survive, intellectual property will come to be regarded with the same revulsion and embarrassment as human beings as property.

—Stephen Schaefer

Protocol Problems

The Linux for Suits column, “The Protocol Problem”, by Doc Searls in the July 2002 issue of Linux Journal, was of interest to me because it is an issue that I think we are seeing several major issues start to develop around. There seems to be an overwhelming amount of “If all you have is a hammer, everything looks like a nail” mentality going around, at least insofar as trying to put nearly the entire world on top of IP—this is only making, what Doc Searls referred to as “the gating factor”, worse.

As technology develops, the lower layers of the technology become more and more abstracted and hidden from the upper layers. Right now, it is doubtful if many people care or concern themselves with whether their IP packets ride over fiber or copper, the particular framing or link-level protocols being used, etc. Because we can connect fiber and copper and wireless networks together, we are not limited by the individual mediums.

Eventually, this is what is going to have to happen with IP and any other protocols occupying this space—the “end-to-end (network) protocol” mentality must give way to the “end-to-end application” mentality. One day, eventually, the IP network will be replaced by something else. It's going to happen—but the general mentality seems to be that an IP network must be IP end to end. Building for that, designing for that, that attitude will slow down the deployment of any future protocol technology...like IPv6.

One day, just as most people do not care whether their data are going end-to-end over a particular physical media (as long as the desired quality of service parameters are met), they will not care about the network protocol. When data networks are not an “IP-only” club where only IP through-and-through networks can play, but a true inter-net (as opposed to the current IP Internet, which is more of a giant shared-address space Intranet) where the intra-net protocols may all be different, then the barriers will be gone. Organizations will be freer to develop and deploy new protocols and technologies internally without getting ostracized by their peers for being different.

Erratum

In my interview in the June 2002 issue, my affiliation was mistakenly listed as “Director of PythonLabs” rather than the correct “Director of PythonLabs at Zope Corporation”.

—Guido van Rossum