Letters to the editor: Trusting open source

Is the IEEE standards process broken?; Can companies trust their networks to open source

Standard treatment

Regarding “Is the IEEE standards process broken?”: As past marketing chairman of the now-defunct HomeRF Working Group, which eventually lost in the market to IEEE 802.11, I offer a perspective that addresses the IEEE political and committee-size problems. The root cause of the problems is how the IEEE handles voting. To encourage early and active participation, an IEEE policy was established to allow one vote per meeting attendee, as long as the person was an IEEE member and attended the previous two meetings, or something like that.

When HomeRF was still active, I heard complaints from members involved in both wireless technologies, saying the IEEE 802.11 chairman and co-chair wielded too much power with their ability to select meeting locations and voting schedules. One story described a meeting in Hawaii that was relatively expensive to attend and difficult for members to justify with their management. Those who did attend reported busloads of attendees from the companies of the chairman and co-chair. You guessed it: Two meetings later a critical vote was held and companies that didn’t send their own busloads were disadvantaged.

A policy allowing only one vote per company would prevent such abuse. The HomeRF Working Group and Bluetooth Special Interest Group both reserved voting rights to the sponsoring companies that provided most of the funding and did most of the heavy-lifting design work. Those organizations had their own political problems, but that’s another story.

A brief history of HomeRF and why it failed in the market is posted here.

Wayne Caswell

Austin, Texas

Regarding “Is the IEEE standards process broken?”: As an observer and former participant of several standards organizations, I feel a major problem is that too many vendors and manufacturer are seated in these committees. These corporate representatives and their (technical) supporters have considerable financial stakes and the temptation to mold the process in their favor is too great.

Unless vendors’ and manufacturers’ powers are curbed, and curbed effectively, in IEEE as well as in other standards societies, this problem will be repeated over and over again. Some 20 year ago I saw such a situation occur in the ISA’s standards committee on industrial field bus standards.

Vendor and manufacturer interest is understandable, and their as well as third party (consumers/users) input is desirable to get well-designed and thought-out devices or systems. But when these individuals want to ram through features and requirements to gain economic advantage (read: greed), or to capture a larger market share, by perverting the normal process, use threats of litigation, then the standards committee has to take some strong and rational actions.

IEEE, as a professional and technical society, needs to decide what its primary mission is. IEEE must maintain a level playing field, develop fair and technically feasible standards for all; and that includes the suppliers, consumers and users of technology. These standards must be based on sound principals, i.e. standards which would provide quality products or systems, which are cost effective, and technically sound because the manufacturers and vendors did follow the standards.

IEEE must maintain its primary focus of technology, and user/consumer interest in the standards process. The various standards committee(s) participants must be ethical and bound by standards of IEEE and not of their corporate financial interest.

IEEE must ensure that specific methodologies or specific features must not become part of a standard to the detriment of the society, a detriment to other competing products, and in some cases, used by one party for crushing other reasonable standard requirement because this major manufacturer/vendor finds the features or requirement inconvenient. The IEEE standards rules and process must stop these corporations and/or individuals from using such tactics.

Perhaps the time has come for IEEE, and other standards organizations, to consider that it would be reasonable to limit the participation of corporations, manufacturers and vendors to certain areas of the standards process only. The corporate or financial stake holders should be excluded from certain aspects of the standards process because it will be not possible for such participants to be fair and reasonable due to their employment or association with certain clients; and these participants would abuse or pervert the standards process.

B. Mahmood

Via e-mail

If you are not going to have vendors and manufacturers, then who will you have? Academics who have no idea how to implement anything in the real world, and more "individual contributors" who have no clue how their ideas can be realized by industry? At the end of the day, all standards work that is worth doing is paid for by products and services that result from its creation. Standards development is just one essential part of the development processes of the industry that supports it.

With more than 30 years of experience of product R&D and national and international standards work, my major observation on the IEEE process is that it needs two major changes: per-organization voting, as others have alluded to, and the ability to move on. Aside from that, if the IEEE process is well run, with attendees who have a real interest in seeing the standard and resulting technology out the door, it works. The work of 802.16 under Roger Marks is a good example of that, completing a mobile broadband standard while .20 ground to a halt in the mess as described.

Having said that, it should be noted that IEEE has major limits when it comes to developing complete suites of standards (e.g. mobility networks). IEEE works OK where the end product is sold in a "retail" manner like 802.11, but it is not at all well suited to developing standards for WANs operating in licensed spectrum, as its current processes do not allow sufficient influence for the end customer, in this case the wireless carriers. So the network development is done elsewhere, resulting in a split of responsibilities which may slow development, as a result of which the IEEE's standards may not be as widely adopted as those from, for example, 3GPP & 3GPP2, which develop as much as the network as is practical under one roof.

Actually, the best standards I have seen were developed offline by small groups of organizations, big and small, that have a common interest in promoting a specific technology or service. Eventually their pre-standard product will be brought to an SDO for a sanity check and publication, but the hard work is already done in a quick process where many of the limitations of the SDOs’ processes are laid aside by the participants.

The IEEE standards folks just need to stand back a bit and look at the whole world of technical standards development, and how they can best fit into it in the 21st century.

James Harvey

Via e-mail

Trusting open source

Regarding your “Face-off” on whether companies can trust their networks to open source: It's funny, there is so much irony in this debate. Ipswitch started out as a shareware company -- remember WS_FTP Lite? FileZilla and other free FTP clients probably have killed Ipswitch’s FTP client business.

Ipswitch’s WhatsUp started out as a low-cost shareware product as well. When the company realized it could make more money selling a professional product, the lower-end product was end-of-lifed. Now open source threatens Ipswitch’s cash-cow product once again. WhatsUp Gold is a great product, but don't expect everyone to pay $2,000 for it when there's an open source solution available that's almost as good.

Paul Misner

Columbia, Md.

I am grateful that in his column, Ipswitch’s Roger Greene notes that open source software is not free, something which the open source and free software communities strive to remind people. In any environment, the cost of software must include installation, configuration and maintenance, and depending on the particular software product in question and your environment, this can run from insignificant to many times the purchase cost of the software. This is true of both open source and commercial software. Similarly, the responsiveness of the vendor to bug reports can range from poor to great irrespective of the commercial/open nature of the product -- the only real difference is that in open source software if you are desperate enough, you can fix it yourself.

In evaluating software products, purchase price (or lack thereof) should be low on the list of criteria. Just as I would assume Mr. Greene does not automatically purchase the product with the highest price on the assumption that it must follow that it has the highest quality, he should not dismiss an open source product simply because there is no purchase price attached.

Tom Payerle

Department of Physics

University of Maryland

College Park, Md.

Learn more about this topic

Join the Network World communities on Facebook and LinkedIn to comment on topics that are top of mind.

Copyright © 2007 IDG Communications, Inc.