# Internet commentary

ISSN: 0305-6120

Article publication date: 1 June 2000

220

## Citation

Ellis, B. (2000), "Internet commentary", Circuit World, Vol. 26 No. 2. https://doi.org/10.1108/cw.2000.21726baa.001

## Publisher

:

Emerald Group Publishing Limited

## Internet commentary

He maketh his sun to rise on the evil (The Gospel According to Saint Matthew, 5:45).

Keywords: Internet, Sun Microsystems, IPC

What I am about to say for my preamble is slightly "off-topic" but not entirely so. Most of us, when confronted with the famous "blue screen of death" or any other problem causing any of the Windows systems to crash have indulged in a little mental Microsoft-bashing. This is a very natural and easy reaction. You have only to pick up a PC-oriented magazine to see just how unreliable even the high end Windows NT-4 is when it is used intensively. However, the problem is certainly exacerbated when using some of the very widely distributed software, such as Internet Explorer or the MS Office Suite. I may be wrong, but my impression of this type of software is that it has been designed so that it functions well for most of the time but it hates being pushed to the limit. Could it be that the designers have instructions not to waste time on very costly perfection? I don't know. But the Microsoft share of this market is stupendous, especially if you consider that other software, such as the Netscape Communicator and the Corel WordPerfect Suite, both direct competitors, are more performing, more reliable and, in my opinion, generally better - and cheaper, to boot. I will not go into the reasons for this apparent paradox: the US Justice Department is fully aware of how it has come about.

However, there are vague signs that the Microsoft Empire is beginning to teeter slightly. Some years ago, the Linux operating system was developed. This is a free system, based on Unix, which anyone can download. I have not used it yet, but it will certainly happen within a few months. This has a hard core of aficionados, representing perhaps up to 5 percent of the operating system market, including those persons using both Linux and any Windows flavour. The great obstacle to the development of Linux has, in the past, been the lack of availability of software and hardware drivers. Both these problems are being more seriously addressed by various manufacturers, now, and some of the big names are beginning to act accordingly. For two or three years, the Linux developers have even been developing a Windows-like interface for Linux, called WINE, which would allow the use of Windows software on the system. This is available in development form but I believe that it still has some weaknesses, inherited from Windows itself. However, the most telling proof of Microsoft being worried is that they have just put up a Web page "proving" that Linux is inferior to Windows 2000, which has not even been released at the time of writing and is not likely to be for another few months. Implicitly, it is an admission that Linux is better than the current Windows systems, such as 98 and NT-4, because it is only compared with what is vapourware for the corporate world, with the exception of a few beta testers. The very fact that Microsoft stoops to publishing such a document must be a measure of the worry that the operating system market may slip out of their hands. I have little doubt that, if the Linux developers wished to riposte in a childish way, they could make up an equally convincing list "proving" that it is better than Windows 2000.

However, if I were William Jefferson Gates Junior, it is not specifically Linux I would be worried about, at least directly, it would be Sun Microsystems. They have just released a new office suite comprising all the features that can be found in other ones, a sophisticated graphics editor and a Web browser/developer system. When I say "new" office suite, this is not strictly true, as it is already in version 5.1. However, if I understand the situation correctly, it is a German development which has recently become an international contender with a tie-up with Sun. I have not yet tried it - but I shall. Why would a comparative newcomer in this market worry Bill Gates when he already has three-quarters of it in his power? Well, there are four reasons:

1. 1.

Possibly the least important, but the most striking, is that it can be downloaded free-of-charge (c. 60Mb, if you please!) from the Internet, purchased on a CD-ROM for under $10 or purchased in a shrink-wrapped box with instruction manuals for under$40.

2. 2.

It is compatible with all existing files (e.g. Word, Ami Pro, WordPerfect, Excel, Quattro Pro, Lotus 1-2-3, presentation softwares, databases from DB-3 through to Access, all the usual graphics formats etc.)

3. 3.

It has been developed with maximum reliability in mind, as well as maximum "user-friendliness".

4. 4.

The clincher is that it is totally cross-platform and the CD-ROM contains the suite for all the common operating systems, such as Windows in all flavours, Linux, OS/2, Sparc (Solaris/Solaris X86) and so on.

The latter feature is what must make Bill Gates start worrying a little. I can just imagine the following scenario:

John (to take a name at random) is a Word user, let us say as a professional technical author making up instruction manuals on a stand-alone machine running under Windows 98. On average, he must reboot twice a day, which is quite typical for the kind of work he is doing, producing heavily formatted documents with spreadsheet tables, photographs and line drawings. One day, in exasperation, he decides to try the Sun StarOffice (yes, this is what it is called) and finds that its new concept of a total suite in a single software suits his work better than the disparate parts of the MS Office suite plus a separate graphics suite, desk-top publishing and Web browser. To boot, it crashes less frequently but, because it runs under Windows 98, it will still do so. The next logical step would be for John to think, "Ah! I've heard that Linux practically never crashes: I have everything I need on the StarOffice CD-ROM to run it under Linux". So John would go to his nearest computer dealer and buy, say, the Red Hat version of Linux for a few tens of dollars or euros. Suddenly, he will find himself running everything he needs under Linux without any problems and without any real hassle because all his files are intercompatible. At first, he will probably have Linux and Windows on a dual-booting system but, as he gains confidence, Windows will be thrown out of the window. Microsoft will have lost a customer on two counts.

This fictitious scenario will depend on many factors, if it is to become reality, not the least of which is that StarOffice must be what Sun Microsystems promises. Now, Sun has a reputation as being a good, solid, serious company with top-quality Sparc workstations and, above all, Java. I have little doubt that StarOffice is the first serious major application which has been developed entirely in Java, already known to be less quirky than the other high-level languages, which are used for the development of the major part of the competitive software. I would therefore think that, in the medium term, Linux/StarOffice may now become a strong contender as a replacement for Windows/Office/Internet Explorer/FrontPage as the "standard" software set-up, first for small corporate requirements and then drifting into the larger ones, possibly as Unix/StarOffice in some cases. Because it is cross-platform, this makes it particularly powerful for networking over different types of computer. The big problem, as I see it, is in support. How can a company with no significant income from a product give adequate technical support on it? At the moment, this is so new that the answer is not yet apparent, other than that there are Internet Newsgroups. But Sun must address this problem very seriously if they wish to penetrate the open software market. Whatever they do, it could not be worse than some of the big boys in the software game, anyway!

Now what has all this to do with a Circuit World Internet commentary? More than is immediately apparent, I suspect. In the first place, the Sun suite incorporates a Web browser as an integral part of the software. Remember, this is an integral part of a single software, not a separate software as is the case with the Microsoft offering. Because MS Internet Explorer is not a very robust software (already in its fifth major moulding in less than as many years!), the tendency of a StarOffice user will almost certainly be to use it as his default browser, especially if he is not working under Windows. Rumour has it that it may contain similar features to Netscape Communicator, already a better Internet suite than MSIE plus Outlook, in my opinion. However, it may be that the new software could also be a better medium for Web site development. Let me explain what I mean by taking the example of Word 97, because it is still more widespread than Word 2000. The generation of a Web page from a Word document is done by passing the DOC file through a filter which analyses all the tags, one after the other. Where there is an equivalent tag or series of tags in Hypertext Markup Language, it replaces it; otherwise it deletes it. The result is that a moderately complex Word document becomes a very tag-heavy approximation in HTML and certainly not good Web page generation. Many users may not notice the difference but, when I see so-called professional sites generated entirely with Word or any other word processor doing a similar operation, I shudder. This is one reason why I like Microsoft FrontPage (you see, I am not just on a Microsoft-bashing binge). It strips away unnecessary tags from the generated pages, as well as being an excellent WYSIWYG HTML generator in its own right. As I mentioned, I have not yet tried StarOffice but, because it is a single integral software, I feel that its Web page generation is probably custom designed to do its job well, rather than an add-on filter converter. If this is so, then I think we may see a new generation of better, more professional-looking, home-made pages. This remains to be confirmed, so I'll reserve judgement on this for a future issue of this journal. However, being a Java software, the system has a potential for automatic JavaScript and Java applet generation, opening a whole new ball-park of Web site generation.

Have I whetted your appetite sufficiently that you wish to know a little more? You may find more details or download it at http://www.sun.com/staroffice Be sure to download the White Paper which is available on this site. It is a 22-page document, dated September 1999. If anyone tries this software, I would welcome your views to meld into my own as I explore it. And, by the way, I am not using Word to generate this article, but WordPerfect under Windows 98!

http://www.ipc.org

OK, so let us get down to business. Regular readers will know of the problem I mentioned in Circuit World, Vol. 25 No. 4 (August 1999). As I wrote the text for my Internet commentary, the IPC site I was criticising changed and left me in a hole on my deadline date. It is now six months since this happened and any teething problems should have settled, so I am taking this opportunity to re-review the IPC site as if I had never done it before. For a view of the Home Page, please refer to p. 43 of the August issue.

So how do I comment on such a complex site? Quite frankly, I don't really know. I could fill the whole of this journal, advertising space included, if I were to do a thorough job. I therefore propose just mentioning a few highlights which I consider good. One thing I particularly like is the fact that the IPC staff is listed, each person with their job function and e-mail address. It is a pity that many other companies and organisations do not do the same. If I send an e-mail to support@joe_bloggs.com, I can never know whether it reaches the desk of the appropriate person, unless I receive an answer (which often tells me it has not!). On the other hand, unless I missed out on something, they do not publish a list of members. This has the advantage that there could be no general spamming generated from such a list. What they do list are links to PCB fabricators, assemblers and designers: these are in the form of a link to some of the members' Web sites with a very brief description of each one's specialities.

Possibly the most important activity of the IPC is in standardisation. A complete catalogue of published standards is available, with prices for members and non-members in hard copy and, in some cases, electronic formats. What I do not understand is why the electronic versions are considerably more expensive than the hard copy ones. I would have thought that a CD-ROM or a couple of diskettes would be cheaper to produce than a printed and bound book (if this were not the case, why do software manufacturers not provide us with decent instruction manuals?). In addition, there are a few documents available for free download, including the entire Test Methods Manual in an economical method-by-method format (see Figure 1). (For the uninitiated, the complete document must weigh about 10kg on lightweight paper, so this free method is fine for individual or updating documents but hardly worthwhile for the complete manual, the downloading time for which must approach the best part of a month!)

Figure 1 A typical download PDF file from the Test Methods Manual

The next section gives a complete rundown of current and future events, in the form of a monthly calendar. This is followed by a so-called catalogue and bookstore which simply shunts one back to the standards section. Then there is a section on training videos and CDs. These seem, at first sight, to be horrendously expensive at, say, \$10 per minute of running time, but they can hardly be assimilated to the price of a BBC video of a Miss Marples mystery, which is probably produced in editions in the 100,000s, against 100s for an IPC video. The following section merits a few minutes' reflection, being devoted to market statistics and research. Some of the data, especially on bare board fabrication, are freely available, but they mostly have to be bought. For the assemblers, there is a special council which handles these matters, membership of which is not included in the IPC subscription.

There are also details of various training and certification programmes. This is probably an important aspect of the IPC's work, in a field where internationalisation has been sadly lacking. Such programmes exist for bare board fabrication, assembly, soldering, design etc. As the actual training is done by private organisations (design excepted), mainly in Europe and the USA, under contract, prices may vary and surfers are linked to them to be able to enquire about course dates, prices and other details.

As may be expected, the section on regulatory and legislative affairs is essentially valid only for US members. It is perhaps a fact that large-scale industry lobbying, such as occurs in the USA, has not been the custom in Europe. On the whole, this is possibly, in my opinion, a good thing for the public economy but, when confronted with new European Directives, it may become necessary to avoid catastrophic technocratic decisions. The impending WEEE proposals on lead in solder is a case in point: without lobbying, it could kill the wellbeing of our industry throughout the world, not just in Europe. Brussels is slowly filling up with lobbying offices and organisations. Maybe our industry needs - unfortunately - to be better represented there, È l'américain? This personal digression has, of course, nothing to do with the IPC.