7th > December > 1998 Archive

The Register breaking news

Intel’s Katmai no humble pigeon pie

Intel will today brief journalists on the progress of its Katmai MMX processors but not us because we daren’t sign non disclosure agreements (NDAs). The chip giant is now expected to roll out the first of its Katmai babies in February and will move to .18 micron technology through the whole of 1999. But not before it introduces faster Celerons and mobile Celerons in early January, as first revealed here in The Register We love a challenge here at the anti-Satanic offices. Look out for further details during the day. ®
The Register breaking news

A year ago: Intel plans NC standard

A couple of years ago Intel and Oracle allied with the intention of accelerating development of Internet video, the general idea as expressed at the time by Andy Grove being that information was information, so if you were going to be looking for it via the Internet you'd be looking for video as well as words. The two companies intended to produce a sort of test service as a shop window for ISPs (we weren't getting into provisioning, no sir…), but it never happened. Oracle went off to push Network Computers instead, Intel was conspicuously not one of the NC backers, and they effectively wound up in separate camps - until last week, when Intel emerged as leader of an effort to define a standard for x86-based NCs. The Intel move has been a long time coming. The company provided support for Oracle and its NC arm, Network Computer Inc (NCI) towards the beginning of this year, but hadn't said anything official prior to the announcement of the latest effort last week. Intel has been working furiously to 'prove' that Intel chips run Java best, and NCI is entirely sold on the idea of Intel being the best base for business NCs (remember the first prototypes were ARM-based), so having Intel define the standard is a logical way forward. But what, precisely, is it that Intel is defining? You'll recall that Apple, IBM, NetScape, Oracle and Sun were devising an architecturally independent NC reference profile back in 1996, but you probably won't recall anybody announcing that they had in fact completed it (due date July 96, ten on the vapour scale). Several companies including NCI do however already have operational Intel-based NCs, so we might suspect that what Intel is actually going to be doing is producing a reference design that is not processor independent in the way that was originally intended. This isn't necessarily sinister – not entirely, anyway – because in order to make high volumes and a low price point NCs will need Intel's special skills in chipset and (no doubt) board design. Significantly, Intel says the design will require an embedded Pentium or better, so you can view that as a signal that Intel finally intends to turn its work on embedded Pentiums into mass-market sales. The other thing Intel is going to be defining is also significant – the server end. Previous efforts at NC definition didn't cover this, leaving IBM, Oracle and Sun to do their own things with their own systems, but that left a rather large hole where the Intel space should be, and obviously seriously impeded NC penetration of the PC OEM business. Intel will produce a spec based on Pentium II or Pro, and will cater for specific remote execution issues such as processor load, memory, I/O and storage. This will make it simpler for OEMs from Compaq down to put out Intel-based equivalent's of NCI's network in a box. Intel has a good number of supporters on board, not as long a list as the original NC endorsers but certainly a more credible one: Microsoft, IBM, Citrix, NCI, Novell and SCO for software; Compaq, Hewlett Packard, IBM, Fujitsu, NEC, Network Computing Devices, Packard Bell-NEC, Siemens-Nixdorf, Unisys and Wyse for hardware. Note the interesting tensions in the software list. Citrix's presence puts the initiative into the Win32 mainstream, while IBM's on the software front brings WorkSpace-on-Demand into the equation. Microsoft's contribution, meanwhile, is CE. Microsoft is in an interesting position here, as although (along with Intel) it originally dismissed NCs, it seems inescapable that what Intel is working on will turn out to look pretty much like a Windows Terminal, so not supporting the latest effort would be silly. But note the key differences between the Intel NC spec and the Intel-Microsoft NetPC spec. The latter requires Microsoft software and network management systems, whereas the former will probably (it's not actually out until next year) define the hardware for management, but will leave the software side open for a range of different approaches. Recent intelligence we've heard on the first beta of Windows NT 5.0, which is the Microsoft OS intended to make the whole of the NetPC show roll, suggests that the situation is not good, probably meaning more delays, and possibly allowing some of Intel's new buddies to grab the ball back, and get an NC-based alternative established. Intel's position is meanwhile equally interesting. Its involvement means it accepts that NCs will take off to some extent, but success for remote computing of any description, NC or NetPC, will inevitably cannibalise its desktop sales, and force it into lower margin, lower spec chip sales. We think, rather than retreating to the server, Intel is hoping vast new consumer markets will open up, but nevertheless it will be a good trick if the company manages to achieve a dominant position there too. ®
John Lettice, 07 Dec 1998
Intel headquarters, Santa Clara

Katmai pricing will mean PII dominoes fall

Intel will release its 450MHz and 500MHz Slot One Katmai chips at the end of February next year and is pricing them at $528 and $760 respectively. But those prices are set to drop to $445 and $675 respectively on the 11 April 1999.
Mike Magee, 07 Dec 1998
The Register breaking news

Intel builds chip to beat the Bomb

Intel is to announce the first nuclear holocaust proof chip tomorrow, according to US reports. The breakthrough, reported in today's issue of Defence Week, could ensure the Great Satan's final triumph over its rivals - albeit at some considerable cost in collateral damage to chip customers. The new chip has reportedly been developed in conjunction with the US government's Sandia National Labs in New Mexico, and will provide the basis for radiation-hardened computers. The possibility of a nuclear conflagration zapping all the chips in their computers has been a long-term worry for governments everywhere. Ordinary PC buyers and electorates in general tend to have other concerns, but they needn't panic immediately. Intel's announcement is scheduled to be made in conjunction with NASA, and the first radiation-proof computers will be deployed in space. This will protect US government satellite surveillance and communications systems against nuclear weapons detonated in space in order to knock them out. ®
John Lettice, 07 Dec 1998
The Register breaking news

Grid teams with Sprint for Brazil phone bid

The UK National Grid's plans to repeat its success with Energis look set to kick off in Brazil, with a joint venture with US giant Sprint. Energis, National Grid's telecoms arm, has been successfully floated off, but the outfit plans to repeat the formula on a larger scale, internationally. Brazil and Sprint look a promising combination to do this with. The Brazilian telecoms market is underdeveloped, but potentially massive. Sprint meanwhile has spent the last five years developing its Integrated On-Demand Network (ION) system, which is designed to be a combination broadband voice and data system covering the US. Sprint's expertise here could therefore dovetail well with an approach similar to the one National Grid took with Energis. The two companies are setting up a £600 million consortium to bid for construction of a national phone network in Brazil. Bids for this, and for four regional ones, are due in by the end of this week. ®
John Lettice, 07 Dec 1998
The Register breaking news

IBM PC martyrs meet untimely end

A story in this morning's Wall Street Journal said that two men were shot dead for smuggling IBM kit into mainland China. According to the newspaper, the men were summarily executed yesterday for smuggling nearly $7 million worth of kit into the country. The WSJ claimed that companies like Big Blue and Motorola "have long sold to offshore distributors" moving kit into Red China without paying the full duty. Now that IBM has decided it will sell direct, perhaps the lives of others will be spared… ®
Mike Magee, 07 Dec 1998

Compel paid £18 million to acquire Info'Products UK

Info'Products is paying Compel £18 million to take IP UK off its hands. The headline price is just £1 -- an indication of Info'Products' disastrous UK performance in the last few years, following the removal of IP UK founder Gerry Redman.
Drew Cullen, 07 Dec 1998
The Register breaking news

Coming soon – Bill Gates, the movie

Where have we heard this before? Bill Gates hasn't read it and doesn't know anything about it, and Microsoft spokespeople dismiss it as "pretty shallow and ultimately forgettable". No you're right, the Redmond spinmeisters haven't used those precise words about the DoJ's case, yet. This time they were talking to the Daily Telegraph's New York Diary column about a forthcoming made-for-TV nerd opera, Pirates of Silicon Valley. It's due for showing in May, and will star Dr John Carter from ER as Steve Jobs, and specialist geek actor Anthony Michael Hall as Gates. From what the Telegraph says, the film doesn't exactly break new ground. Jobs is portrayed as a maniac, and an acid-dropping one at that (we're sure that bit can't be true). Gates, meanwhile, expresses admiration for Napoleon at age 12, and races Porsches. It's not clear if he gets to 'liberate' and race earth-moving equipment in the film; we heard he used to do that before he took to moving APIs around instead. ®
John Lettice, 07 Dec 1998
The Register breaking news

IBM, AMD acquisition rumours take wing – again

Rumours that IBM and AMD are in negotiations over a merger have re-surfaced again following frantic share movements on Wall Street. For some weeks, investors have questioned whether IBM should split its shares which closed on Wall Street last Friday at $164 1/4, up three quarters on the day. Meanwhile, AMD's share price rose by nearly two dollars on Friday on rumours of the acquisition and stood at $31 3/8. Over four million AMD shares were traded that day. According to speculators, IBM has realised it must make a pre-emptive strike on the PC front, and with the AMD price rising so high, so fast, now would be the time to do so. It is not the first time such a rumour has been mooted. It is known that earlier on this year the two companies were in talks. (Story: IBM set to buy AMD, Gateway? also Yield rises at AMD, but so do rumours) No-one at AMD UK was prepared to comment on a possible acquisition. ®
Mike Magee, 07 Dec 1998
The Register breaking news

Java guru beats-off Microsoft attorney with ease

Sun's Dr James Gosling proved to be a highly credible witness last week, and outclassed Tom Burt, the Microsoft associate general counsel who cross-examined him. Burt didn't cross-examine Gosling on his written testimony, but engaged in a debate with the inventor of Java -- and lost. Only once did Burt question Gosling on his written testimony in nearly two days, and that was a request for an elaboration rather than a challenge to what Gosling had said. While Gosling was searching for his testimony at the bottom of a pile of documents, Judge Jackson pointedly remarked that "I have his testimony. It has been a long time since we have averted to it." It seemed clear that Burt had blown it by ignoring the testimony and trying to discuss other issues raised in marginally interesting emails, and in magazine articles. The only other reference that Burt made to the testimony was to remind Gosling that he had said in a footnote that he would not broadly address issues raised in Sun versus Microsoft. At one point, after Burt crowed at the performance of Microsoft's Java virtual machine, Gosling said: "Where we became significantly concerned about Microsoft was not in the fact that they were outperforming us, but in the process of re-engineering their VM, they were deviating from the specifications, which, you know, the court in San Jose has now found twice in our favour." Burt couldn't move on fast enough, and away from the testimony. At least the judge will accept Gosling's testimony -- and the signs in court were that he had read it. In it, Gosling set out the background to Java in a clear and concise manner, and without rancour. With no challenge from Burt about the reasons for Java development, and with the evidence that Microsoft was losing out in its case with Sun in San Jose, Microsoft's actions with respect to Java are likely to be viewed by the judge as illegal and anticompetitive. Some data in Gosling's testimony are worth mentioning: more than a thousand copies of the Java source code have been licensed for reference and evaluation, and more than a hundred corporations have executed commercial sourcecode licenses for the Java Application Environment. Gosling's opinion that "there are no significant efficiencies to be gained by the inclusion by Microsoft of the Web browser in the operating system that could not also be achieved if computer manufacturers or end users were permitted to install and remove web browsers themselves" is likely to stand as an expert view that Microsoft integrated IE in Windows 98 for anticompetitive reasons. Gosling also pointed out that Microsoft did the same for the JVM - made Microsoft's non-standard version mandatory and non-removable in Windows 98. The broad thrust of Gosling's testimony was to show that far from Microsoft not being allowed to innovate, Microsoft has been doing everything in its power to stop the most innovative technology of the decade by using its market power to foreclose distribution channels. Burt clearly lacks experience in cross-examination (or perhaps has not watched enough afternoon television re-runs of Perry Mason). His attempt at ring-fencing Java failed because of his ignorance in the face of Java's creator. A good example was when Burt attempted to suggest that cross-platform programs only support the lowest common denominator functionality. He shouldn't have brought this up, because it gave Gosling the chance to explain simply and effectively that Windows did not have, for example, anti-aliasing, so that circles drawn in Windows have jagged lines, whereas Java's anti-aliasing makes the lines smooth. Burt's desire to show that Java was slow was so strong that he took examples from the earliest versions of Java. Judge Jackson caught on to this immediately, so that when Burt produced an undated extract from a book, he asked for the date. Burt then went on to what he called "some more contemporary articles in the computer industry press" and came up with a June 1997 item - an article by Jesse Beerst of Ziff-Davis. The judge noted that he was interested in the ad in the left margin, but didn't think it was for Java. Burt forgot to ask Gosling a question, and moved on to another Microsoft-sympathetic article in PC Magazine US from April 1998, but that turned out to be a mistake too. Burt started reading extracts that suited his case, without asking questions, so David Boies for the DoJ objected. The judge upheld the objection, and Gosling proceeded to demolish Burt's inadequate probes. Burt shouldn't have brought up the fact that there were no fully-compliant version 1.1 JVMs for Windows 3.1, because it gave Gosling the chance to mention that "Windows 3.1 has some architectural problems... even Microsoft failed to do a decent implementation... Many people impaled themselves on that particular sword." Ouch. The documents and video-taped testimony that Burt chose for his non-testimony-based challenge of Gosling presented no challenge at all. They did reveal a one-liner from Sun CEO Scott McNealy in one of the exhibits. After McNealy had seen an early demo of Java from Gosling, he wrote an email that said: "Charge: kill HP, IBM, Microsoft and Apple all at once!" The court went into closed session twice to consider Sun evidence that was accepted under seal. It seemed that Microsoft must have enjoyed its fishing expedition when it served subpoenas on Sun to obtain possible evidence. In some very pointed instructions at the end of the session, the judge reminded Burt that "as in-house counsel, you and Mr. Neukom are both precluded by your ethical obligations from imparting the information which has been placed under seal... [this applies] through the corporate hierarchy... remind Mr. Neukom of that obligation". Whether key people at Microsoft already knew the content was not disclosed. Gosling's testimony during cross-examination was meticulously presented, at an appropriate level for the judge, though without condescension as was the case with Burt. At times, the judge questioned Gosling directly, in a way that indicated that he was interested and wanted both to understand the subject better and to get some information that Burt was not bringing out. Burt tried to introduce some testimony from the Sun-Microsoft case, but in a flash, Boies was up, objecting that Microsoft had applied successfully to have a deposition from the Caldera versus Microsoft case withheld, at least for the time being. Burt hastily withdrew his request to include testimony from Sun VP Bill Joy. Burt fearlessly asked Gosling if, by "the evil empire" he meant Microsoft. Gosling confirmed it. Burt then went on to explore the Sun slogan "write once, run anywhere", but seemed unable to appreciate that it wasn't just Microsoft that indulged in a bit of hype - and besides, the slogan was becoming increasingly true over time, although Burt favoured ancient examples. Burt chose the wrong person with whom to pick a fight with on this particular issue, and was stopped by the judge: "You're arguing with the witness now. Put your next question." Sometimes Burt had to curtail Gosling's evidence, because it became rather damaging for Microsoft. In talking about binary compatibility, Gosling went on to say that "there are certainly, for instance, Windows 3.1 programs that do not work on Windows 95... (Burt's interruption was ignored by Gosling) despite extensive testing and diligence on Microsoft's part." Burt again tried to use documents that did not have the original date on them, but didn't get away with it. By this stage, Gosling had the initiative, and Burt's feeder questions tended to underline Gosling's evidence, rather than challenge it. Burt made little headway in trying to suggest that Corel's effort with a Java version of its office suite (using JDK 1.1) proved that Java was a failure. Gosling showed that the problem was in part that it was written before the JDK performance improvements were introduced. So far as the Netscape Java browser was concerned, Gosling thought it too big an engineering project for Netscape after Microsoft had made IE free and so cut off Netscape's revenue possibility. Burt began to repeat some arguments he had made earlier, and found himself being warned by the judge against "cumulative testimony". Again Microsoft tried to establish that a competitor had attempted a market carve-up against Microsoft. Burt suggested that Sun had agreed with Netscape that it would not to enter the desktop browser market with HotJava. Microsoft's evidence was not very convincing, but even if it had been, it was Microsoft that was on trial, not Sun, and excuses that "everybody does it" were an implicit admission by Microsoft of anticompetitive acts. The efforts of Burt to elaborate on a supposed Sun-Netscape agreement came to naught when Gosling said that Sun did not have a commercial browser product and therefore was not competing with anybody in that sector. At one point, Burt introduced an exhibit, which prompted the judge to tell him: "All I'm concerned with is the fact this is from somebody whom this witness does not know to somebody this witness doesn't know about matters as to which he has no knowledge, and I'm curious as to what use the document has other than to get evidence." Gosling's evidence will be resumed after evidence from Professor David Farber on Tuesday. The court will not be in session on Monday. ® Complete Register trial coverage
Graham Lea, 07 Dec 1998
The Register breaking news

Cyrix takes axe to high-end MII prices

Cyrix has slashed prices on its two high-end MII processors today as a battle-royal begins between it, AMD and Intel. Prices on its MII-333MHz part drops to $80/1000 from $90/1000, and on its MII-300MHz processor to $59/1000 from $67/1000. Its other two processors, the MII-233 and the MII-266 remain stable at $48 and $55 respectively. The reason for the price cuts on the higher end parts is to compete with Intel's Pentium II family. Last week, Intel announced it was phasing out its PII/300 and PII/266MHz processors, even though there is a scarcity of low end Pentium IIs in the marketplace. A representative of AMD said she knew of no plans to drop its microprocessor prices. ®
Mike Magee, 07 Dec 1998
The Register breaking news

Quantum spreads distribution wings wider…

Storage firm Quantum said it had appointed Summit, part of the Datrontech group, as a distributor for its products. Summit joins Arrow, Ideal Hardware, C2000, Ingram and RK Distribution in the UK. John Barnes, regional director of Quantum northern Europe, said that while the company is still big in desktops, at the high end it is gaining strength with its Viking and Atlas products. It had appointed Summit to help widen its presence in the UK. He did not say whether Quantum had now completed its distribution strategy. There is now a question over whether Quantum is over-distributed in the UK. One source at a Summit competitor said: "This is not untypical of Quantum". ®
A staffer, 07 Dec 1998
The Register breaking news

Digital River and Netsales branch out into UK

Warwickshire-based retailer Software Warehouse announced its deal with a US company to sell software over the Internet has so far proved a major success. The electronic software distribution (ESD) agreement, with Digital River -- which boasts the world's largest online database of software and other digital products -- has boosted Software Warehouse's online sales by 25 per cent in just four weeks, the company reported. Software Warehouse claims to offer prices that are on average 40 per cent cheaper than manufacturer's recommended retail prices. Consumers visiting Software Warehouse's site can browse, review, compare and then download products from Digital River's site. The Minneapolis company will also handle the financial management of the order. News of Software Warehouse's deals coincides with an announcement by Birmingham-based start-up PC Crazy that it has joined forces with another US online computer store, Netsales, to create an Internet shop selling 53,000 computer products. As well as being able to download software, Netsales has said it will create a UK distribution centre to supply consumers in the UK and possibly Europe. As further evidence of the emerging e-commerce market these latest announcements are to be welcomed. But Tim Waggett, managing director of DNA, the UK's first on-line ESD company, believes the sale and distribution of software online is going to take off in the new year. Moreover, this electronic model of doing business could threaten traditional software channels. "In the IT market the software industry has a real opportunity to exploit the commercial benefits of e-commerce," he said. "While existing distribution channels have taken the first steps by offering on-line ordering, for physical distribution, the real and profitable opportunity lies in true electronic distribution. "[If this takes off] the existing methods of distribution are made redundant overnight," he said. ®
Tim Richardson, 07 Dec 1998
The Register breaking news

Survey: Net more important than TV, VCRs, hi-fi

The Internet is fast becoming such an integral part of people's lives they are finding it increasingly difficult to manage without it, according to the 1998 Cyberstudy, conducted by researchers Roper Starch and America Online. Of the 1001 Americans questioned, nearly half of those with laptops said they take them on holiday and a quarter said they check their e-mail while on vacation. And if given the option of being stranded on a desert island with either a TV, telephone or PC, two-thirds said they would prefer to have Net connection to communicate with the outside world. Around 80 per cent of people confirmed that that the Net makes many activities easier and more convenient. And almost half (47 per cent) of them believe that being online has a more positive influence on their children than watching TV (35 per cent). "This first-of-its-kind study confirms that our vision for the interactive medium is becoming a reality," said Bob Pittman, AOL's president and COO. "The Internet is becoming a necessity -- in fact, other Roper research suggests it has surpassed VCRs, stereos and cable TV as a necessity for those who have access to them." "Whether it's keeping in touch with friends and family, getting information to make better buying decisions or trading stocks, people are clearly seeing everyday tasks are easier and more convenient when they're done online -- and the longer people have been online, the more benefits they notice," he said. The survey also found that 51 per cent of people prefer using e-mail as a business communication tool compared to 35 per cent who prefer the telephone and five per cent who prefer traditional paper mail services. ®
Tim Richardson, 07 Dec 1998
The Register breaking news

Sun modifies JDK licence to ban benchmarks

Sun's licence agreement for the recently released Java Development Kit 1.2 contains a curious amendment from previous licences -- anyone who uses the kit is effectively forbidden from saying publicly how good it is. The licence agreement contains a clause which says: "You may not publish or provide the results of any benchmark or comparison tests run on Software to any third party without the prior written consent of Sun." In short, anyone -- say, a journalist reviewing the JDK -- isn't allowed to say that it runs applets more quickly or more slowly than previous releases, or how well it performs in comparison with other Java development tools, such as Symantec's VisualCafe, Inprise's Jbuilder or even Microsoft's Visual J++ (assuming it is made to be fully Java-compliant as per the ruling in the recent case brought by Sun), without the Great Satan of Workstations' say-so. Of course, whether it actually would sue is another matter. Microsoft also includes such a clause in its software, but that didn't stop users pointing out how slow some of its Office apps are. Microsoft, as far as we know, didn't sue anyone that did publish benchmarks for breach of licence, and it seems unlikely that Sun would. So why include the clause at all? "It's a lawyer's cut and paste job, taking the lines from other licence agreements," suggested one Java user. Or could it be that Sun is so worried Microsoft will now do a better job of Java than it can, in order to make a point over the 'Microsoft Java not up to scratch' case that it doesn't want to allow Microsoft (or anyone else) to point that out? ®
Tony Smith, 07 Dec 1998
The Register breaking news

Microsoft releases update for 98 Y2K problems

Microsoft has today released a Year 2000 update for Windows 98, correcting a surprisingly substantial list of problems in a product just a few months old. Windows 98 is however just one of a long list of "minor" Year 2000 compliancy issues Microsoft has to tackle over the next 12 months. Sources in the channel claim Microsoft has told staff they may not take leave in the first weeks of 2000, and this is scarcely surprising, considering the number of Microsoft products currently listed on the company's site as Year 2000 "compliant with minor issues." (MS Compliance List) Fixing these "minor issues" will require the application of patches to large numbers of clients and servers in corporate networks, and in the absence of Windows 2000, beta 3 of which has now slipped to Q1 1999 (Windows 2000 beta slips), Microsoft doesn't have a ready mechanism for rolling out patches automatically. And even if Windows 2000 does go gold before the end of 1999, it's unlikely that many companies will see it as a reasonable solution for Y2K problems - deploying a new server OS while at the same time attempting to eradicate Y2K bugs sounds like a recipe for IT management career suicide. The general corporate view of Microsoft and Y2K is that the company didn't take the issue seriously early enough, and still hasn't got its act together properly; the Windows 98 situation is unlikely to change their views. Last week Microsoft's Y2K site listed Windows 98 as compliant with "minor issues," whereas today it's listed as compliant. Referencing the downloadable update for Windows 98, Microsoft talks about the "minor issues" the update fixes. And puzzlingly, the Windows 98 entry in the Y2K site claims it was last updated on 15th October 1998. The Y2K problems of Windows 98 are indeed minor, but in common with a lot of other Microsoft minor problems, they still need a patch. Microsoft lists them as follows (for the sake of completeness, we've quoted the post in its entirety: Date Rollover: If a system is booting at the precise fraction of a second when the date rolls, the system clock may display an inaccurate time or date. The occurrence of this would be extremely rare because the exact time frame varies from machine to machine and lasts typically less than 1 second. Date/Time Control Applet: If a user opens the Date/Time applet in the control panel and sets the date to February 29 of a leap year and then uses the up and down buttons to change years, February 29 might be displayed even for non leap years. This is simply a display problem, however. The user cannot apply this incorrect date. Dialer.Exe Log: When a user makes a phone call using the Phone Dialer applet, the log file created after completion of the telephone call displays the year portion of the call date incorrectly. DOS XCOPY: When using XCOPY in real mode with the optional parameter /D:date, xcopy does not accept years entered as two digits, except for the years 80 through 99. The message "Invalid date" is displayed. When using xcopy in protected mode (from within Windows), two-digit dates are accepted but are recognized as being within the 20th century (for example, 02/05/01 is seen as 02/05/1901). Java Virtual Machine: Some year 2000 issues have been reported in connection with Java virtual machines based on the Sun Microsystems Java Development Kit (versions 1.1.1 to 1.1.5). For example if a Web site uses Java and makes use of the java.txt.SimpleDateFormat class library, and the user enters four digits for the year, the date functions may truncate the year and use only the first two digits. Lagging IP Lease Dates: If a user logs onto a LAN on or after March 1, 2000, and runs Winipcfg from the Run command or Ipconfig from the DOS VM, the DHCP client reports the IP lease date as having been obtained on the previous day. Microsoft Foundation Class Library: After the year 2000, programs that use the COleDateTime function may improperly parse a date. For example, 02/05/2000 may display as 2/05/100. To view an example of this after the year 2000, from Programs\Accessories\System Tools users can run System Information (msinfo32.exe) and save the file. Users should then open this info extension file and from the File menu select Properties. The date stamp may display the improper date. Microsoft Wallet: When entering credit card information in versions of Microsoft Wallet that precede 2.1.1383, users must enter the month, day, and year for expiration dates beyond 2000. Otherwise, information may be parsed incorrectly. For example, entering an expiration date of 5/01 could be parsed as May 1, rather than May, 2001. Users can correct this by installing Microsoft Wallet version 2.1.1383 or later or by downloading the Windows 98 Year 2000 Update. Programming to Data Access Components: If a programmer codes to ADO or OLE DB, and uses data access components such as adDate, adDBDate, Dbtype_Filetime, chooses to use an international date format with periods as separators, and specifies a year by two digits earlier than 60, then the date may be translated as a time. Regional Settings Date/Time Picker: If Regional Settings in the control panel is set to use two-digit years, then the date/time picker function may not return the proper date. Only two digits are accepted at a time. To ensure proper handling of dates, the user can either set Regional Settings to four-digit date handling or download the Windows 98 Year 2000 Update. WordPad Custom Properties: If a user selects properties, custom on a WordPad or Word document, the custom date setting will not accept 2000 as a valid entry when entered as "00". All two-digit dates are assumed to be in the 20th century and if the time zone is set to Far East, the date properties will lose a day when the year is entered as 2000.
John Lettice, 07 Dec 1998
The Register breaking news

ARM cosies up to OEMs

ARM has unwrapped a channel accreditation programme, designed to cosy up to System on a Chip (SoC) designers. The vendor says an enlarged ARM design community is necessary to handle growing demand for ARM “design engagements”. Under the ATAP or ARM Technology Access Program, design houses can try-out for ARM Approved Design status, following a competency audit. In return, ARM will make its training, software tools, and design resources available to qualifying companies. First qualifying member is the Cadence SoC Design Centre in Livingston, Scotland.®
Drew Cullen, 07 Dec 1998
The Register breaking news

Europe's largest independent chip foundry founders

Newport Wafer Fab Ltd (NWL) has become the latest UK-based chip plant to be threatened with closure by the intense competition in the semiconductor market. NWL went into administrative receivership last week -- administrator Price Waterhouse will run the company while it seeks a buyer. The chip maker, the largest independent foundry in Europe, claimed the decision to call in Price Waterhouse had been made when banks called in loans owed by NWL's Hong Kong parent, QPL International Holdings. The news follows concern over the future expressed last month when it emerged that NWL was considering a raft of jobs cuts (see NWL cuts could affect Wales). At that time, chief executive Steve Byars said the business was "fundamentally sound". That statement, plus a recent cash injection of £230 million, suggests the closure is more about QPL's problems than the plant's. NWL's plant currently employs around 550 people. It runs 1.5-micron and 0.35-micron CMOS processes and an 8in wafer line. NWL was formed in 1992 out of QPL's buyout of a former Inmos plant. ®
Tony Smith, 07 Dec 1998
The Register breaking news

Exclusive: Free Internet access set to become the norm

Paying for Internet access in the UK is to become a thing of the past following the announcement that Callnet is to offer companies the chance to offer free access to their customers. The London-based company is in the final stages of negotiations with a number of major blue chip companies and an announcement is due later this week on Callnet's first corporate customer. "This is the death knell of subscription Internet services," said Aaron Goodman Simpson of Callnet. "As far as I know, this is the first branded service of its kind." "We provide exactly the same service as any ISP except we don't penalise people with one-off sign-up fees, forced advertising or expensive phone calls to premium call centres when they need help." Callnet -- which is running the service hand in hand with Cable and Wireless -- has been running a pilot for the last couple of months to prove that the model actually works and to iron out any glitches in the system. According to Callnet, companies who sign up to the service have to undertake to generate a set number of users -- anything between 15,000-20,000 a year. In return, Callnet will provide the service, manage the technical support helpline and take on the responsibility for administering the system. For their part, companies signing up to the service will receive a slice of the income generated by the local dial-up calls and benefit from the branding and marketing exercise associated with such a free service. Although its likely that such a move could accelerate the number of people signing up to the Net in the UK, it's more likely that subscribers will jump ship from their existing paid-for ISP to a free alternative. This announcement comes at a time when Dixon's Freeserve -- much criticised for its £1-a-minute charge for helpline support -- has notched up around half-a-million subscribers to its free Net access service in two months. ®
Tim Richardson, 07 Dec 1998
The Register breaking news

Dixon shares soar on FreeServe bubble

Now we know the City is mad. UK consumer internet customers are worth more than £1,300 each –- according to the latest valuation of Dixons Stores Group. It is set to return to the FTSE 100 index of the UK’s biggest companies on the back of its extraordinary success with new ISP service FreeServe. Shares soared £690 million on news that the company signed up 510,000 users in its first month of service -- equivalent to £1,352.94 per (non-paying) customer. That’s an awful lot of ads to sell to justify that sort of premium UK Trade valuations for paying consumer Internet accounts used to come in at £180 each – these days £50 or £60 is nearer the mark.®
Drew Cullen, 07 Dec 1998
The Register breaking news

Phone giants and MS square-off in Bluetooth standards war

The Beast of Redmond breathed fire in the direction of the Bluetooth SIG today - or did it? According to "sources" referred to by our good friends at InfoWorld, Bluetooth is running into trouble in trying to recruit Microsoft, which is just about the only significant non-member, and existing members are getting restive about the standard being controlled by the five founders. The big questions (aside from why InfoWorld has to get up so early these days - 6:38am, Ephraim and Dan? Come on…) are who the sources were, and why. The five founders, Ericsson, IBM, Intel, Nokia and Toshiba are the only members currently who have a vote on what goes into the final Bluetooth spec, and a 4-1 vote is enough to decide. According to the sources, if Microsoft joined it would want to be a voting member, and would also want the rules changed so decisions would have to be unanimous. So any of the (now six strong) steering committee would have a right of veto. You don't have to stretch your imagination particularly far to see these as natural requirements for the Beast to demand in exchange for its approval. But The Register has sources too, and they say Nokia, Ericsson and Intel have specifically devised the system so that the three of them will carry the day, and they're not about to change it. Three of them? "It would take a very strong person from Toshiba to argue," say our sardonic sources. It probably hasn't escaped your attention that Nokia and Ericsson are shareholders in Symbian, the Psion-inspired dagger aimed at stopping Microsoft dead in the mobile phone and pocket communicator market. So they're not about to change their rules to accommodate the Beast - not when they think they're winning. The attitude of Microsoft's old buddies at Intel can be gauged by a cursory glance at the trial coverage of a few weeks back - hardware Satan isn't going be doing software Satan any favours, if it can avoid it. Which means that if Microsoft absolutely insists on its terms, it doesn't join Bluetooth - it's what you might call an undone deal. But there's another reason why it's difficult to see Microsoft joining - Bluetooth SIG membership requires you to give up all intellectual property rights for technology contributed to Bluetooth, and this isn't the sort of deal you'd ordinarily expect Microsoft to sign. It's also not the sort of deal you'd expect Qualcomm, Microsoft's partner in the WirelessKnowledge joint venture, to sign. Qualcomm has been raising merry hell over IP relevant to G3 wireless systems, and isn't about to give anything away there. But Qualcomm is a Bluetooth member, so it signed. InfoWorld's sources claim Microsoft and "current members who did not read the fine print" are concerned about the IP pooling arrangements, and the thought of the ordinarily truculent Qualcomm screwing up in this way is certainly chucklesome. Has its new friend just discovered it's given away Manhattan for a bunch of beads? But although Microsoft might have a certain interest in having Bluetooth either stopped or amended to its own satisfaction, we at The Register are seriously considering the possibility of other outfits being the ones with axes to grind. Proxim, for example, was bad-mouthing Bluetooth at Comdex , and as Nick Hunn of TDK Systems, a noted Bluetooth aficionado, says, "Breezecom and Proxim have the most to lose if Bluetooth succeeds." Hunn's view is that the as yet largely anonymous critics of Bluetooth are approaching it far too much from a PC industry perspective. "Bluetooth pulls in together two of the most diverse industries we've ever had, PCs and telephony," he says. "But less than 10 per cent of Bluetooth devices are likely to be PCs." That's an obvious problem for companies whose business is based around the PC model - say, Microsoft and wireless networkers like Proxim (Intel is doing some fancy footwork that may mean its business isn't based around the PC model, RSN). They'll raise objections, as apparently Andrew Seybold does, about Bluetooth not supporting IP, because that's the way the whole world is going. But Hunn argues first that this is "totally incorrect" - Bluetooth has been designed to be able to work with IP, and second that a lot of Bluetooth applications will be way, way outside the PC space, and won't need IP. "A lot of the software is going to be embedded, and a lot of it at the moment is ARM-based type OS, VXWorks [that outfit that got friendly with Intel recently] and OS9. That's what you use in the real world. Microsoft might not like it, but it's true." Hunn suggests that there is going to be a conflict over Bluetooth, but that it's likely to be a phone-style turf war rather than the company versus company standards war the computer industry is used to, and that InfoWorld is writing about. The Nokia and Ericsson engineers, he says, are ETSI (European Telecommunications Standards Institute) veterans and they're driving the spec from an ETSI perspective. "They know how to write a spec that works," he says, as opposed to "802.11, USB and PCMCIA." Hunn, a good European, takes the view that the ETSI approach to mobile phones worked, and the US approach did not - anecdotal experience of nascent US digital phone networks suggests he may have a point. And Bluetooth, although theoretically global, is being driven hard from Europe, and is likely to succeed first in Europe. You could therefore envisage a phone-style continental rift where the US peeled-off in the direction of HomeRF and/or WAP, while Bluetooth succeeded in Europe, GSM-style. One of the objections to Bluetooth, incidentally, has a particularly interesting aspect to it. It uses the same frequency as 802.11 and HomeRF, so it obviously clashes with them. But a side-effect of Bluetooth design is that this doesn't matter - for Bluetooth devices, that is. "Bluetooth will kill all of them," says Hunn. "It's far more aggressive." ®
John Lettice, 07 Dec 1998