Bij Gamer's Depot hebben ze een review van een stel speakers van Midiland (nooit van gehoord hier in .nl). Maakt niet uit, hier wat info:

Of course being a gamer, the first thing I threw at these were games. For their price these little guys really surprised me. The subwoofer, as small as it is, really pounded out the lows with decent punch. I’m sure this is helped by it being housed in a real wood enclosure. With each explosion or gun fire, I was pleasantly taken into sound that didn’t seem like it should be coming from speakers this cheaply priced. The next thing was to run Mp3’s, CD’s, and midi. Here to, the sound quality was equally impressive, I could actually turn it up to a decent volume without much distortion or coloration in the sound. With the controls placed on the right-hand satellite speaker adjusting your volumes is easy, and if you don’t’ happen to own a 3D sound card, they even include a nice little 3DSP button to add a nice 3d-effect to most games and music sources. One of the things I did do was fill the sub enclosure with dampening material (poly-fill) which greatly helped to control the woofer movement at higher volumes. I think the only real drawback is the lack of power, if Midiland would include a more powerful amp, I’m sure these babies would scream pretty good.

Fresh3D heeft een review in elkaar gezet van Epox's ep-51mvp3e-m. Ja, kunnen wij een moederplank een cryptischer naam geven?

Anyway, hier een plak uit de review:

So what makes this board stand out? Well, for starters, it has 1 Meg of onboard L2 cache, as opposed to the 512 KB standard with all other boards. This alone makes for a huge speed increase. It also has some very nice monitoring functions on it. Before I looked at the BIOS while changing a few settings in it, I never really gave it any thought, but when I did notice it, I also noticed that my K6-2/350 was heating up like Peewee Herman at a Sex theater. Quick fix: Put it in my fridge.

Bij C!T Tips&Trucs (een lullig goedkoop computermagazine gemaakt voor en door mensen die niets van computers begrijpen) geven ze maar weer eens blijk van hun enorme kennis van computer technologie, zie deze posting van Reinder op het forum:

Met droefheid in mijn hart kondig ik u aan: ik heb u allen voorgelogen.

Al die tijd heb ik u doen geloven dat ik verstand had van computers. Vandaag dan komt het er eindelijk uit.Ik weet NIETS. Al mijn adviezen en meningen over wat u moest kopen... die facade heb ik vandaag in rook op zien gaan.

zoals namelijk te zien in Computer!Totaal-Tips en Trucks dec.-jan. Hebben wij allen het verkeerde mobo gekocht.

Hardware, stel zelf uw systeem samen heeft namelijk aangetoond dat het plankje dat ik u allen aanraadde geheel verkeerd is. U had namelijk een plankje moeten kopen zoals daar geadviseerd: een plank met 4Dimm sloten, en boven alles..... 2 (jawel, u leest het goed... TWEE) Accelerated Graphics Poort- Bussen. Die zijn namelijk geweldig handig voor o.a. Videokaarten.

Ok, nu ff weer normaal.... afgezien van het feit dat er geen BX bord is die er 2 heeft, Die hele chip set HEEFT niet eens de ondersteuning ervoor. Enne... AGP-bus?? AGP is zoals de naam het al zegt: een POORT, en DUS geen BUS. er waren nog wel andere dingen hoor... zo was bijv. de snelste PII een PII 400 (en ik maar denken dar er ook een 450 was...) en wel meer van die flauwe kul.

C!T-redactie.... ik kreeg langzaam aan al het idee dat jullie werkelijk geen idee hadden waar je het over hebt, maar dit is dus een klassiek voorbeeld van de bel horen luiden, maar niet weten waar de klepel hangt.

AGP2x houd dus NIET in dat er 2x een AGPslot op een mobo zit he??? Wat het wel betekent zoek je maar eens lekker uit, daarna praten we wel verder... Journalisten, kenners...... pah....

"C!T Tips en trucks" is overigens in alle supermartkten en bij tijdschriften winkels te krijgen. blader maar eens voor de grap door het inkijk exemplaar.

LOL, wat een sukkels.

Het was gisteren oorlog bij de Aldi, waar kudde's computernerds een poging deden een goedkope PII te bemachtigen, daarbij lichamelijk geweld niet te schuwen.

Zie dit bericht in de Telegraaf (OK, 't is een kutkrant, maar wel goed in dit soort nutteloos nieuws):

Een supervoordelige computeraanbieding heeft gisteren bij diverse vestigingen van supermarktketen Aldi tot chaotische taferelen geleid. Met de aanbieding, waarvoor Aldi gisteren grote advertenties plaatste, konden liefhebbers vele honderden guldens op de normale aankoopprijs besparen.

Per vestiging was een beperkt aantal computers beschikbaar. De belangstelling bleek vele malen groter dan het aanbod. Aldi heeft in Nederland zo'n 330 vestigingen.

Bij de vestiging in Oosterhout waren om 6 uur 's morgens al talloze klanten op de been. Bij aankomst van de verkoopsters drongen zij de zaak binnen. Toen bleek dat er slechts 24 computers beschikbaar waren, drongen zo'n twintig mannen het magazijn binnen om toch nog een exemplaar te bemachtigen. De verkoopsters liepen rake klappen op, toen zij de mannen probeerden tegen te houden.

Vechtpartijen

Ook buiten ontstonden enige vechtpartijen. Ook zouden diverse mensen zonder betalen met artikelen aan de haal zijn gegaan. De verkoopsters bleven huilend en overstuur achter. Enkele klanten toonden zich verontwaardigd dat de meisjes niet beter op de aanbieding en de daaruit voortvloeiende stormloop waren voorbereid.

Ook in Groesbeek zouden er klanten onder de voet zijn gelopen en zou het bijna tot vechtpartijen zijn gekomen. Bij de Aldi-vestiging in Hoogeveen waren de beschikbare 30 exemplaren binnen tien minuten na de opening uitverkocht en moesten tientallen klanten teleurgesteld naar huis worden gestuurd.

Aldi weigert ieder commentaar: "Wij geven geen informatie, noch negatief noch positief." Wel erkende men dat de belangstelling voor de computeraanbieding zeer groot was. De regionale werkmaatschappij in Ommen gaf toe de hele dag door gebeld te zijn door teleurgestelde en boze klanten.

Erg bizar... (en thanx to Emiel voor de posting in het forum).

Sysopt.com heeft een artikel gepost met handige info over het gebruik van koelpaste ter verbetering van de warmtegeleiding tussen CPU en heatsink. Errug handig als je dat vage goedje op de juiste manier op je CPU wilt toepassen.

Hier wat stuff over wat nou eigenlijk het probleem is:

A heat sink is made of an extremely low heat resistance material. The larger the heat sink, the more heat it can hold and more importantly, the more surface area it has to dissipate the heat radiantly. To help increase the ever important surface area, fins are part of the mold with fans blowing air over these fins. The more air that flows over the fins, the more heat that is wicked away. In essence, this is how a heat skink pulls the heat away from the CPU out to the surrounding air.

The largest heat sink and fan will not help cool your CPU if you do not have a good connection between the two. Simply latching on a heat sink is not adequate as it will result in about a 1% surface area connection. The problem is that both surfaces are not machined to fit against each other. There are massive variations on the surfaces when you examine them at a microscopic level.

Hallo,
Jullie mogen hier mijn eerste nieuws posting lezen!
Ik las vandaag een aantal leuke stukjes in usenet en in de nieuwe Computer Techniek, waar in stond dat Microsoft in hun documenten bijna openlijk schrijft dat ze open standaard op hun eigen manier implenteren, zodat er een microsoft standaard komt, die op een gegeven moment de oude standaard 'over-ruled' want iedereeen draait toch Microsoft software....
Deze gegevens zie je op dit moment al in de praktijk, kijk maar naar de Java van Internet Explorer, en ook het afwijkende JavaScript model van Internet Explorer. Nu is daar ook nog het volgende bij gekomen:
De 'geupdate' DHCP server in Service Pack4 van Windows NT 4 is niet in staat om andere machines dan windows van een IP-adres te voorzien....
Over het algemeen hoort men de laatste tijd steeds meer negatieve geruchten van Microsoft door verschillende mensen. Wat daarbij toch wel apart is dat bij een onderzoek gebleken is dat Bill Gates er niet meer of minder door wordt aangezien(dit laatste hoorde ik op Radio Online).

Het houdt niet op met goeds bij Tweak-it, met een vergelijking tussen de Xeon en Pentium II:

Whenever a manufactures release a new processor, it's always expected to blow away the previous generation. The fact is that in most cases it does not. The XEON is no exception, posting only marginal gains over it older brother. What the XEON does bring to the market is scalability. By design you could have up to 8 XEON processors on one board with out the need for extra logic to control them. This alone is enough for some highend users to jump at the chance to own them, but for most of us, it doesn't justify the large price tag. What Intel did with the XEON was give us back technology they had taken away with the release of the Pentium II, cache running at the same speed as the processor core like in the Pentium Pro. They also increase the address and cache range to a whopping 64 gigabyte. The problem here is even their best chipset the 450NX will only support 8 gig of RAM installed. Intel also claims they made some other architectural enhancements to improve performance in a server environment. Unfortunately I don't have the resources to call them on it, so will have to take them at their word for now.

Nog meer info van Tweak-it. Dan heeft er op deze pagina zijn ervaring van de Comdex neergepleurd.

Hier wat interessante stuff over de K7:

Probably the single hottest technology displayed at Comdex could not be found out on the main display floor for everyone to see, no, it was squirreled away in one of the meeting rooms which was the only show presents that AMD had this year. The technology is of course the K7.

Although no formal benchmarks were allowed on the alpha system, it is assumed that the performance will be stellar to say the least. The K7 will plug into what AMD calls a Slot A, which is identical to a Slot 1 in appearance but not in function. The Slot A is based on Digital's EV6 bus protocol and will operate at a 200MHz FSB. Because the Slot A is the same physical connector as a Slot 1, AMD flipped it around so that people wouldn't mistake it for a Slot 1 and insert a Pentium II in it. According to the AMD rep, the first testing of the K7 was done on a Digital motherboard designed for the Alpha processor without modification. This is just speculation, but there is a possibility of motherboards supporting both the DEC Alpha and K7 in the future. The test board they showed code named Gomez was in the Alpha stage along with the Irongate chipset and the K7 processor. AMD's goal is to have all the technology laid out for manufacturers to allow for a smooth product flow when the K7 is ready for release. They have also entered into agreements with both ALi and VIA to allow them to come to the market with chipsets supporting it. The K7 will enter the market based on a .25 micron process with 128K L1 cache and from 512K up to 8MB backside L2 cache.

Introductory speed, although not confirmed, is expected to be at 500MHz with speeds beyond 600MHz reached with the release of a .18 micron processor later that year. With the help of a technology agreement with Motorola that will bring AMD into a copper process, they have stated they will reach 1GHz in the year 2000. The first chipset (Irongate) supporting the K7 will make use of SDRAM, with a move to Rambus (RIMM) in a later releases.

The K7 will include 3D-Now and AMD says they have the licensing for KNI if they need to include it. Entering into territory no other x86 processor has been other then ones for Intel, the K7 will also support scalable multi-processing. Word is, AMD is working with third party chipset designers to bring up to 8 K7's together in one solution.

Targeted for a early second quarter 1999 release, the K7 has a chance of knocking Intel on their butt, lets just hope AMD can pull it off.

Tweak-it.com is terug! en nog wel met een hele vette update over SDRAM DIMMs in combinatie met Abit's holy BH6.
THG was nogal kritisch over de BH6 en Dan van Tweak-it heeft dan ook errug leuk nieuws voor BH6 bezitters:

Ok, here is the scoop. As far as I can find the BH6 has absolutely no problems with memory compatibility. I have tested every stick of SDRAM I have laying around, including stuff I wasn't sure still worked and the BH6 passed with flying colors. I even played mix and match with PC66 and PC100 SDRAM to see if I could find a weakness. The only piece that does not work 100% is the new 512MB registered SDRAM DIMM I received from Gyser Electronics. It still works in the board, but is only reported as being 256MB. Other than that it ran flawlessly. This should not be a surprise as Abit doesn't claim support for Registered SDRAM or DIMMs larger then 128MB.

Op Tweak-it staat een hele lijst met DIMMs die zonder problemen in staat waren om in peace met de BH6 samen te werken.

Dus THG's review van de BH6 is bullshit!

Ook Sharky Extreme heeft een review van Half-Life. Als we de review mogen geloven is dit echt een megavet spel:

Ordinarily, games that have been as hyped as Half-Life has been tend to fall short of the public's inflated expectations. After the release of Epic's Unreal and Ritual's SiN and the heart-crushing realization that they didn't live up to their hyped standards, we started fearing that the same fate would befall Half-Life. Earlier last month however, we were given the opportunity to play through Half-Life: Day One, an OEM demo which is bundled with Diamond's MX300 sound card, Guillemot's MaxiGamer Phoenix Banshee card and a few other pieces of hardware. The game didn't live up to the hype generated by two years of print and online coverage… it completely surpassed it. Does the full game continue the excellence provided by the demo? We'll say it right here: Hell yes, and then some. Read on to find out why Half-Life gets our coveted Game of the Year Award, and as Highlander's Duncan MacLeod so eloquently put it, "there can be only one".

CPU Central heeft wat info over het gebruik van een Peltier koeler als wapen in de holy war der overklokkers tegen het eeuwige gevaar van overhitting:

Okay, I've finished testing the Pentium II Peltier cooler, and here's what I've come up with: the peltier unit keeps the Pentium II a lot cooler than the Glacier 4500B, however, my SL2W8 still won't overclock to anything past 450MHz. With the Glacier 4500B, the Pentium II would not get into Windows at 464MHz, however, with the Peltier unit, it would get into Windows, however, it was too unstable to use. My advice: if you've managed to get your Pentium II up to 464 or 500 MHz and are having a little trouble with stability, this peltier unit may give you that added cool needed to keep your processor stable. If you've overclocked your Pentium II to 450MHz, and are happy with it's performance and stability, stick with your current cooler, and don't unnecessarily introduce the extra heat into your case associated with peltier cooling units.

BTW, the unit I tested consisted of one fan (although it was much larger than the ones on the Glacier 4500B), and a massive heatsink, again, bigger than the one on the Glacier 4500B. The unit I tested was a prototype unit which came from Mushkin Online. It's not yet available for sale, however it should be available soon. Keep checking back to their site if you are interested in it.

The Super 7 Hardware Guide heeft een review gepost van Quantum's hete Vuurbal:

With today's higher speed processors, memory and buses, upgrading you hard drive can really improve your overall system's performance.  The Fireball EX platform is an outstanding drive and shows performance increases of up to 45% over Quantum's former Fireball SE platform.  And although Quantum has released the specification for it new Fireball CR platform, whose new Ultra ATA/66 interface could improve performance even more, the EX is the one to buy for Super7 users.  The Ultra ATA/66 interface will require a mainboard upgrade for most of us.  The EX on the other hand is ready to go now (in fact, I believe the CR and EX drives are identical except for the interface) and virtually eliminates disk waiting time in most applications

OK, het is een beetje laat: Fastgraphics heeft een HCC Coverage gepost met spul over de afgelopen HCC dagen.

Hier wat stuff over Diamond:

I've been talking to Diamond's PR manager for Northern Europe, Neil McGuinness. He told me some interesting things about Diamond. First of all Diamond will (surprise, surprise) most likely be among the first to jump onto the Voodoo 3 bandwagon. Secondly, Neil pretty much agrees with me that 3Dfx's choice to make the Voodoo 3 chipset incapable of using the AGP DME functionality might not be the smartest move out there, most of all since the initial specs show that the maximum memory size will be 16 Meg... Time will tell about that though.

Unfortunately Diamond wasn't able to show me the new Rio MP3 player, they had been shipped in from the far East, but they didn't get to the Netherlands in time.

I did get to see (well, hear actually ;-) one of the new Monster Sound MX300's and it sounded great.. Unfortunately the environment was very noisy, and only two speakers were connected, but it sounded great anyway.. I've been playing a bit with the new A3D 2.0 demo's and the positioning with two speakers seems to have increased quite a bit over the previous A3D 1.0 products. I will get a MX300 very soon, and you can expect a head to head with Creative's Sound Blaster Live within two weeks.

Een paar weken geleden schreef Overclockers.com een wedstrijd uit voor wie het eerst een PII-333 stabiel op 620Mhz kon krijgen.

Asetek is het gelukt om 'm als eerste op 620 te draaien, daarvoor maken ze gebruik van hun leuke koelsysteempie (zie de site van AseTek voor info):

I have been corresponding with Asetek (www.asetek.dk) in Denmark - these are guys who have integrated a cooling unit into a case to cool the chip to below freezing. They are claiming to be the Contest Winners by running an SL2WY week 37 (98371032-0203) at 620! I have seen screenshots and they tell me they have run it for 2 days while stressing the chip and it is stable. Take a look at their site under Tests for performance data.

I have been skeptical about what supercooling can achieve and must now admit there may be something to this after all. If supported, it opens up a whole new avenue for perhaps using peltiers to attain some remarkable speeds. More on this as I get supporting details. Crypto at bxboards.pair.com is acting as a judge and I will defer to his considerable expertise.

Humphrey at PC Nut (www.pcnut.com) was kind enough to offer a $100 Gift Certificate as the prize - Question is what do you give a guy who has a supercooled case - Keg of Beer??

Mikroschoft heeft z'n aandeel van 10% in RealNetworks geselled. Niet echt verbazingwekkend, aangezien Mikroschoft en RealNetworks de laatste tijd toch al niet al te vriendelijk met elkaar om gingen. Hier wat info van Techweb :

Considering the acrimonious relationship between Microsoft and RealNetworks, Microsoft's decision to sell its 10 percent stake in the streaming media company took few by surprise.

The partnership had been rocky since October of last year, when RealNetworks' released version 5.0 version of its technology without support for the ASF format Microsoft is pushing as a Web standard for streaming media.

This week, RealNetworks released the final version of its G2 player -- which backs the SMIL streaming media format that Microsoft said is unnecessary. That won't do anything to warm things up between them, although RealNetworks said it actually would prefer that Microsoft remain a shareholder.

Mozilla.org heeft een eigen reactie gepost over het gebeuren rond de overname van Netscape door AOL. Hier ff het belangrijkste:

Some people have the impression that the Mozilla agenda is set by Netscape, and to some extent that is true: because Netscape is paying more than a hundred people full-time salaries to work on the Mozilla code base -- and to give their code away.

In addition, Netscape is funding mozilla.org, those of us providing management and infrastructure and tools to this large, distributed software project.

So, with Netscape being acquired, what does that mean to mozilla.org? Hopefully, it will mean nothing: hopefully, AOL didn't buy Netscape with the intention of turning Netscape into something that it is not; it's hard to imagine that they would spend $4 billion dollars on Netscape just to throw away the client.

So, assuming that they still want to have a Netscape Navigator, it is not unreasonable to assume that they will adopt the same attitude that Netscape has: that open source works, and that the best way to have a top-of-the-line web browser is to keep it open.

But let's think about some worst-case scenarios. Let's think about the nightmares. What if AOL hates ``open source''? What if they want to undo everything we've done, and make Mozilla be evil and proprietary again? What if they just think that browsers are a waste of time, and that they should just use MSIE forevermore?

Well, they simply cannot undo what has been done. The Mozilla code is out there, and it cannot be recalled. It has been distributed under an open source license, and nobody can ever take that away from you. Ever.

If AOL hated open source, or didn't want to build their own browser, what they could do is fail to contribute to Mozilla in the future. They could stop paying those hundred-plus full-time salaries, and they could stop funding those of us who are mozilla.org's full-time employees.

But be clear on this: the agenda of Mozilla is set by those who contribute to it. If you believe that mozilla.org is just a smokescreen, that the organization exists only to swindle you out of your hard work for the benefit of some shambling inhuman beast of a corporation, then don't contribute to it. Take the source code, and build your own browser based on it. Fork the tree. Do what's right.

The Register heeft weer zo'n leuk artikel over de Mikroschoft rechtszaak. M$ wil nu dat de hele case gedropt wordt, nu Netscape is overgenomen door AOL. Een stukkie:

Microsoft will introduce a Motion to dismiss the case after the DoJ has finished presenting evidence from its twelve witnesses. This is a common legal procedure, but usually has more in common with fisticuffs than legal merit.

It would perhaps have been better if Microsoft had not pre-announced its intention before all the evidence had been heard, but such moves are all part of the running PR campaign outside the courtroom. David Boies, the DoJ's special trial counsel, said that this is the sixth time that "Microsoft has proclaimed the government's case dead".

Microsoft is of course trying to suggest that the AOL-Netscape deal shows just how volatile the industry is, and is cheeky enough to claim that the case should be dismissed as a result. Perhaps Microsoft should re-read the Complaint that started the case: nothing has changed so far as Microsoft's business practices are concerned.

A cooler interpretation is that Microsoft's actions made Netscape vulnerable to the deal that will cost it its independence. Boies said: "What you see here is an exit strategy for Netscape."

Mensen met een dermate vet spaarvarken dat ze er nutteloze dingen van gaan kopen (een 4de Ferrarie enzo) hebben nu iets leuks om naar uit te kijken, want het domein computers.com staat te koop voor het leuke bedrag van 500.000 dollar:

Posted 26/11/98 1:12pm by Tim Richardson

For sale: Computer.com.

It's a sad fact of life but $500,000 doesn't buy much these days. It might just about get you a good time at Comdex. It may even buy you an even better time in Vegas while you're there.

But if you're after one of the hottest domain names on the Web, it will barely pay for the deposit.

Kevin Sinclair, who currently owns computer.com, is looking it sell it to the highest bidder. And his asking price of half a million bucks -- cash -- is just for starters. He thinks it's worth a lot more and expects the price to go higher -- much higher.

Het vuile idee van Mikroschoft om de patch voor Win95 die de problemen met snelle K6-2 CPU's moet oplossen te verpatsen voor 35 dollar zijn mooi om zeep geholpen door AMD, die de patch nu gratis op z'n site heeft gezet.

Hier wat info van The Register:

Posted 26/11/98 2:46pm by Tony Smith

AMD posts Windows 95/K6-2 bug fix for free

Microsoft's penny-pinching attempt to force owners of K6-2 based PCs who have experienced a bug in Windows 95 to cough up $35 for the patch has been scotched by the chip manufacturer itself.

Faced with the Great Satan of Software's apparent refusal to admit its mistake and eliminate the charge, AMD has made the fix available from its Web site free of charge.

As reported here earlier (see AMD users go through roof at $35 fix for K6-2 crash), Microsoft not only charged $35 for the patch but, according to users, didn't exactly go out of its way to point K6-2 owners to the patch's location.

Users quickly alleged the company was using the $35 charge to encourage users to migrate to Windows 98 rather than fix a glitch in 95.

Wil je toch iets nuttigs bereiken met een Mac, dan kun je er Linux op gooien. Yellow Dog gaat Linux distributies voor de Mac verkopen:

Mac software developer Terra Soft Solutions (TSS), best known for its Terra Firma clip-maps, is attempting to become the Red Hat of the Macintosh world with the launch of Yellow Dog, a subsidiary dedicated to distributing a PowerPC version of Linux.

Yellow Dog will be offering two Linux packages. The first, Champion Server, is aimed at Web hosts using PowerPC 750-based PowerMac G3 hardware. Due to ship around the middle of next January, Champion Server contains Linux 2.2, a standard set of Web server-oriented applications and utilities -- Gimp, Netscape, Apache, Python, etc. -- plus TSS' own Web Manager site content administration package.

Gone Home, Yellow Dog's home-oriented Linux package, is scheduled to ship in Q2 1999. It offers the same set of tools and kernel as the server edition, but omits Web Manager. It will also run on most PCI-based Power Macs.

Both products come with only limited installation support -- 30 days for Gone Home and 90 days for Champion Server -- so Yellow Dog clearly reckons its some way off attempting to target corporates in the way Red Hat currently is. Champion Server will cost $74.95; Gome Home $49.95

Afkomstig uit dit artikel van The Register...

Sharky heeft een report in elkaar gezet met info over de waar-blijft-ie Nec PowerVR Series 2:

First off, this isn't about benchmarks. It is about marketing, stockholders, corporate political necessity, a journalist and his toys and somewhere, buried way down there, it is about a group of engineers working very hard to achieve something innovative and possibly great so that you and I can Kill and Maim and Shave the Princess at ever higher resolutions and frame rates.

Actually, it isn't anything nearly so grand, but we do have an update on what happened to VideoLogic/NEC's PowerVR Series 2 PC chip production schedule, which is almost as good!

Ook PC.IGN heeft een review van Half-Life. Hier de intro:

ou've got to hand it to these guys. Valve Software is a start-up, a relative unknown, with no track and, really, no business making such a mighty game. But mighty it is, the best single player shooter I've ever played.

It will be hard for some--mainly those who haven't played the game--to put their finger on exactly what everyone is raving about. There's nothing exceptional about the engine--it's a heavily modified version of Id Software's Quake. Its genre, the first-person shooter, has nearly been beaten to death. Its design couldn't be simpler: get some guns, run, shoot, jump; no laborious cutscenes or misplaced RPG elements here.

The weapons aren't spectacular either, just your usual array of realistic modern firearms and a few alien gadgets with some new twists. The creatures in some respects are fairly standard, and many of the levels have been done before. Taken separately, no particular element in Half-Life shines above the rest, nothing makes or breaks the game.

No, the real secret is that Half-Life is truly the sum of its parts. It takes every element in this oh-so-predictable genre, re-examines it, improves it, and then meshes it with the rest to create a constantly surprising, internally consistent, and always entertaining game.

Gamepower heeft een review van Motoracer II in elkaar gegooid:

I mentioned that Moto Racer 2 has its limitations. Unlike Motocross Madness, all tracks in Moto Racer 2 are bracketed by "invisible walls" that keep the player confined to the course. It prohibits taking shortcuts, but detracts from realism, too. You also won't find any spectators at these tracks and the AI bikers never crash, unless you run into them. Moto Racer 2's greatest weakness, however, is its lack of provision for saving the game's excellent TV-style replays. You can watch them at the end of a race, but that's it. Bummer!

Planet Unreal heeft nog wat info over de vorderingen met de D3D patch voor Unreal, zie ook deze nieuwsposting eerder vandaag.

Hier een stukje c&p:

It's still a few versions away from losing its "beta" status and becoming final. Right now, we and our partners are mainly working with 3D chip makers to get driver issues ironed out. The TNT performance is rising quickly (as mentioned in yesterdays MegaBoard post from Mark), and NVIDIA has become very responsive in optimizing their drivers for the texture-intensive workout Unreal gives them.

On the move to Raleigh and general info:

We're also now finalizing Epic's move to Raleigh. Most of the team has already moved here, and others will be coming here over the next couple months. We're holed up in a temporary office space working on the project currently known as Unreal Tournament now, and are shopping around for a new permanent place. The team is psyched to finally be together permanently in a cool location. Raleigh is exactly the kind of city we want, lots of cool stuff to do yet not too crowded, good cost of living, and a great climate (it's November 25 and warm and sunny outside). The hard part of the move, which is in progress now, is closing down our old office in Maryland. We're eliminating the direct-sales part of our business(selling games via mail and our toll-free 800 number), which served Epic very well for our first 6 years in business--from Jill of the Jungle to Epic Pinball and Jazz Jackrabbit II. That business has been going downhill since we began focusing on retail games, and has been eclipsed by our game development and licensing. We decided to cut it and focus on doing what we do best, and doing it with a lean-and-mean group. This has been rough on everyone, but in this kind of business, the conditions change radically every few years, and companies need to change and adapt in order to survive and be able to thrive.

Gamer's Alliance heeft een review gepost van het extreem heersende Half-Life. Hier een stukkie over de AI:

Well. Where do I start? The AI in Half-Life is, in a word, astounding. No longer do enemies blindly charge you, no longer do they work alone, and no longer are they suicidal. In the case of the soldiers, they coordinate attacks by layering both their offensive and their defensive, they communicate to one another and have specific tasks to perform, and they maneuver to obtain the best fields of fire with fallback positions and regrouping areas. This ability really hits home the first time you see a soldier and run back to hide; after a few seconds, it gets very quiet, and then a grenade or three comes arcing around the corner. You run forward to assault their position and find that not only have they moved, but there is one on either side of you, and another grenade is coming courtesy of the commander hiding behind that box in front of you. Or you’ll be sneaking up a flight of stairs and hear the commander call out "Recon!", upon which a lone solider will appear at the top of the stairs, look down, and then, instead of charging or shooting, run away… with a half-dozen grenades raining down shortly thereafter. Finally, all of the soldiers in Half-Life want to live, not die; if they’re wounded, they’ll hide or run away (or both), holding their leg and crying out "Medic!" The aliens in the game are similarly advanced, but their actions are harder to categorize; in any event, the little three-legged dogs seem to have scouts, and will exhibit pack behavior when attacking. Absolutely brilliant work, and it definitely sets the standard for future games to emulate or enhance. Well done, Valve.

Mercurycenter heeft een artikel met info over RDRAM en andere ontwikkelingen die momenteel het gebied van RAM geheugens aan de gang zijn en die de komende tijd voor een snelle toename van de geheugensnelheid zullen leiden.

Hier een plak 'nfo dat ik uit het artikel gejat heb:

Next year's computers will run significantly faster, not just because Intel, the world's largest chip manufacturer, plans to deliver faster versions of the Pentium processor, but because PC manufacturers will start to move to a faster type of memory known as RDRam (Rambus Direct Random Access Memory).

Rambus of California has been developing RDRam for several years, but the first samples have only just been delivered to PC manufacturers. However, 15 memory manufacturers have signed licenses to produce the chips, including Fujitsu, NEC and Toshiba from Japan, Siemens from Germany, LG Electronics and Samsung from South Korea and IBM. RDRam chips should appear in reasonable volumes in the second half of 1999.

RDRam can provide data at speeds of up to 1.6 gigabytes per second, twice as fast as today's fastest synchronous (SDRam) memory chips. But they won't fit today's motherboards which take DIMMs (dual in-line memory modules -- small cards designed to hold memory chips). Instead, they'll require new circuit boards with slots for Rambus memory modules known as RIMMs.

Intel has supported the development of the Rambus system and took out a license in 1996. Last month Intel also invested $500 million in Micron, an Idaho-based memory chip manufacturer, to ensure an adequate supply of memory components, particularly Direct RDRam. The firm called it an ``investment that supports our microprocessor roadmap into 2000 and beyond.''

Where DIMMs operate at clock speeds of 66MHz or 100MHz, like today's PC motherboards, RIMMs will be able to run at up to 800MHz. This is more important now that processors typically run at 300MHz to 450MHz.

Nou, ik heb m'n sinter nikolaasgedicht al weer in elkaar geknutseld:

De sint zat te denken, wat-ie die gozer van een Femme moest schenken.
Natuurlijk, ik geef 'm een paar SIMMs.
Nee zei Poeter_Piet, want Femme wil DIMMs.
Maar oh! zei de sint: simsalabim,
we geven die harrie een echte RIMM!

Hmm...

Het kan nog erger bij Mikroschoft. Ze waren er namelijk ook nog op het idee gekomen om AMD op te kopen.

Microsoft OEM division senior VP Joachim Kempin seems to be a specialist in thinking the unthinkable, being a loose cannon, or both. Last week his radical recommendation that Microsoft move over to a rental model for its software became public (Click for story), and now it emerges that at the same time he was suggesting Microsoft buy NatSemi or AMD.

Given the timing (December 1997, in a suggestion to Gates), our verdict on Kempin has to start tilting towards loose cannon. Microsoft was immersed in the Windows 95/Internet Explorer integration battle with the DoJ, and had for quite some time been taking a ‘heads down’ view of major acquisitions. The high tide came when the DoJ effectively warned it off from buying Intuit, and although the company slipped WebTV in afterwards, that didn’t necessarily look such a big deal at the time.

A bid for a major x86 company, on the other hand, would certainly have drawn further unwelcome attention from the antitrust authorities. Kempin was making the suggestion in connection with the scheme to move over to an annual fee system for software sales. He saw an Intel entry into the operating software market as probably the most dangerous threat to the plan, as Intel could conceivably undercut Windows drastically (Kempin saw $50 a PC, no matter how low the PC price, as ‘no pasaran’ territory for Microsoft). This could screw-up the fee plan, but even if the fee plan didn’t go ahead immediately, it could slash Microsoft’s OEM revenues.

So to counter this, Microsoft could buy a cloner, presumably tying clone x86 hardware and software closely together and slowly squeezing Intel out. This of course would have been a dangerous plan – maybe Microsoft undercutting Intel on hardware while Intel undercut Microsoft on software would have been absolutely great for the industry.

Could Microsoft have gone ahead with such a plan? With hindsight, from this distance, it would have been suicidal. But in December 1997 Microsoft was (rightly) confident that it could overturn the DoJ injunction, and Bill Gates quite possibly thought defeat for the DoJ on integration would be the end of the matter. He and his lawyers had been taking a line by line, narrow legalistic approach to the consent decree that was being argued over, and as his deposition videos now are making clear, Gates did not take the DoJ seriously. If he thought that Microsoft would soon leave the whole antitrust thing behind and steer into clear blue water, then yes, maybe he could have gone for it.

Meer info vind je hier.

Voor het geval je 't nog niet gehoord hebt: Mikroschoft heeft ooit het belachelijk plan opgevat om een jaarlijks bedrag te gaan vragen voor Windhoos. The Register heeft er een artikel over, waarvan hier een stukkie:

Microsoft has considered charging an annual fee for Windows use, according to a memo to Bill Gates produced in court yesterday. The document, from December last year, came from senior VP sales Joachim Kempin, and was headed: "Licence for limited time and create annuity business."

Kempin suggested the scheme could start in 2001, apparently because in order to do this in the consumer market, the move would have to be made from the launch of a new rev of the operating system. He points out to Gates that he's made the suggestion before, and that it now needed serious consideration. Awkwardly for Microsoft, which was yesterday denying accusations of putting its prices up over the last two years, the memo shows that Kempin was a year ago anticipating PCs down at the $500 level by Christmas 1998, and saying "our royalties could be as high as 10 per cent of total system price…" Nevertheless Microsoft should resist royalty price decreases firmly, he said - i.e., Microsoft's pricing should not fall in line with PC pricing, and Microsoft's percentage share of a growing market should therefore rise. Oops…

And oops some more: "We have increased our prices over the last ten years… other component prices have come down and continue to come down." Nevertheless Microsoft's goal was to get the "highest amount of dollars" it could from its OEM customers. Triple oops…

Intel heeft nieuwe drivers gereleased voor hun i740 3D chippie. Je kunt het spul hier halen.

Hier ook nog ff wat er verbeterd zou moeten zijn in de nieuwe drivers:

The primary focus of these drivers were to fix the various freezing problems some I740 owners were plagued with. Third party Chipset support was finally added for those not on an Intel motherboard, the Intel740 as Primary adapter in multimonitor systems is enabled, and a whole bunch of other fixes are included

Avault heeft een review gebakken van Heretic 2. Hier maar weer wat info:

In essence, there are genuine game-related motivations for using the third-person; it is more than a gimmick, its implementation is a fresh approach to 3D action. The game offers the benefits of a first-person shooter with the cool moves and versatility of the third-person perspective. This enabled the designers to explore new facets of mapping, such as taking greater advantage of vertical space, and permits gamers to bust out of the stiff boundaries of the first-person and truly experience what it is like to navigate a 3D world. The execution is not absolute, but it is the most comprehensive seen to date. Sure, using the third-person was a smart marketing move that differentiates Heretic II from the current deluge of first-person titles, but it was a marketing move that ironically resulted in great gameplay

Lithtimes heeft een preview gepost van Earth 2150.

Hier wat info over de graphics:

First off, Earth 2150 supports 3d acceleration. Gameplay takes place in a stunningly rendered 3d terrain, complete with looming mountains, rolling hills and gaping canyons. Imaging of the terrain and objects is done through "Gouraud" shading, and from the pre-alpha screen shots, it looks real enough to leave the hardened gamer with his (or her) jaw on the floor. There are trees which actually stick up out of the snow, with individual branches, providing cover for advancing units.

The buildings and units themselves look like something from a Star Wars movie set. The attention to detail is Myst-like, and this is only from the pre-alpha shots. Dynamic lighting makes night missions eerie and suspenseful. Both the buildings and the units under your command have spotlights which illuminate the area and cast shadows across the terrain, bringing home the sense of atmosphere which the game bathes in. (These lights can be extinguished to sacrifice visibility for stealth) Product Manager Achim "Mirek" Heidelauf says there will be anywhere from 5000 to 12000 triangles in a single scene as well as limitless number of colored lighting sources.

Overclockers.com heeft een best wel interessant artikel over de relatie tussen de CPU, cachechips en overklokbaarheid.

Hier een stukkie over de Klamath, wat ik als PII-233 gebruiker wel nuttig vond:

The Klameths (233, 266, 300) all came with 7ns cache. (I think it really was 6.67ns cache, since I doubt Intel deliberately O/Cd the cache on the 300.) So, taking 6.67 as our figure, and applying our formula, we get (((1000/6.67) X 2) X 1.15)= 345Mhz. This accords pretty well with the O/Cing record of the Klameth: 333 was fairly easy, fewer got to 350, very few claimed more than that. Since the 233 and 300 had the same cache, which was much quicker than needed for 233Mhz but about what was needed for 300; that explains why the 233 was a good overclocking chip, and the 300 was not.

Mijn PII-233 draait alles rock solid tot 309MHz, daarboven gaat-ie vaag doen in games enzo. Het is me maandag gelukt om de 392MHz te bereiken en dat liep goed in Windoos maar uiteraard niet stabiel bij heftiger werk. Het probleem zijn duidelijk de te trage cachechips die de CPU niet bij kunnen houden.

Bjorn3D heeft wat info over de Rendition V3K en V4K. Er zijn hier in de polder geloof ik nogal wat Stealth II gebruikers, die vast wel geïnteresseerd zijn in hoe de zaken daar bij Rendition lopen (nogal chaotisch, V3300 gecancelled enzo).

Voor wie het niet meer weet hier wat (erg veel) info van Bjorn:

I had the chance of chatting with Chris a bit more yesterday and asked him a few more follow-up questions from the chat I had with him a few days ago.

* How can Micron/Rendition say they won’t make any money on the V3k chip is they would release it?

As long as they fab’ed the chips I thought they at least would be able to get back some of the money by selling cards with them on even if they don’t make a profit. Chris though said that it’s not that easy. The first quantity of chips made is about 1000 chips. If they would go commercial with the chip they would have to make another production run with 100 000 chips. So by not going ahead with a commercial version of the chip they would save the money they have to invest for those 100 000 chips. Of course one might say that they still would get the money back when selling boards but I understand that it does involve some consideration before deciding to go forward and produce 100 000 more chips.

* One of my thoughts has been this: If the V3k and the V4k share parts of the core and any debugging and tweaking of the V3k benefits the development on the V4k, would this also apply after the chips has been fab’ed? In other words - would the v4k development benefit from continued debugging and tweaking of the real silicon? Chris said that the chips share some of the logic. This is the part they now are concentrating on debugging and tweaking in the "chip-simulator". This changes when you go silicon as they won’t even be made by the same fabrics. So according to Chris any continued debugging/tweaking of the silicon V3k won’t benefit the v4k (or at least not as I understood it). This brought me to my next question:

* With the specs that I’ve seen on the chip it sounds odd that they haven’t rushed it out yet and released it as it is very competitive to both the Voodoo2 and the TNT (including the 0.25 micron TNT).

Chris reminded me that as they don’t have the chips in silicon yet there is no way to guarantee it will match those target specs right away. The numbers that have been quoted are what they target the chip at. But as we all may remember it’s not always the target specs is reached right away. The V2200 came out with some serious bugs that made it drop it’s performance very much. It took a lot of debugging and tweaking to get it up to speed again. There is always a risk that this happens to the V3k to (although as far as I remember Rendition has been quite good on matching the announced specs - that’s why I am at least semi confident they eventually would be able to do it with the v3k to).

In connection with this Chris asked me this question:

If the v3k is fab’ed and it turns out it doesn’t perform up to the specs. How much time and manpower should they allocate for debugging and tweaking the chip when it at the same time means that the development of the v4k is delayed?

I guess this is the kind of decisions and questions the high-guys at Rendition/Micron must consider all the time. It’s easy for us to say to them what to do as we don’t have to take any responsibility for any consequences ….

Chris question though lead me to another thing I’ve been thinking of.

* Considering that Rendition/Micron now should be a fairly large company together it sounds a bit odd that they can’t continue develop the v4k and at the same time work on the v3k. It can’t be good that they have to choose what to concentrate on instead of being able to work on several larger projects at the same time. We should be able to get v3k boards out now while they still can keep the v4k development on track.

Chris said that this is one of the reasons they are hiring so many new people. Unfortunately it doesn’t look like they will be there during the v3k/v4k development. I think it’s imperative that they come there soon though as it can’t be good that you have to basically choose between 2 products like the V3k and the v4k when the natural thing would be to release both as they are chips from 2 different generations.

I also asked how long after a chip is fab’ed it would take to get a board out on the market (some extra debugging and tweaking included). He said he didn’t have enough knowledge to give a estimated time-frame as it depends on a lot of variables (including moon-phase, colour of shirt etc. ). Other people have said that it might take up to a few months though.

Finally - there’s a new article at TechWeb which says that a decision about the V3k will be made this week. Well, I asked Chris about it and he said that as far he had been told the decision of manufacturing the v3k’s had already been made. The decision that still is pending is if it will be a commercial product and this decision won’t be made until the chips come back from the factory. Chris also said that this would not happen this week. He would though check up if there had been any decision changes while he was away at Comdex..

In the end we still have no clue what will happen. I feel that the v3k is treated like the ugly duckling by Micron/Rendition. It still has no real place in their product portfolio other than a stepping stone to the v4k. Let’s hope the "ugly duckling" turns out to be a beautiful swan just as in the tale by H.C Andersen …

I thought I close up with some comments about Comdex. I expected a low key show (after being warned about it before the show) but this low-key? Ok, they showed some prototype SocketX stuff but where’s the announcement? Or what about a little press-statement? Did I miss them? And didn’t we expect at least a few announcements? Even though the "bomb-shell" stuff would come later? Sorry, this really doesn’t work! Ok, I can understand the confusion regarding the V3k. But what about the v4k? It’s the "Golden child" at Rendition/Micron right now. Both Nvidia and 3dfx announced products that should come out around the time frame we could expect a V4k (although the v4k probably will come end of Q2 start of Q3). Isn’t it time for the exec’s at Rendition/micron to have laid out the long term plans soon?

What I’m worried about is the fact that Rendition won’t be included in any 3D card overviews that will be made in January about the cards that are coming. If no one reads about you, you don’t exist. What I want to see real soon is a announcement where Rendition/Micron announces the v4kE for a Q2-Q3 release. It should include some of the features (I guess some of the FOW features might still apply) + a rough target fillrate and clockspeed. Market it as Rendition’s new revolutionary technology (embedded DRAM) etc. etc. As soon as the decision has been made for the V3k a press-release should be issued regarding the "imminent" release of boards using Rendition’s "third generation" 2D/3D chipset. Once again include features and fillrate and some other benchmarks. I’m no big marketing guy so I’m sure others ahve much better ideas but I feel something needs to be done … However I know Chris has ideas that he wants to get approved so it’s not all lost I guess.

Oh, I totally forgot to ask him about drivers and the FFVII patch. I'll see if I can catch him later with those questions.

Anand heeft op de frontpage wat info over de nieuwe K6-400, die hij momenteel aan het benchmarken is:

I'm about to start the K6-2 400 review, I finished most of the testing necessary for the review, and I've found the new features of the K6-2 400's core quite interesting. Overall performance is improved approximately 7% over the original K6-2, that is in addition to the clock speed increase. Unfortunately the system isn't stable enough to run as many comparison tests as I would like, however it is enough to illustrate that AMD did in fact improve the core of the K6-2 with the 400.

Verder nog wat stuff over de dagelijkse handel en wandel van Anandje, waar we nu geen interesse in hebben ;)

Tom en Patrick van THG hebben een PC Shopping Guide in elkaar geknutseld met info over welke harde waren je wanner moet kopen. Vooral handig voor mensen die een nieuwe bak willen kopen (hmm...wie doet dat nou?).

Anyway, hier weer een stukkie ctrl-c, uit de intro:

In the last time I got vast amounts of email with questions or recommendation requests. As you can imagine we spend a lot of time doing our normal work, which often doesn't leave enough time for answering emails. Thus I have to apologize first that many emails cannot be answered.

Christmas is near once again and this is the classic time for many users to buy a new computer. Maybe you know this situation: Working with your two year old computer becomes boring and your new graphics software as well as some new games just don't run very well. It just seems as if the PC had become slower over night.

Many of our readers have decent computer knowledge and improve their system's performance by upgrading some components, e.g. the CPU or the graphic card. But still more users don't want to put their hands on their computer; they will rather go for a completely new machine.

That's why this article is split into two major parts: Single hardware recommendations for most relevant computer components and complete buying recommendations depending on your wishes and requirements. Always keeping an eye on the market, its developments and innovations isn't always possible; so I will try to sum up the most important facts to make it as easy as possible for you to find the right components or the right computer.

Andy heeft een review gepost van STB's welbekende TNT kaart, de Velocity 4400. Hier maar weer een stukkie koppie&peest:

The TNT, despite nVidia's initialial downgrading is still an impressive beast - all the standard feature set are supported the highlights being 180M pixels a second fillrates, 1 pass bi-linear and tri-linear filtering (performed by the twin texture engine), AGP x2 sideband support, 32-bit Z buffering. The stencil buffer is also exciting developers like ID Software who are currently making very enthusistic noises about the TNT right now. STB add to the potent mix TV Out support, which is actually quite a useful feature for those of you with a DVD Player, as the TNT makes a wonderful DVD accelerator when used with software decoding. TV Output is supported by a display tab in the Control Panel which allows the selection of various display schemes for busniness apps, DVD Movie Playback (DVD does not conform to standard resolutions), and game play. The 2d features of the Velocity 4400 are excellent with features that were found on very highend cards only a year or two ago. This means the inclusion of the now standard 250Mhz RAMDAC which, when coupled with the 16 megabytes of RAM (more than my first PC!) allows resolutions of up 1920x1200 at 65536 colours.

CPU Central heeft een preview van de AMD K7 in elkaar gepleurd. Er zitten ook een paar erg vette foto's van het dingetje bij (hij lijkt erg op een PII).

Hier wat info dat ik eruit geript heb:

AMD claims that the K7 is being engineered on-schedule, and the will be able to deliver the K7 in the first half of next year at clock speeds starting at 500MHz. The K7 will initially make it's debut around early August or late July at .25 micron, and then (hopefully with no delays), move into a .18 micron process around October or November.

How much will it cost? At the moment, no one is absolutely sure. However, since the K7 will probably be able to outperform Intel's extremely expensive Xeon processor, the K7 definitely won't be too cheap. AMD could unfortunately not quote me a price, however, speculative prices are floating on the net that are between $700 and $900, depending on the L2 cache options you choose.

Gamespot is erg produktiek, nu al weer een review van Ubisoft's Monaco Grand Prix:

The newly introduced 'scenario' modes give you an objective that requires more than just the skill of fast driving. These represent a nice addition which gives the game more flexibility and long-term appeal. The 1950s mode is fun but little more than a graphical rehash and further jittery vehicle dynamics.

Riva3D heeft een review van KQ8: Mask of Eternity in elkaar weten te draaien.

Hier wat ik er uit geplukt heb:

he King's Quest series of adventure games has been around a long time now. Roberta Williams made her mark in the industry in 1984 when the first King's Quest title debuted, and was the first adventure game to utilize animation. The series also introduced the icon based interface to gamers in 1989 with King's Quest 5. Along with these innovations, Roberta has honed the art of an involving story mixed with player interaction to a high level of expertise. And now she sets out to put the series back on the cutting edge by moving the kingdom of Daventry into the third dimension. And by the looks of the preview version of King's Quest 8: Mask of Eternity, Roberta Williams has once again delivered the goods. The storyline of King's Quest 8 unfolds with the destruction of the Mask of Eternity, the holiest of all things Daventry, by none other than one of the protectors of the relic. The storm released by this destruction sweeps across the land, turning all mortals to stone. All that is, except for an adventurer named Connor, who luckily has one of the 5 pieces of the Mask land at his feet just as the tempest reaches his village. He is thereby protected from the evil spell, and soon discovers his destiny: To find the other 4 pieces of the Mask, and restore the Kingdom to its former glory.

Gamer's Depot heeft een review gepost met een vergelijking tussen de Diamond Monster Fusion en de Quantum3D Raven.

Hier een stukkie:

From the benchmarks we can see that both cards are good D3d (Direct3D) performers, as seen by the Forsaken scores, but thanks to the fact that the Banshee chip only has one Texel processor, any game like Quake2 is going to perform kinda lame. It�s very apparent that after the headache and frustration caused by the Voodoo Rush chip last year, 3dfx has made strides in trying to make the Banshee what they wanted it to be. A good, low cost chip to bring decent performance to the average user who doesn�t care about hitting 100FPS+ in Quake2. This isn�t however, really a chip I see the hardcore gamer would want in their system.

Accolade heeft een demo van TD5 gereleased die je bij Avault kunt downen. Hier wat info over wat die 10,8Mb doen:

he clash of modern super cars vs. muscle cars from the past continues in the sequel to the best-selling Test Drive 4. Test Drive 5 features twice the cars (28) and three times as many tracks (18). The addition of new features like split-screen mode, branching roads, and breathtaking photo realistic graphics make Test Drive 5 the ultimate racing challenge

Mark Rein van Epic heeft een vette posting op het Epic Megaboard gepleurd met info over de Riva TNT in Unreal:

Nvidia has been hard at work analyzing their D3D drivers to find out why Unreal's performance on TNT has been less than stellar up until now. The symptoms of Unreal's problems on TNT have been visual "stuttering" as well as serious control lag problems which causes mouse and keyboard inputs to be very sluggish.

Well the good news is that Nvidia believes they've located the reasons for these problems and that they'll have solutions soon in the form of new drivers.

Here is what David Kirk of Nvidia had the to say about Unreal on TNT:

"Unreal uses several texture download paths in the TNT driver that were unoptimized (they went through generic code). Those paths are optimized now (hand-coded, special purpose code). The reason that they were unoptimized before is that they were rarely used by other games, and certainly not frequently and dynamically during gameplay."

There may be other TNT driver releases between now and the time these particular optimizations are released but we'll make an announcement once the proper drivers are available for download.

I want to thank TNT users for their patience and Nvidia for their diligence in keeping at this problem until it gets fixed.

On other TNT related news: Creative Lab's told us yesterday that the TNT stencil buffering real-time shadow patch, shown at Comdex Fall 98 last week in Las Vegas, will be released to Unreal users. Right now the code is still in an alpha stage and they have some work to do on it before it's ready to release. You can read more about this and see some screen shots at http://www.unreal.com/comdex98/main.html or PlanetUnreal's mirror of this site at: http://www.planetunreal.com/index.asp?features/comdex.shtm

Gamecenter heeft een preview van Westwood's Lands of Lore 3:

At Lands of Lore II's release, Westwood really took it on the chin for shipping a product with dated graphics. Surprisingly, it is once again making a calculated risk in this area. At a time when polygon 3D characters are the norm, Westwood is sticking with pseudo-3D voxel characters. Castle explains why: "With voxel characters we can have 10 to 20 characters on the screen at a time. It was a question of visual fidelity versus performance ratio."

Happily, Westwood's prolonged experience with voxel technology has paid off. Now, characters are nicely rounded and multiple-sided. Best of all, though, is Lore III's new real-time lighting system. This alone has made a miraculous difference in the game's overall look since it was shown at E3. Six months ago, we would not have praised Lore III for its eye candy; now, it's actually quite pretty.

The Lands of Lore series has always been aimed at the first-time RPGer and at the gamer who wants to be immersed in a rich fantasy world. Lands of Lore III will satisfy both camps, but it also has a lot to offer players who think they've "been there and done that." Prepare to journey to the newest Lands of Lore in March 1999.

Gamespot UK heeft een preview van Mechwarrior 3. Hier een stukkie:

MechWarrior 3 should really be called MechWarrior 3.5. It isn't a giant robot step up in quality from the other successful MechWarrior games ­ it's a soaring jump-jet leap. The leap includes deformable terrain: the buildings have destructible geometry, incoming artillery rounds pockmark the ground with craters, and even the Mechs leave distinctive footprints. Similarly, the artificial intelligence has been improved considerably: it can ambush you and use the terrain against you. And the command interface is considerably more friendly: you can use pop-up windows to position your subordinates (who are significantly 'smarter' than their predecessors) and direct them into action.

Na de overname van Netscape vragen wij ons allen af wat AOL en Netscape gaan doen met het Open Source gebeuren rond Mozilla 5 (die bdw op dit moment nog zwaar instabiel is).

The Register heeft een artikel met info:

AOL's acquisition of Netscape has thrown the latter's commitment to open source software into doubt and could provoke a battle royal over just who owns its browser technology.

Back in January, Netscape finally conceded that, with Microsoft shipping Internet Explorer free of charge, it would have to do the same for Communicator. It also made the cunning move of announcing it would release the source code of Communicator 5.

The plan allowed users free access to the code and the right to tinker with it, provided any modifications, amendments and new features were submitted back to Netscape's open source admin team, collectively known as Mozilla, for verification and incorporation in the next release of the source alongside the work of Netscape's own coders.

It's the way Linux is now developed, and so Netscape was able to win even more anti-Great Satan of Software kudos than it had already.

That was then. Now, AOL technically owns the Communicator source code -- or, rather, it will do once shareholders and regulators have given the deal the go-ahead -- and may prove rather less keen on open source than Netscape.

For the moment, Mozilla-ites reckon they're safe -- as one member, Jamie Zawinski, wrote on the group's Web site: "What if AOL hates open source? The Mozilla code is out there, and it cannot be recalled. It has been distributed under an open source licence, and nobody can ever take that away from you."

True, but the "open source licence" only appears to refer to Communicator Standard Edition 5.0 -- there's arguably nothing (but Open Sourcers' opprobrium) to stop AOL pulling version 6.0 (Navigator 2000, anyone?) back into the fold. Of course, open work on 5.0 could continue, but it's hard to see it retaining widespread support once it's been superseded by a probably more advanced version.

Given AOL clearly has its eyes on wider markets than PC-based Web browsing (see Netscape deal founds Sun-backed 'AOL Everywhere' scheme), it's highly likely AOL will want full control of the software that's likely to form a large part of the interface to the new devices that plan is founded upon.

At the press conference, Sun and AOL said they will both develop the next generation of Navigator and Communicator clients, which suggests that it will, at least initially, be pulled out of the open source arena.

Of we daar zo blij mee moeten zijn :(

AOL suckt stevig en ik hoop dat door de overname Netscape ook niet meegenomen wordt. Wat moet je dan? Overstappen op Opera (best leuke browser, v3.5 is net gereleased)?