Slashdot Mirror

* resizing text works, in fact you can easily resize/zoom the entire flash movie. I never have a problem with unreadable flash files,

It barely works. The word "easy" is not valid here. If a page is made with flash and tiny fonts, I can individually right-click each Flash area to zoom in the size by 2x. The amount of space allocated on the page doesn't change, so now I"m stuck dragging it around with the mouse to see the whole thing. Is there a way to resize a flash within a page, not just zoom in on it? I can't find one- and it certainly doesn't automatically bind to the browser's text magnification setting, which will be an advantage of Mozilla's native SVG.

* That key intercepting thing was something I hadn't realized, and I agree something should be done about it. But it's something I personally never encountered, and is a problem with the implementation not the format.

The flash format allows the author of an SWF to intercept keystrokes for his own use.

if google can parse PDF's, why not SWF's, it's not like it's that much harder...?

Google makes a mess of many PDFs, but at least all PDFs can evaluate somehow to a linear sequence of text. Flash programs can be arbitrarily complex, in a Turing-complete way. For computer scientists, "Turning Complete" is the trivial way to demonstrate that something cannot be comprehended by a computer program. (Postscript is also Turing complete, but it's non-interactive, so non-degengerate files will complete after a reasonable time). If someone makes a website with 20 pages all in a Flash file, a web-crawler has no sure way to tell how those pages are linked relative to each other.

Maybe it can find an interesting text string inside the SWF, but then what? Does it print you an ascii dump of the file in its cache? Or just take you to the live SWF itself, where there will be no guarrantee that you can even navigate to the desired page within the flash?

(Some content providers will want to use Flash for this reason- because it makes deep-linking and other 'subversive' user techniques more difficult)

There are other problems with the Flash format, in that they enable or imply bad behavior by the implementation. Primarily, the assumption that Flash files are an executable program, and that they are constantly advancing through time (and animating as they go). Some websites (like Tom's Hardware will have 4 embedded Flash files per page. Playing them uses a significant chunk of CPU power, even if you're not viewing the page. But if someone tries Mozilla's tabbed-browsing feature to keep reviews of several different products handy for comparison, the CPU consumption just keeps climbing and climbing.

Other problems with Flash are purely with the implementation- I can't easily turn down the default rendering quality, for example (or just set them to not even animate until I click on one). But Macromedia has no desire to improve the end-user's ability to control his view of SWF files. To do that would make them less attractive to advertisers. "If consumers could de-activate the constant bouncing, they might be able to focus on content they want to read!"

You need at least a R9600/FX5600 to play at a descent frame rate.

There 's a nice article on Tom's hardware.

You need at least a R9600/FX5600 to play at a descent frame rate.

There 's a nice article on Tom's hardware.

You need at least a R9600/FX5600 to play at a descent frame rate.

You are partially correct. Intel did make a Pentium 4 1.3GHz. As for performance, you may be right. I could not find any direct reviews of those two processors going head to head so it's hard to tell. Indirectly we can compare the Celeron vs Duron then the Duron vs Athlon then Pentium 4 vs Athlon. But this only underscores my point, even though we both speak the jargon, there is confusion as to the understanding of the subject.

You are partially correct. Intel did make a Pentium 4 1.3GHz. As for performance, you may be right. I could not find any direct reviews of those two processors going head to head so it's hard to tell. Indirectly we can compare the Celeron vs Duron then the Duron vs Athlon then Pentium 4 vs Athlon. But this only underscores my point, even though we both speak the jargon, there is confusion as to the understanding of the subject.

. The Athlon is comparatively simple, and the 400MHz FSB is overkill.

Uh-huh. Thats why AMD does sooooo well in memory performance.

Now that all of this has been released, I wonder if we will be seeing alternative firmwares with support for new features (detailed external logging, radius server, wireless VLANs like the cisco APs, traffic shaping, oh, and MeshAP could be cool too.) A friend of mine already bought his WRT54G and likes it, but after I found this out (and submitted the story to slashdot) I ordered mine from Amazon($130 with free shipping) along with the NetGear WAG511($85 - 802.11a, 802.11g, 802.11b) as recommended by a Toms Hardware review. I'm so excited that I am going to have a dope 54mbps wireless network in my dorm room for only $215. A little excessive, maybe, but hey I will probably keep this for another 5+ yrs. Especially the 802.11a 802.11g cardbus card. Oh hey, if you are shopping, check out the SeattleWireless Hardware Comparison. They have all sorts of info there and it made it easier to decide what to buy.

THG did a nice rundown a while ago on (still-)existing audio chipsets on Mobos and sound cards, comparing bells&whistles, CPU usage and IIRC quality.

Cheers.

You mean this wasn't a Tom's Hardware review?

Ahhhrrrrggg. Its too late for me. You can read more here at my blog. I had to replace my power supply just a month ago. I paid $75 with shipping and tax for a Zalman ZM300A-APF which is high on the list you linked to. There's nothing custom about it, unless you mean how its fan speed is thermo-controlled. At least its silent relative to my Noise Control Silverado. I could have paid half of what I did...

I have a Noise Control Silverado on my 1.2GHz Athlon. That cooler generates only 38Db in quiet mode, which I'm using, and its STILL more noisy than I would like. My machine is 1 meter away from me and when I'm not listening to music, I want silence. I want to fall asleep without noting how loud the computer sounds 2 meters away. Sure plenty of people are used to sleeping with their jet engine on, but I'm not and I don't intend to get used to it.

The next step in my quest for quiet is adding sound absorbing material the the inside of the case. I've been considering vinyl matting used as walkway in buildings. It looks and weighs like rubber gasket material available in 1 foot wide rolls at the hardware store. It does have ridges on one side, but its much cheaper, about $17 for a square yard instead of $5 a square foot for the gasket rubber.

I have a Noise Control Silverado on my 1.2GHz Athlon. That cooler generates only 38Db in quiet mode, which I'm using, and its STILL more noisy than I would like. My machine is 1 meter away from me and when I'm not listening to music, I want silence. I want to fall asleep without noting how loud the computer sounds 2 meters away. Sure plenty of people are used to sleeping with their jet engine on, but I'm not and I don't intend to get used to it.

The next step in my quest for quiet is adding sound absorbing material the the inside of the case. I've been considering vinyl matting used as walkway in buildings. It looks and weighs like rubber gasket material available in 1 foot wide rolls at the hardware store. It does have ridges on one side, but its much cheaper, about $17 for a square yard instead of $5 a square foot for the gasket rubber.

Soundproofing your case produces results, but the biggest bang for the buck is definitely replacing the PSU.

I built a TV PC and I was annoyed by the hovercraft-like PSU, so I invested in a silent PSU. There are lots of custom quiet PSUs recommended here, but you pay for the styling and mods.

For me, the most economical approach was to pay $50 for a standard Fortron/Sparkle PSU with inside-case 120mm fan intake. There's a review of it at Tom's.

After the PSU replacement and replacement of the PC case, the PSU is literally inaudible. The loudest ambient noise in the apartment now comes from the fridge compressor in the kitchen one room over.

It's a minimal system though, an underclocked XP2400, a single hard drive. If and when I put in some more drives, I may line the case with soundproofing...

...get a quieter fan?

I'm recently went from wired to 802.11g. However, it wasn't without a struggle. I did a good deal of research but still got suckered into buying a Broadcom-based card only supported in Windows. As it turns out, Broadcom doesn't support Linux well (Or at all, in this case). To add to the confusion, most of the cards that I checked out that had once boasted Linux compatibility had been 'upgraded' to use a Broadcom chip. Even 802.11b hardware that used the supported Prism2 chipset is damn near impossible to find these days as much of it has been changed over to use cheaper hardware (Not necessarily Broadcom, but other non-supported brands as well). Model names / numbers are virtually the same as they were before. It's basically like searching for a PCI non-Winmodem these days.

My advice: Go with a nice ethernet bridge and don't get burned by bad / non-existent drivers. I ended up with a Linksys WET54G, which just so happened to be reviewed by THG earlier. It works flawlessly after I plugged it into my NIC under Linux. It also leaves my options open for other OSes that don't even have as much support as Linux. So long as your network card works (And interconnects via RJ45), you'll have a reliable wireless connection using the bridge. Not only that, but it has a configurator accessable through any web browser, much like their routers. This means configuring the bridge for use with encryption and such will work the same on Windows, Linux, MacOS, etc.

Only problem is they're a bit expensive (Roughly $130). if you don't use Windows full time, it's worth every penny.

I sort of want one. A screen like that should get me through a year or two of college...

But I will wait till they ship with the Athlon64. There's really no sense buying 32-bit hardware right now, unless you just get some cheap desktop solution like a 2500 Barton to keep you going untill 64bit is affordable.

btw, if you haven't seen them yet, THG had posted the first pics of the Athlon64

Just serves to make me more curious what Intel has up their sleave.. I wish they'd be more open about it.

Article Info
Building A Home Network From Scratch
Created: June 30, 2003
By: David Stellmack
Category: Networking
Summary:No matter if you choose a wired or wireless network, planning and flexibility are the keys in getting a sucessful installtion that you will be happy with. We were lucky enough to select the right contractor who helped us achieve our goals and exceed or expectations. The result was a state of the art infrastucture that we will not outgrow anytime soon.

Intro (image)

The Ins and Outs of Building a Home Network

With interest rates at an all-time low in the U.S., many Americans are now able to build or buy their first home or, if they are already a homeowner, to upgrade to something better. Buying or building a home offers those who are technology-minded the opportunity to consider the integration of home connectivity up-front. Many new, custom-built homes now offer some level of basic home connectivity as an option; but this is a new phenomenon. Considering that the Internet has become a mainstream feature of our collective lifestyle in just a few short years, unless the home you are buying is fairly new it is likely that you are pretty much left to your own devices when it comes to networking.

First, you have to consider the pipe into your house. Normally, your choices are either DSL (Digital Subscriber Line) or Cable modem; but as many have discovered to their dismay after they have signed on the dotted line of the purchase contract, neither of these options may be available. Generally speaking, the further the home is located from a metropolitan area, the fewer options for Internet connectivity there are. While satellite and ISDN options do exist, they are probably the least desirable solutions.

While we don't suggest that Internet connectivity options should outweigh other factors in a home purchase, if high-speed Internet access is important to you, then it is wise to do some basic investigation as to what options are available at your home location. An hour spent on the telephone with local cable and Internet Service Providers (ISPs) may make the difference between satisfaction and regret with a home long after the sale has concluded.

Wired or Wireless?

No discussion of home networking would be complete without a brief discussion of wired and wireless solutions. While wireless solutions offer perhaps the most convenient solution for many home computer users, they may not present the best solution.

"Wireless" is the current buzzword these days, and it seems every hardware product vendor is touting his or her particular ability to provide it. The word "wireless" evokes visions of being able to wander around various rooms in your home or even outside without the network line that tethers home users to a single place. However, depending on the version of wireless you select and/ or are able to afford, your results might not measure up to "manna" grade, or even a mediocre grade, over time. The physical construction, design and layout of your house can affect whether or not wireless is a good solution, and you may have to invest in multiple access points (in infrastructure mode) linked together via Ethernet to provide coverage for wireless access to the various locations within your house.

If wireless coverage isn't an issue, a more important factor might be connection speed. From our testing, many 802.11a 54-Mbit solutions certainly don't have the amount of signal penetration that is required for wireless access in many multi-story homes and office environments. For example, if you have computer systems in the kids' room on the second floor and your family room is located on the first floor some distance away, you may or may not be able to achieve a connection from both places, depending where the access point i

Fortunately, the friendly folks at Hardocp have stepped in to provide some legal assistance for the guy, to make sure he is not railroaded into pulling down his editorial describing THG's media shenanigans.

More bullshit drama from HardOCP. Last week, it was accusations about Futuremark. This week it's about some supposed media blackout, and accusations that THG might be suing some other website. I mean, come on. Read THG's news page. ABC News and WHAS 840 radio even did a broadcast from the event. Does this sound like a "media blackout" to you?

"as you can see from the coverage on www.sudhian.com, any other press that came along was treated respectfully and was not hindered in any way. You can also visit the organizers at www.lanwar.com, and you will see clearly that there was no media blackout. In fact, I personally recommeded to the organizers that they get as much as press as they could. Local press such as WHAS 11 TV was there."

And what business is any of this of HardOCP? At what point did Hard|OCP stop being a hardware site, and start being a fucking soap opera? I've seen twelve year old girls that are less into gossip and rumors.

I was a regular [H] reader until today. This is ridiculous. Check out what's on the front page.

More bullshit drama and conspiracy theories. This time it's about some supposed media blackout, and accusations that THG might be suing some other website. (a completely baseless accusation)

At what point did Hard|OCP stop being a hardware site, and start being a fucking soap opera? I've seen twelve year old girls that are less into gossip and rumors than this.

I know that people here seem to hate Tom's Hardware, but this is a
pretty good article:

http://www4.tomshardware.com/howto/20030428/index. html

-ft

I take everything I read with a grain of salt these days --THG, Yahoo!, Google, /. and the zillions of other articles I read online included. I think you have to in the world we live in. I still consider THG a useful source, thier reviews are more thorough and detailed than 90% of the other hardware reviews I find online. So what if so-and-so might be in bed with so-and-so on occasion...unfotunately, it happens all the time in big business and media. We'd all be hard pressed to find a resource thats completely ethical as well as credible ALL the time.

WBGG

Check Tom's Hardware. You might be better off on the nVidia side with a Ti4200, the same or better performance as the 5600 but cheaper.

What I look for as a consumer is this - a head-to-head comparison of several generations of cards. That's where you can find the sweet spot.

You have to take benchmarks like this in context and with a grain of salt.

WBGG

WBGG

WBGG

WBGG

As is already well known, Springdale and Canterwood chipsets are completely identical during wafer production. Intel in fact makes a preselection, which then leads to different packaging. Intel explains this step with the following: only when a chip does not pass the PAT test does it become the lesser-value Springdale (provided that all other quality tests have been passed).

From the beginning, Intel put safety precautions into place that would ensure that the board manufacturers couldn't easily activate the fast memory access (PAT) with the less expensive 865 chipset. The price difference with the chipset alone is at least $16, which results in a difference of up to $55 per board for the end user.

No. They can't. In fact the article clearly states that they were pleased to reach over 1,000 attendees.

But who really cares if they can or can't hold X amount of people. I just want their Network Equipment when they are done

No. They can't. In fact the article clearly states that they were pleased to reach over 1,000 attendees.

But who really cares if they can or can't hold X amount of people. I just want their Network Equipment when they are done

And for those wondering how to activate it in 865PE, there's an article at Tom's Hardware.

Tom's Hardware has the latest and greatest Pentium 4 3.2GHz running Quake III at 450 FPS. So this computer is almost 5 times slower than the top of the line, at least as for as OpenGL/Quake/games go.

Unfortunately the more egregious benchmark was the Quake benchmark.

1. Tom's was using Q3A 1.16 instead of 1.32 (PunkBuster code is thought to be a little slower)
2. Tom's used set s_initsound 0 to disable sound while Apple noted default settings, which would imply sound was left on.
3. Tom's used demo_001 while Apple used demo_4

You mean like the G5 Quake3 benchmarks on this page?

337 FPS? is that all?

Looks like you just barely managed to beat the low end hyperthread P4

It's hard to tell from that Apple page what tests they ran on Quake3 Arena.

However, Tom's Hardware ran Q3Arena on a P4 3.0GHz/Radeon9700Pro at 1024x768/32Bit and the
result was 402.9 FPS and not 275 FPS as on the Apple page.

The Tom's Hardware review is available Here

The reason I don't know what kind of card I have is that, although I purchased a brand new, shrinkwrapped Radeon 9500 Pro retail box, I noticed that the card looked somewhat different from the OEM (I think Sapphire) 9500 Pro I'd used for a few days previously. Didn't think much of this until I started seeing indications in places like the XFree86 log that the card might be a 9700, instead of a 9500 Pro (I do realize the two cards both use the R300 NE chipset). Also, that Antalus flyby score is meaningfully higher than what Tom's Hardware found for 1280x1024x32 on a much faster system than my two-year old Athlon 1.4GHz. Haven't bothered to crack the case open again since, so until I do I'll happily enjoy the illusion that I somehow ended up with a 9700 for the price of a 9500 Pro.

The driver I use is a binary-only one from a German reseller. They appear to be betas of forthcoming versions of ATi's own drivers. As I have XFree86 4.3.0 this has been a great blessing, as ATi's own Web site only has drivers for XFree86 4.1 and 4.2.

Erm...that's not all that surprising, seeing as how the 5200 is supposed to be the budget version of the FX line (analogous to the MX in the earlier generations), whereas the Ti was the top-of-the-line version of the GF2. Even though your 5200 had 128 MB RAM, it's still bottlenecked by the intentionally crippled chip.

Don't blame the company because you didn't do your reading.

It was in an article on Tom's Hardware Guide. An impressive display of how Intel and AMD processors responded to loss of heatsink. AMD's processors instantly turned to toast, while P3's locked, and P4's slowed down to reduce heat output until the heat sink was put back on (a really cool and useful ability). Since then, AMD's new processors will shutdown (similar to the P3's response I think) if on a motherboard that supports the ability. I'm not talking about the little heat sensor thing most MB's have, which couldn't respond to a 300 degree increase in temperature in 2 seconds.

For those who care, there is also a comparison of AMD 3200+ to P4 3.2 GHz at tomshardware: here

Are. you. licking. frogs? Again?

here is a nice picture that will give both a current and historical perspective on AMD and Intel CPUs.

Please note that except where specified, the Intel processor wattage is given as an average, while the AMD wattage is their maximum thermal load.

The current top Athlon dissipates 76w max while the current Pentium 4 dissipates 82w average.

The Pentium 4's max is 101w.

Just in case you missed it, the Pentium 4's maximum heat dissipation is 101 watts.

Now that we have solved that matter, the reason the Xeon does not do to well in chess is that fritz is a single threaded floating point intensive task.

When the Pentium 4 architecture can not take advantage of Hyperthreading, it's floating point performance is quite marginal.

"We" have not cleared anything up. Shut your touchy, feely, non-researched, "quite superior at doing tasks that are very mundane and repetitive," patter up. You are spewing marketing crap like you own shares for god's sake.

Opteron is not innovative but it will take a 3.8 Ghz Xeon to match its performance. Remember Opteron? No, I know, you have been trying hard to forget. Shut your pie hole before any more hallucinogens fly in there without you knowing.

Tom's had an article that showed that WineX was within a couple percent of Windows, and it was emulating the Direct3D calls.

By a couple percent you mean 50?

That article shows that X has comparable speed, sometimes a percent or two slower, sometimes a percent or two faster, than Windows 2000. The article also shows that the same games running under WineX are significantly slower than under windows. It goes on further to show that the slow down is not related to a delay in displaying graphics, and thus X is doing good, it's just WineX which is providing a bit of a bottleneck. This was one year ago, WineX may have improved since then.

I will admit to ignoring the problems of X being a network protocol rather than a graphics one
I, however, will not. X being a slow network protocol is largely a myth. Closer to the truth is that X *can* be a slow network protocol, but with the extensions in recent versions of XFree86 and a video card with a good driver (e.g. nVidia) X can be just as fast as Windows. This month's Linux Magazine has a couple acticles about it and one claims that in some benchmarks (on a local display) X is *faster* than Windows. And last year, Tom's had an article that showed that WineX was within a couple percent of Windows, and it was emulating the Direct3D calls.

I believe Microsoft did not use the MIT-licensed code for their version of Kerberos: post

It is interesting that Microsoft wrote their own Kerberos even when code already existed with a BSD-style license.

This prevents M$ from stealing his work, ...

With the direction Microsoft is taking with file systems, I strongly doubt they will want his work.

Worse than that, THG can't even get their facts straight. For example, when discussing fsutil.exe on page 4, the caption of the picture calls it a DOS app (it's not) and say it's from Sysinternals (perhaps they meant ntfsinfo, like the picture shows), yet the article text properly calls fsutil a "command line utility" (which it is) from Microsoft (which it is). While they do mention that it works on XP and not Windows 2000, they don't bother to mention that it's also available on Windows Server 2003, and that it's a system utility that's installed with the OS (c:\win[dows|nt]\system32\fsutil.exe). And just to add insult to injury, the "fsutil fsinfo" command they suggest you run is not quite correct. You need something more like "fsutil fsinfo ntfsinfo c:". "fsutil fsinfo" by itself just gives you another help screen, and not "scads of fascinating statistical information on the file system, volume and MFT."

All this article does is reinforce my dislike for Tom's Hardware Guide, and gives me ammunition I can use to convince others that THG is crap, too. If you want good hardware reviews, go somewhere good like AnandTech or Sharky Extreme. Hell, you could even go to Blue's News for the daily Hardware Reviews and still get better info. (I've not once seen Blue's link to THG from the Hardware Reviews ... I wonder why?)

While looking into the differences of 802.11 a|b|g I found this article over at tom's hardware. It appears that the second generation 802.11a protocol devices now have much better range than the previous first generation 802.11a devices. This can also be seen by a recent whitepaper by Atheros (The company whose chipset is used in most widely available wireless devices from such companies as Netgear, Linksys, and D-link). With better bandwidth performance than most 802.11 b|g devices on the market and equivalent range to b|g devices, all while operating in the 5 ghz range, perhaps 802.11a will make some what of a comeback. The potential really comes from the fact that right now you can buy wireless APs now from Linksys (WAP55AG) and D-Link(DWL-7000AP) and wireless cards from linksys,netgear,and d-link that do all 3 protocols. I know that right now I will be looking at running my AP in 802.11a mode and not worrying about interference coming from the 2.4ghz range.

See Tom's Hardware for an article on D-Link's Repeating WAP.

If the cooling fan fails, only AMD is toast, Intel's P4 can slow itself down (P3's crash but not burn) when overheated. Haven't you seen THG videos? Also, the current motherboards will power down when AMD CPUs go over a certain temperature, although you'd loose any precious unsaved data.

Then again, it's been a while since they made the video. I wonder if the current Northwood will crash and burn. Anybody dare to try? :)

I do have to agree though, another fan in the case is another breeding place for the dust bunnies.

That article showed that lower latency doesn't mean higher bandwidth (and this is only true if your original latency is low enough, mind you!), but it didn't consider overall performance. Latency has indeed an impact on the performance -- look at Tom's Hardware article on performance improvements when Intel's PAT is enabled. All PAT does is lower latency by 2 cycles.