Send News. Want a reply? Read this. More in the FAQ.   News Forum - All Forums - Mobile - PDA - RSS Headlines  RSS Headlines   Twitter  Twitter
User Settings
LAN Parties
Upcoming one-time events:
Chicago, IL, USA, IL 01/27
Chicago, IL USA, IL 10/19

Regularly scheduled events

User information for Tom

Real Name Tom   
Search for:
Sort results:   Ascending Descending
Limit results:
Nickname None given.
Email Concealed by request - Send Mail
ICQ None given.
Homepage None given.
Signed On May 21, 2000, 15:34
Total Comments 1539 (Pro)
User ID 4838
User comment history
< Newer [ 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 ] Older >

News Comments > Christmas Metaverse
12. Re: Christmas Metaverse Dec 25, 2012, 16:27 Tom
Frags4Fun wrote on Dec 25, 2012, 14:19:
Speed tests show that I am getting up to 25 Mbps but I never get those speeds on an actual download. I'm wondering if I'm getting screwed. Also, I've asked relatives who live within an hour of my location and they get the same exact numbers. Perhaps TWC throttles everything except for the speed test websites.

Indeed, those speed tests always show the full speed, yet real world results, like YouTube and game downloads/updates, are dog slow. Yet other people who are lucky enough to be on FiOS get fast speeds on those things. So what does this tell us about TWC? What we already know... they suck. The only thing they're interested in investing in is lobbying to preserve their monopoly status. They still haven't even upgraded to DOCSIS 3 here. Even Charter, arguably the worst ISP in the US, did that years ago.

If I had actually got the speed I'm paying for, the same speed I always get from speed test sites, it would've taken about an hour to download Mass Effect 3 the other day. Instead it took about 10 hours. I believe they are throttling and/or their infrastructure is just garbage.
Reply Quote Edit Delete Report
News Comments > Morning Tech Bits
9. Re: Morning Tech Bits Dec 22, 2012, 09:52 Tom
I apologize for the caps lock, as well as the improper (but perhaps amusing) use of the word "perpetrating". This is a major pet peeve of mine and it drives me crazy that the issue has been swept under the rug by everyone: SSD manufacturers, OS vendors, tech reviewers, defrag software vendors, everyone! And none of them ever bother to explain what's really going on. They just say "oh well defrag is pointless on SSDs, doesn't do anything except wear out the flash, don't do it, don't pay any attention to that, it doesn't apply, you can pretend it doesn't exist".

You mentioned that Windows disables its defrag on SSDs. This is not a good thing. This is an irresponsible easy out for MS. Over time, filesystems will become more and more fragmented (NTFS is VERY prone to fragmentation) and this problem will absolutely contribute to the "Windows feels slower over time" problem. Not like they're strangers to that though. Just buy a new PC. Buy a new SSD. Format+reinstall. Everyone's happy, right?

Even though the problem is absolutely solvable, nobody in a position to do anything cares. This is a sad state of affairs.

If you're interested in understanding the truth, don't just take my word for it. Here are good starting points - note that even Wikipedia usually fails to cite sources for claims related to this point, and even sometimes contradicts itself on this point within the same article:

Also, here's a tool you can use to easily measure fragmentation and defragment on a file by file basis:

I base my statements on a moderate understanding of how filesystems work (having implemented a simple one) as well as my direct observations. I have personally measured a 35% difference in time taken to read a downloaded file from an SSD whether it was in 150 fragments or 1, taking care to make sure the reads were not coming from the OS's disk cache. I encourage everyone to make their own direct observations, though I know it's not easy.
Reply Quote Edit Delete Report
News Comments > Morning Tech Bits
6. Re: Morning Tech Bits Dec 20, 2012, 18:51 Tom
Verno wrote on Dec 20, 2012, 14:14:
DangerDog wrote on Dec 20, 2012, 14:05:
I thought I read that you shouldn't defrag SSD drives

That's correct, there's no purpose in doing so and it usually conflicts with the drives internal garbage collection.

WRONG! Please stop perpetrating this myth. The truth is that filesystems on SSDs are just as prone to fragmentation as they are on traditional hard drives, and defragmentation has the same purpose on all drives: to eliminate fragmentation. The key differences with SSDs are that 1) the slowdown caused by fragmentation is not nearly as severe as it is with traditional hard drives (but it still exists!), and 2) there's a real wear cost to defragmentation. So things get a little complicated.

SSDs are much faster at reading data scattered around different locations, since they don't have to move a drive head over the correct portion of a spinning disk. But even on SSDs there can still be a performance penalty imposed by the overhead of fragmentation. This can be a measurable and sometimes even very noticeable impact when files are split into hundreds or thousands of fragments. Extreme fragmentation happens more often than you might expect, especially if the volume is used a lot while nearly full, and fragmentation breeds fragmentation. When reading a heavily fragmented file, the OS can no longer issue relatively few large, regularly sized read requests. It has to instead issue a larger number of smaller, irregularly sized read requests. If you know anything about SSD performance and benchmarking, you know that SSDs perform better (sometimes DRASTICALLY better) at larger I/O sizes. Now, good modern SSDs may achieve a comparable throughout on many small I/O requests vs. relatively few large ones, but this is hardly guaranteed. Chances are, if you actually measure, you will find a significant performance difference between reading a heavily fragmented file from an SSD vs. reading an unfragmented file.

Please understand that fragmentation happens at the FILESYSTEM level. The SSD has no notion of filesystems. It's a block level storage device that makes many small blocks of storage available to the OS. The OS then makes this storage available via a filesystem. Therefore, any garbage collection or wear leveling done by the SSD is not happening at the filesystem level and has nothing to do with filesystem fragmentation. The only way the SSD is even remotely aware of the filesystem structure is via the TRIM command, which the OS uses to inform the SSD that data has been deleted. The SSD can then take this information into account when doing its own internal garbage collection and preemptive erasing of flash cells. This improves performance of future I/O operations and reduces unnecessary wear on the flash.


It it worth defragmenting a filesystem stored on an SSD? Usually not. Not because "there's no purpose in doing so" but because it's not worth it from a cost/benefit standpoint. The cost of defragmentation is that you cause physical wear due to the many erase/write cycles caused by the OS trying to make things contiguous again via rearranging blocks, without the SSD understanding what's really going on. The benefit is improved performance, which may or may not be noticeable or significant depending on the situation.

In the future, it would be nice if OS and SSD manufacturers would make it possible for the OS to remap block addressing on the SSD in an intelligent, coordinated way. This would allow true defragmentation of the filesystem without wearing out the flash memory on the SSD. The SSD could then lazily rearrange data in flash as it sees fit in order to optimize future requests. But given how much trouble it was just to get TRIM (a relatively basic level of coordination between OS and SSD), and how readily the public buys these lies about the nature and effects of fragmentation, this may never happen.
Reply Quote Edit Delete Report
News Comments > Morning Metaverse
3. Re: Morning Metaverse Dec 19, 2012, 20:56 Tom
The solution is very simple. Don't go public!  
Reply Quote Edit Delete Report
News Comments > Morning Tech Bits
6. Re: Morning Tech Bits Dec 13, 2012, 23:31 Tom
Kitkoan wrote on Dec 13, 2012, 23:02:
What would really help give vectors I push I think would be something like an MMX technology that is just for vectors or even a x87-style chip.

What do you think a GPU is?
Reply Quote Edit Delete Report
News Comments > Morning Tech Bits
3. Re: Morning Tech Bits Dec 13, 2012, 15:13 Tom
Microsoft created a GUI framework called WPF (Windows Presentation Framework) that is vector-based and contains the infrastructure needed to make a real GUI without requiring bitmapped images. Yes, it truly supports vector-based icons (and other "doohickeys") and will scale to any resolution, using 3D hardware acceleration. However, for various reasons, WPF was doomed to an early grave.

Modern 3D hardware acceleration - even Intel HD Graphics - is plenty powerful to render an entire OS worth of GUI using purely vector graphics. We're just waiting for software developers to stop being lazy and make it happen. That's really all that is necessary.

With high-DPI displays finally going mainstream, there's starting to be reasonable incentive to actually do the hard work to make it happen. For example, Surface Pro is a 1920x1080 10.6" screen. A good old 16x16 icon is going to be !@#$ing small on that screen. But vectors are too hard so we'll just make 10 versions of the same icon instead. Great plan.

As for animating vectors, what is the problem? Games are animating (3D) vector geometry in far more demanding ways than a regular old GUI. Do you think 2D is slower to render than 3D? No, the problem is purely in how software developers are making use of the hardware. And that is the usual bottleneck that holds back true progress.
Reply Quote Edit Delete Report
News Comments > Morning Mobilization
19. Re: Morning Mobilization Dec 10, 2012, 18:36 Tom
Proper use of a GPS navigation device can be quite the mystery, apparently. Here it is in a simple and familiar format:

- entrust it with your life
- use it as your only source of info on a trip involving an unfamiliar route
- assume it will keep working for the duration of your trip
- blindly continue on when things don't seem right, hoping for the best
- drive into water, onto railroad tracks, off a cliff, etc.
- blame the technology for your failings

- doubt its every instruction
- PAY ATTENTION to signs and which way you're going
- plan ahead and find out at least roughly how to get where you're going BEFORE you start driving
- double-check its suggested route(s) against known landmarks, preferably prior to the trip using a regular map view AND satellite view
- know how to check its suggested route(s) using its mapping capabilities
- attain at least basic familiarity with how GPS and navigation devices work (it's not magic), and what their limitations are
- plan for emergencies. what if the GPS suffers catastrophic failure? what if the car breaks down? what if there's no cell or satellite signal? etc.

It ain't rocket science.
Reply Quote Edit Delete Report
News Comments > Dark Souls II Announced
9. Re: Dark Souls II Announced Dec 8, 2012, 16:32 Tom
Prez wrote on Dec 8, 2012, 16:01:
Hell, I had to download an easy mod just so I could survive the first sector in FTL.

What?! Come on, FTL is not that hard!
Reply Quote Edit Delete Report
News Comments > New Steam Greenlight Approvals
8. Re: New Steam Greenlight Approvals Nov 30, 2012, 17:45 Tom
EURO TRUCK SIMULATOR!!! If you haven't seen the "let's play" for the original one... enjoy:
Reply Quote Edit Delete Report
News Comments > MechWarrior Online Patch Adds Camo, Bitching
5. Re: MechWarrior Online Patch Adds Camo, Bitching Nov 27, 2012, 23:23 Tom
Great headline, Blue.  
Reply Quote Edit Delete Report
News Comments > Free Bridge Constructor
8. Re: Free Bridge Constructor Nov 27, 2012, 23:18 Tom
Bridge It is the one I remember fondly.
Reply Quote Edit Delete Report
News Comments > Intel Killing Off Desktop/Enthusiast PCs?
80. Re: Intel Killing Off Desktop/Enthusiast PCs? Nov 27, 2012, 20:31 Tom
PHJF wrote on Nov 27, 2012, 20:13:

... you are batshit insane.

It's an artform. One cannot help but marvel.
Reply Quote Edit Delete Report
News Comments > Intel Killing Off Desktop/Enthusiast PCs?
77. Re: Intel Killing Off Desktop/Enthusiast PCs? Nov 27, 2012, 18:20 Tom
AMD ain't the answer. They're trying to shift their focus away from CPUs for PCs. They haven't been competitive with Intel for a long time anyway. With even less focus / resources put into PC CPUs, I wouldn't expect too much out of them in the future.,18546.html

The future is Intel vs. ARM. And ARM isn't going to be in gaming PCs anytime soon, so that just leaves Intel.
Reply Quote Edit Delete Report
News Comments > Sunday Legal Briefs
10. Re: Sunday Legal Briefs Nov 25, 2012, 19:27 Tom
PHJF wrote on Nov 25, 2012, 18:12:
Everybody already has a cell phone.

Accordingly, most people are already in the state where their "movements and activities are monitored at all times". So, do these people realize they're being monitored? Do they accept it? Is it an ok tradeoff for these modern conveniences? As far as I can tell, the vast majority of people are either oblivious or ok with it. If they're oblivious, maybe some in-your-face monitoring might prompt some needed discussion. If they're ok with it, then cries of "but this is raising kids to accept it!" seem a bit dubious to me.

There is truly little privacy for those who lead typical lives in modern developed nations. Cell phones are just one of many ways to track people nowadays.
Reply Quote Edit Delete Report
News Comments > Out of the Blue
5. Re: Out of the Blue Nov 24, 2012, 18:45 Tom
Sepharo wrote on Nov 24, 2012, 17:38:
Uh, rambling aside, and back to my original question... Do you guys think Google Docs is a sufficient replacement for very occasional home office needs for someone who's only previously used Office?

It's not even a sufficient replacement for Notepad.
Reply Quote Edit Delete Report
News Comments > Strike Suit Zero Funded
5. Re: Strike Suit Zero Funded Nov 20, 2012, 21:39 Tom
Also never heard of it. The whole robot thing seems to change things a lot at first glance. Kinda like ... Thexder in space?  
Reply Quote Edit Delete Report
News Comments > Star Citizen Kickstarter Reaches for the Stars
64. Re: Star Citizen Kickstarter Reaches for the Stars Nov 19, 2012, 21:27 Tom
Dev wrote on Nov 19, 2012, 21:18:
ASeven wrote on Nov 19, 2012, 18:55:
$6M... Publishers have to be worried at this point since $6M is enough to produce a AAA game, without marketing.
It maybe used to be, don't think it is anymore.

I think it could still be enough. But they'd need to stay focused, use an existing and mature engine without too much modification, only target one platform (PC), not take too long (2 years tops), and spend the money very wisely.

Of course, with Star Citizen, they've promised so many difficult and expensive things that it will certainly take more than $6M. But they have more than $6M, so there's at least a chance.
Reply Quote Edit Delete Report
News Comments > Star Citizen Kickstarter Reaches for the Stars
61. Re: Star Citizen Kickstarter Reaches for the Stars Nov 19, 2012, 20:59 Tom
The "putting up walls" is probably referring to the wall separating backers from non-backers with regards to various information releases from here on out.

It's described in the last KS update here:
Reply Quote Edit Delete Report
News Comments > Star Citizen Kickstarter Reaches for the Stars
59. Re: Let Us Just Remember 3 Words: Nov 19, 2012, 20:11 Tom
SlimRam wrote on Nov 19, 2012, 18:37:
Wing Commander Movie....

Ok ok I KNOW that was a low blow but.....

THAT is what CAN HAPPEN when MILLIONS of dollars are being thrown around; I mean Chris Roberts WAS the director....

Jus sayin :)

Ouch dude, ouch! Yeah, that was a disaster, but ... Wing Commander 2!

SlimRam wrote on Nov 19, 2012, 18:37:
P.S. They REALLY need to bring back the Full Motion Video of the old games to really clinch this. I want to tell Maniac to fuck off/ not fuck off just one more time

"Where the hell are you guys?!?!"

We're still out here.
Reply Quote Edit Delete Report
News Comments > Croteam on Windows 8 Issues
30. Re: Croteam on Windows 8 Issues Nov 17, 2012, 17:58 Tom
NegaDeath wrote on Nov 17, 2012, 17:20:
-Performance improvements. I have a small 60GB SSD and 4 megs of ram (that I will not be expanding until mid next year) and Windows 8 takes up less of that. Nice.

Windows 8 runs with 4 megs of RAM? Now that's what I call a performance improvement!
Reply Quote Edit Delete Report
1539 Comments. 77 pages. Viewing page 36.
< Newer [ 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 ] Older >


Blue's News logo