CataclysmCow

I'm guessing you have a pre-assembled computer? Companies like Dell, Compaq/HP, Gateway, etc. often get specialized versions of hardware for use in their machines. It's not uncommon to find a piece of hardware in your machine that doesn't match with the specs on the manf's site. I think a recent one was where Dell got their hands on some early 7200.8 Seagate's with 16MB caches. Diagnostic software, Seagate's website and Dell's specs didn't all line up.

Doesn't matter which SCSI ID he uses as long as it's not the same ID as the host controller. Non-wide can use 0-7 for IDs and wide SCSI can use 0-15 for IDs. I haven't found any rhyme or reason to the default ID used by the host controller so you'll have to look for it. Doesn't really matter though if he's already got it working. Just change the boot order in the BIOS. You can have a practically unlimited combination of ATA, SATA and SCSI devices.

Don't give "paged to disk" a bad name - it doesn't deserve it. It's being paged to disk because it's not being used. If it's not being used it can't be slow (or fast) can it? I agree - completely. I futz about with my boxes all the time. I've done some really stupid things too and still have been happy with the results (trying to recreate a typical *nix folder tree on Win98 for one). This isn't about tweaking or getting that extra smidgen of performance though. These are just downright stupid pieces of advice that accomplish nothing other than crippling your system. I know it sounds cl

Ok, I tossed my post through babelfish and this is what popped out: The chunk missing in those pictures is for a wire that no longer means squat when connected to a drive. The manf cut that part out to cover their own culpability of their previous products. Better?

"Cable select cables" only came in the 40-pin/40-wire variety. All 80-wire/40-pin cables are "cable select cables"; it's part of the spec from ATA/ATAPI-4 and up (ATA/ATAPI-3 was the first mention of 80-wire cable). Cable select (CSEL) is pin #28 which would be wire #55 on a 80-wire cable. Counting from #1 (red wire) I end up with the funky wire being #67. Wire #67 would end up being pin #34 which is PDIAG and CBLID. I'm a bit unclear on what role pin #34 played in pre-ATA3 setups, but with 80-wire/40-pin setups it's there to indicate the presence of a 80-wire cable; it's grounded at the

If you don't mind paying a little extra I strongly reccomend Lian-Li cases. They usually run $300 and up, but you can find previous years models for around $100 (new). They are the best made tower cases in the world with maybe the exception of one of Zalman's $450 cases. They're for for someone who appreciates workmanship, quality parts, functionality and clean looks. People who are fans of cases that look like they came out of a cheap sci-fi flick need not apply. Don't let the online pics fool you. These cases are stunning in real life. Pics just don't do them justice. Try and find a fr

The campus has several open point-to-multipoint bridges. A friend of mine who recently graduated from there worked in the computer services dept and setup a bridge to his house which is just a few hundred feet from the school. From his network there's another point-to-point bridge to my house. You could do it without using bridge mode, but it's preferable. Since he's moving though I'm going to have to find a solar powered option to put on that relay hill behind the school (and hope no one questions it). Yes, but it only really becomes an issue when you are dealing with longer links over a

I disagree with a lot of content on that site, but since his main argument is that disabling services is advantageous I thought I'd hit that one. This is the big mis-conception that he works off of. That disabling services will free up resources. This just isn't true. Even if you disabled all services you'd only free up about 1.5-2MB of memory. When a service is idle most of this memory is paged to disk anyways. Even if you were to free it up it wouldn't be accessable to applications/user-mode processes because services are allocated memory from the non-paged pool. CPU time isn't an iss

It's just a 802.11b link (PtP means Point to Point) from me to a friend near my old university (it's about 8 miles). For the cards I used 2 Senao NL-2511CDs (200mw). For the antennas I used 2 24dBi parabolic grid antennas (Pacific Wireless). For the platforms we used Soekris's Net4521 boards running Pebble linux. This is real overkill for such a distance and if you did something similar you'd could easily get away with less powerfull/senstive caqrds and would probably want something much less focused than 24dBi grids - they are a PITA to line up. The connection is being taken down right no

PG7, just curious. What DSL modem make/model do you have? Did the telco tell you that you had VDSL?

I love getting insider information, but do you know of any 3rd parties that can confirm this, PG7? When a bin goes bad manf's usually report it within a few days after the reports roll in. Don't think of this as a scar on Maxtor's reputation. Every manf has had their goofs. I personally LOVE Maxtor (I'm an Atlas fiend). I think if PC users just spent an extra $100-$150 on decent storage solutions they'd never have to worry about issues like this. A decent enterprise drive will easily run 24/7 for more than a decade and you very rarely run into these types of bin flaws. You can find a Ada

I thought VDSL had higher speeds than that? <{POST_SNAPBACK}> It does. MUCH higher than that. I think PG either had to many margaritas tonight or he got shammed by a telco salesman. (read my post above)

VDSL? You sure? VDSL is usually symmetric and typically has >8Mbps speeds. Not to mention you have to practically be neighbors with your telco to get it. AND it costs more than you'd pay for DS service.

Thank you for the opportunity to clean this up, Matt. Here's what I supplied to Matt and Jeff: I'd love to hear any other opinions on BV's main premise or any of his other suggestions.

So I make a post explaining that BlackViper's site is a bunch of hooey and will cause readers looking for a learning opportunity to take 10 steps back. 4 days later and my post is lost in an archive of threads and this thread still stands. I guess I'm just suprised that a site like BestTechie is supporting a site like BlackViper.

Where would be a proper place to discuss the complete ineptness, misinformation and flat out FUDness of BlackViper's site and suggestions on this forum? Not in this thread I'm guessing?

My T1 ISP: AT&T Cost: Paid for by an employer ($360/month) 1.544Mbps down 1.544Mbps up (real world numbers are more like 1.4Mbps) My Cable Broadband ISP: Adelphia Cost: $39.99/month 4Mbps down 768kbps up (real world is more like 3Mbps/500kbps) PtP WiFi link ISP: Me (my old university carries the physical link) Cost: Free after equipment ~8Mbps down ~8Mbps up (varies a lot, but I can usually keep it around 8Mbps)

In 80-wire/40-pin cable only the odd numbered wires are used for signaling. The even numbered wires are all grounded to prevent crosstalk on the other wires. By my count that's wire #67. A signal wire by the even/odd rule, but wire#67 is the one exception to that rule. A cut in that wire should have no effect. [edit: oops, counted wrong]

It's not every packet. The SSID is only included in management frames related to finding and associating with APs (probe request, probe response, association, reassociation and of course beacons). On a typical home network these packets are actually pretty rare. You only see them as a client comes online, times out or if there's a lot of interference. It can take hours/days to sniff out out of these packets on a small home network. The reason why it's so vulnerable is that you can force one of these management frames to be generated by forging a disassociation frame which causes the client

What's the deal with the censorship here? Weird thing to run into on an "independant" board....