• Content Count

  • Joined

  • Last visited

Posts posted by btb4

  1. And how do you (REAL) Americans feel about this?

    REAL Americans oppose government involvement in education and pay for their kid's education (like they do their housing, and their food, and their horseback riding lessons, and their Cub Scout camp outs and everything else that is a parent's responsibility to provide for their kids) at private school where of course any (or no) religious practices can be tolerated per the private group that runs the school.

  2. Funny, it does not have to be an actual scam (though it probably is). Consider that they are offering to sell you 400G for $600, and as a "special bonus" if you pay another $600 you'll get another 400G plus a "free" 400G. So, 1.2T of SATA storage for $1200? Heck, I'll sell you that much SATA storage for $1150.00...LOL.

  3. Since this does not seem to be going anywhere under the "experiences category", perhaps some questions in theory:

    These cases support 4, 8,or 15 spindles, and yet have only a single U160 connector.

    So, particularly for the 8 & 15 spindle models, is there much advantage to faster HDs? For example, 15 spindles of 300G Maxtor 5400 Maxline II can be deployed at or perhaps below the 250 Plus 7200s - think there'd be much difference in performance? The seek times are actually listed by the manufacturer as similar for both, and I tend to compare seeks when considering drives for RAID.

    Assuming all other things are equal, does drive cache size matter when in a RAID like this? Can it hurt, particulalry when considering the controller cache?

    These would not be server drives - I stick to SCSI for that - but I might want to use them for 3-4 simultaneous file transfers of moderate to large size, or perhaps to capture video. Of course, my questions about "better" do not necessarily encompass my intended use, as I am sure that 15 spindles of 5400 can capture video...LOL!

    Hmm, it looks like some of these device might actually be able to go to two channels, but for the sake of all my 29160s, let's stick to one channel.

  4. I spoke with Ken at Granite today and he had kicked my questions up to engineering. He thought I should have something back by the end of the week. He did mention that they are unique in having a license to allow them to write their own code for the Oxford firmware. Still, I have heard of this issue with 911-equipped ADS cases, too.

  5. I timed the transfer and it took 160 seconds to transfer 100 MB of data (5, 20MB files).  Thats approx .63 MB/SEC.

    Better (or simpler) test is to try it with a single 100MB file. May be a lot faster.

    Maybe there is a networking issue here, maybe not. Your source machine is not too fast, and disk fragmentation etc. can make things worse as does multiple files, though 5 files is not too bad. You could also see if you can copy each file one at a time faster.

    On Cu Gig we do better than a GB/sec all the time - for single, large files. It can take a whole lot longer to do a G in the form of a few K files.

    Just a thought.

  6. Hmm, nice graph, but seems prettier than accurate. That monograph looks like something "cooked up" more for marketing, and the graph to go with it. Sort of thing tech sales guys could sling around to their VARs or whatever.

    A simple question that begs to be asked: Who sets the "recommended temperature" if lowering it by 5 degrees could improve "reliability" so much? Is this evidence of planned obslolesence?

    This to me looks suspciously like a graph pulled from a materials handbook - conductivity or something - and dressed up and the axes relabeled.

    Being cold may not help a HDD - what is the effect on the viscosity of the lubricants in the moving parts? What happens as current skitters around due to changes in resistance, etc.?

    HDDs are not CPUs - and if they are designed reasonably well, they should perform and last pretty well within a fairly broad range of operating conditions.

    Of course, don't drop the silly buggers....

  7. A question that I've had that seems pertinent - and I do not see the in the voluminous yet valuable discussion here, is to question the entire notion of RAID and these SATA drives.

    Much is made about the firmware of the SATA devices and how they can outperform ostensibly faster SCSI drives under "real world dektop situations."

    What does RAID do to this? Presumably the controller at some point and at some level imposes a different pattern of usage on the devices, even when at the user-level the pattern is a "real world dektop situation." Perhaps striping will be less affected by the presumed phenomenon than other RAID levels, but overall the SATA questions seems to me to be one of demonstrated good performance in one direction leading to the drives being used in another direction where that demonstrated performance does not necessarily apply.

    Skeptically yours.

  8. I have the 911 chip in all external storage devices (Win2k & WinXP) and have not had any trouble.

    I don't benchmark max throughput, but I do store (and restore) disk images from external drives - frequently, burn CD's, DVD's, etc. 

    All of the enclosures have been in use for at least a year with a couple with more than 2 years on them.

    Just curious what issues I'm supposed to be having?  ;)

    A sample discussion of firewire issues can be found here:

    As a general rule the answer seems to be in one of either two camps: (1) This is/ these are in some way an intrinsic 1394 issue (though some sub camps seem to think it could be fixed or worked around) or (2) It is bridge/firmware/driver - or OS - related and is not universal.

    Hunting the web quite a bit on these issues I tend to not find them discussed on manufacturer sites, FAQs, etc. but there's actually quite a bit on various boards. Since the issue is write-specific, operations like burning discs off the drive would not be affected.

  9. Oops - I hit "Report" instead of "Reply".  Wonder what that does?

    You just blew some CIA operatives cover. Right now he is desperately trying to avoid capture by the Chinese secret police and make his way to the secondary emergency extraction LZ just outside of Guangzhou. Thanks a lot. :angry:


  10. I have never seen this problem, though your "solution" suggests something: A fragmented swap file can cause lousy performance similar to what you describe. If you set your swap file to a fixed size, usually at least the size of your physical memory, you can reduce this. Using a defrag utility that will defrag your swap file (e.g., DiskKeeper) is also a good idea.

  11. Oops - I hit "Report" instead of "Reply". Wonder what that does?

    What brideg do you mean in the Audigy? I am referring to the 1394<->IDE/ATAPI/Whatever bridge for non-native 1394 drives, such as the Oxford 911 or the Initio inic-1430.

    As far as Audigy goes, I have seen more errors across those boards used as 1394 interface than across other boards we've used (e.g., TI-based Adaptec 4300).

  12. Does anyone out there have an opinion on what is the current best 1394 bridge chipset? The Oxfords, once highly regarded, I know have several issues. I would say that first and foremost reliability would be the issues, esp. as most of the "performance" differences are really just firmware optimizations that may or may not even apply to a given usage pattern.


  13. Thanks for the update!

    Well, personally once I had kids none of the rest of it mattered. I put it off a while, too, but once I plunged in there was no looking back. Though I am thankful that I don't, I would glady dig ditches to keep them fed. Probably things were better in the days when you turned 18 and just started cranking them out - better not to think too much about some stuff. If a stupid vole can pull it off....

    As to the boat, I think it is much nicer w/o a payment. You can have a lot of good times on a used boat - let some other sucker pay for the depreciation. Really, unlike a car, if you get a good, well-constructed hull and take care of it you'll only need one for life. A '60s or '70s Bertram is IMHO still a better hull than you can have constructed today at any price - and you could sell it in 10 years for at least what you paid for it today.


  14. would if the building was owned by a public company

    what if...

    Actually, this bit about photos has been explored pretty heavily, including and up to the Supreme Court.

    If you were on public property when you took the photo you could do with it what you like, including all commercial uses, purposes, etc. On the other hand, you could very well expect to get sued for it, and unless you had deep pockets or garnered the support of some group or another you'd probably get pasted - law or not.

    A great example of this is the Rock & Roll Hall of Fame & Museum. Fancy building, all kinds of copyrights on the structure, etc. A local shooter starts to make a pretty lucrative business selling photos, posters, post cards, etc. His photos were better than the "official" photos, so his swag sold better. The Rock Hall did not like this, so they sued - never mind that all of his photos were from public property. He won. They appealed. The PPA helped him out with legal, and he won again. Repeat aud naseum.

    Interestingly, this even applies if it is a recognizable photo of you. The NY Times did a series on "Young Urban Black Gays" or something similar. A stringer went out and got a shot of a black guy in a suit on the city streets. They put his photo on the cover of their magazine ssection under the headline. The guy did not like it much, nor did his wife. Too bad - unless they NYT had said "This guy Robert Smith in the photo is a gay black man" or something to that effect they could do as they pleased with the photo. Legally correct, but rather rude.

    On the other hand, if you are not on public property and take the photo but have permission - even if you are working an assignment for the owner, subject, whatever - and the permission is not an explicit work for hire agreement that includes transfer of rights to the subject, owner, whatever then the shooter still owns it. Not only can the shooter make whatever private, public, commercial, or whatever use of the photo they wish, but if you copy it - even if it is a picture of you - you are liable for copyright infringement. Ever notice the © on HS graduation photos?

  15. Yes, looks familiar.

    I recall a post (now fragged) where I once made reference to a lifestyle that will become known as "living like Super C".

    The bit about "no longer a spring chicken - I just hit 33" or whatever was particularly reminiscent of the joys of days past.

    Oh, and by the way Super C, if you pull the outboards out of the water, they last longer, which is why so many mid-sized saltwater vessels are outboards.

  16. Hi,

    Has anyone here set up and used an external RAID based on the Promise UltraTrak series, esp. the SX8000? What drives did you use? How well did it work?

    At the moment, this appears to me to be a fairly astonishing external RAID price point. Compared to Medea (we have I think 6 Medea RAIDs - they are only OK, and have a high RMA issue, but are cheap cheap cheap) which has a 5 spindle 1T for about $6K, it looks like this 8 spindle would run about $4K for around 2T.

    Are there any other suggestions out there? Portability and capacity are important. I tend to move the Medeas around a lot, but they are not very robust and this tends to break them. If you bump one of the Rorkes we shoot you, so that is right out (not to mention the fact that at >$10K for 800GB, if we coudl afford enough fo them, we would not need the work).

    Firewire simply is not reliable enough for my uses, and external single drive SCSI is expensive and space-limited. External SCSI enclosures with RAID controllers (e.g., Mylex, Adaptec 3400, etc.) are not very portable, and I have yet to find an internal RAID controller that does not screw up video (though SCSI to RAID, such as on our Medea or Rorke RAIDs obviously work fine). External SCSI enclosures with soft RAID are also not very portable.