5U 24 drive racks go for about $2k each [...] 144 TB per rack
If density is all, you're doing better than TFA at 4.8drives/U, but you can get 8U 40-drive racks for 5 drives/U or 150TB/rack. Pricey though, and you need at least 3 raid cards per box to run that many drives, which rules out mini-ITX.
I expect you can do even better with those pull-out drawers of drives that Sun used to go in for, but probably not in SATA.
We (Rio) once had three CF-sized 1in hard drives turn up on a pallet. Yeah, as in forklift. Admittedly they were pre-production samples direct from the manufacturer's labs, but even that's not really an excuse for a 3ft*3ft*2ft box.
I'm still looking for any graphics card with amd64 Linux drivers that supports either dual-dvi with accelerated portrait mode (1200x1600 x2), or dual-link DVI (2560x1600). Matrox have some that will do it, but only with proprietary drivers and only on ia32.
Media transport protocol - specs so it can be implemented in a GPL-friendly app.
Ooof, you want to use that crock? IMO MTP is an example of how Microsoft can really cut off their nose to spite their face -- saddling themselves with an inferior system, just so they can deny access to it to others. All sane devices also implement mass-storage class.
(But in fact it may not even be Microsoft's fault that MTP is GPL-unfriendly: can its baseline PTP be implemented in a GPL-friendly app itself? PTP was originally put together by camera manufacturers, who typically beat even Microsoft in the field of closed systems and customer lock-in.)
Exactly what I thought when I read this in the article: Microsoft had effectively kept everyone on a project within a square-mile radius of Redmond. A radius of a square mile? What dimensionality is the space around Redmond these days?
And sometimes also on Slashdot, although not when they post articles about us at well after beer o'clock on a Friday evening, UK time.
The car-player was a failure only in the sense that not everybody who would have loved one, ever got to hear about it. Several thousand were successfully made, the company more-or-less broke even making them, almost all were sold for over a grand a pop with never a penny spent on advertising, and a large majority are probably still in everyday use five years later.
If that's a failure and that Ipod Shuffle thing is a success, I'll take failure over success any time, please.
I don't know anyone that uses anything besides iTunes or Windows Media Player because they have full Unicode support. Having menu in 20 languages is not the same as being able to display song names in the correct language.
Rio's final products -- Karma, Carbon, Forge -- all have Unicode support, and fonts covering all the languages of Europe plus Japanese.
I still wonder if anyone else found it as funny as I did.
As a student, my one moment of joy in a long and boring lecture course on databases was when the lecturer brightly said, "Alternatively, you can take a dump every night -- and then process the massive log you've produced".
This kind of anti-corporate behavior reflects poorly on the entire country
This behaviour isn't anti-corporate. It's pro-corporate. What happens when Amazon decides that the purpose of their listings is only to buy stuff from Amazon, and that all other uses of that scraped information is illegal? Allowing spam harvesters is IMO a small price to pay for the rest of us being allowed to use the contents of websites for purposes unintended by their owners.
The streak angle, BTW, is exactly arctan(2/3) -- the streak goes two pixels up for every three across. (It goes 652 pixels up and 978 across, which is less than 1% different from 2/3, smaller than the error of me pointing at things in the Gimp.) To me this makes it very likely to be an artifact.
My guess is a very bright event (the failure of the streetlight, probably) causing CCD overexposure and subsequent temporary ill effects on the rest of the CCD scan line. Any Canon geeks in the house who know about the CCD scanning direction of a Powershot G3 and can compare it with the streak "trajectory" angle?
Or, they could install this thing, called a driver...
Not if the device is USB mass-storage class, like some of our recent products. In that case Windows' own mass-storage class driver owns the endpoints and all we can do is send custom USB commands through the existing driver -- which is what requires admin privileges. Before we started doing mass-storage class players, we did indeed have our own driver and could run entirely unprivileged.
Peter
Re:It's not just the shady companies
on
The Spyware Inferno
·
· Score: 5, Interesting
I don't say they're delivering ads or sending back personally identifiable info to their manufacturers, but they are using my resources without giving me what I consider to be any perceptible advantage.
Rio Music Manager has one, too, and the reason we put it there is because there are certain things that Rio Music Manager needs to do (such as send custom USB commands to portables) which can't be done by an unprivileged user under Windows. So at install time -- assuming it's installed by an administrator -- the service gets run with admin privileges, and then later, when unprivileged Rio Music Manager runs, it can send custom USB commands via the service.
On Linux it's probably just "chmod 660/dev/sdwhatever ; chgrp portable/dev/sdwhatever" and adding people to group portable, but on Windows it's not so easy. Not all background tasks are necessarily malicious.
I thought we did not have patents as of yet in the EU....
They're not enforceable, because of, coincidentally, the Munich Convention on patentability. For some reason, this has not prevented many patent offices, including the UKPTO, from "awarding" them.
Perhaps there are devices that can survive having an upgrade interrupted, but I've never seen any.
Empeg car-player, Rio Karma, Rio Central -- in fact, basically the entire Empeg/Rio line, possibly even the little MP3 flash players. You have a tiny bootblock, not field-writable and only smart enough to boot or reflash the actual firmware. Many flash chips have a special block at address 0, for just this purpose, which can be made non-field-writable.
OP is right, making a device that can be doorstopped by a failed upgrade is just laziness. Of course, you'd better hope you got the bootblock right first time...
...and never followed. In particular the bit about text being either ISO Latin 1 or UTF-16 (or, in later versions of ID3v2, UTF-8), which is a very sensible idea, is always completely ignored; the overwhelming majority of tag writers, both on Windows and Linux, write text in arbitrary 8-bit encodings (shift-JIS, GBK, whatever) and then mark them as being Latin 1. There's nothing a tag reader can do about that, as there's no way to work out what the writer's locale was. Taglib can write Unicode tags correctly, but no front-end for it that I've seen does the Right Thing: use Latin 1 tags if all the characters used are available in Latin 1 (or, given the problems above, US-ASCII) and UTF-16 tags otherwise.
The problem isn't the standard. It's the implementors.
So, do you have to print those out as flipbooks or what?
Hey, don't knock it until you've tried it. I once worked on an animation package for schools in which print-as-flipbook was one of our users' favourite features.
Ugh, any chance of ZeroConf support in a future version?
Well, ZeroConf is the same as UPnP as far as IP address acquisition goes. As for the upper layers (multicast DNS and what-have-you), it's a possibility, but it's never clear how much those things are going to take off and we thought we'd test the waters with one standard at a time.
Considering that Rendezvous and UPnP are both being developed in a reasonably open-standards kind of manner, it's a bit crap IMO that there's two of them.
Does anybody know ANY portable player that is able to show UTF-8 encoded filenames and/or tags?
Karma keeps all track information in UTF-8, and the transfer software fully understands UTF-8 and UTF-16 tags. Unfortunately the very first release of the Karma firmware won't have Unicode fonts, but we're currently intending to offer a subsequent free upgrade including glyphs for Cyrillic, Greek and Kanji. The Rio Nitrus (the 1.5Gb micro-hard-disk player which we've also just announced) has UTF-8 support including Cyrillic, Greek and Kanji from the word go.
So is that actually usb 1.0 or 1.1 renamed as usb 2.0 (usb full speed) or usb high speed incorrectly labelled as usb 2.0?
Without necessarily wishing to express an opinion on the nitwits who thought that that renaming was a good idea, Karma supports the 480Mbits/s variety of USB, or, as I'm tempted to call it, proper USB2. (That is, the wire speed is 480Mbits/s; you don't get the whole 60Mbytes/s in practice as that's more than the head rate of the winchester.)
I guess part of why it grates slightly to hear the car-player described as "hackable" is that it's like a "parkable" car or a "openable" door. It ought not to take extra words to explain that a gadget hasn't been made incapable of doing some of the things it's clearly capable of. It ought to take extra words to explain when it has.
5U 24 drive racks go for about $2k each [...] 144 TB per rack
If density is all, you're doing better than TFA at 4.8drives/U, but you can get 8U 40-drive racks for 5 drives/U or 150TB/rack. Pricey though, and you need at least 3 raid cards per box to run that many drives, which rules out mini-ITX.
I expect you can do even better with those pull-out drawers of drives that Sun used to go in for, but probably not in SATA.
Peter
We (Rio) once had three CF-sized 1in hard drives turn up on a pallet. Yeah, as in forklift. Admittedly they were pre-production samples direct from the manufacturer's labs, but even that's not really an excuse for a 3ft*3ft*2ft box.
Peter
Put nipples on the reset and power button.
You think you're joking, but ICL PCs of the 1990s had power buttons whose colour was officially described as "nipple pink".
Peter
I'm still looking for any graphics card with amd64 Linux drivers that supports either dual-dvi with accelerated portrait mode (1200x1600 x2), or dual-link DVI (2560x1600). Matrox have some that will do it, but only with proprietary drivers and only on ia32.
Peter
Media transport protocol - specs so it can be implemented in a GPL-friendly app.
Ooof, you want to use that crock? IMO MTP is an example of how Microsoft can really cut off their nose to spite their face -- saddling themselves with an inferior system, just so they can deny access to it to others. All sane devices also implement mass-storage class.
(But in fact it may not even be Microsoft's fault that MTP is GPL-unfriendly: can its baseline PTP be implemented in a GPL-friendly app itself? PTP was originally put together by camera manufacturers, who typically beat even Microsoft in the field of closed systems and customer lock-in.)
Peter
Are you posting from a different dimension?
Exactly what I thought when I read this in the article: Microsoft had effectively kept everyone on a project within a square-mile radius of Redmond. A radius of a square mile? What dimensionality is the space around Redmond these days?
Peter
I can still find them on the empeg bbs.
And sometimes also on Slashdot, although not when they post articles about us at well after beer o'clock on a Friday evening, UK time.
The car-player was a failure only in the sense that not everybody who would have loved one, ever got to hear about it. Several thousand were successfully made, the company more-or-less broke even making them, almost all were sold for over a grand a pop with never a penny spent on advertising, and a large majority are probably still in everyday use five years later.
If that's a failure and that Ipod Shuffle thing is a success, I'll take failure over success any time, please.
Peter
I don't know anyone that uses anything besides iTunes or Windows Media Player because they have full Unicode support. Having menu in 20 languages is not the same as being able to display song names in the correct language.
Rio's final products -- Karma, Carbon, Forge -- all have Unicode support, and fonts covering all the languages of Europe plus Japanese.
Peter
I still wonder if anyone else found it as funny as I did.
As a student, my one moment of joy in a long and boring lecture course on databases was when the lecturer brightly said, "Alternatively, you can take a dump every night -- and then process the massive log you've produced".
Peter
He didn't say "unpopular". He said "lame". Many hella lame products become commercially successful.
Peter
This kind of anti-corporate behavior reflects poorly on the entire country
This behaviour isn't anti-corporate. It's pro-corporate. What happens when Amazon decides that the purpose of their listings is only to buy stuff from Amazon, and that all other uses of that scraped information is illegal? Allowing spam harvesters is IMO a small price to pay for the rest of us being allowed to use the contents of websites for purposes unintended by their owners.
Peter
monstrosities with cameras, video games, color screens and picture messaging
Completely agreed. I'm intending to go to these guys if I ever need to replace my 8890. Which I hope I don't.
Peter
the streak "trajectory" angle
The streak angle, BTW, is exactly arctan(2/3) -- the streak goes two pixels up for every three across. (It goes 652 pixels up and 978 across, which is less than 1% different from 2/3, smaller than the error of me pointing at things in the Gimp.) To me this makes it very likely to be an artifact.
Peter
Canon PowerShot G3
My guess is a very bright event (the failure of the streetlight, probably) causing CCD overexposure and subsequent temporary ill effects on the rest of the CCD scan line. Any Canon geeks in the house who know about the CCD scanning direction of a Powershot G3 and can compare it with the streak "trajectory" angle?
Peter
Or, they could install this thing, called a driver...
Not if the device is USB mass-storage class, like some of our recent products. In that case Windows' own mass-storage class driver owns the endpoints and all we can do is send custom USB commands through the existing driver -- which is what requires admin privileges. Before we started doing mass-storage class players, we did indeed have our own driver and could run entirely unprivileged.
Peter
I don't say they're delivering ads or sending back personally identifiable info to their manufacturers, but they are using my resources without giving me what I consider to be any perceptible advantage.
/dev/sdwhatever ; chgrp portable /dev/sdwhatever" and adding people to group portable, but on Windows it's not so easy. Not all background tasks are necessarily malicious.
Rio Music Manager has one, too, and the reason we put it there is because there are certain things that Rio Music Manager needs to do (such as send custom USB commands to portables) which can't be done by an unprivileged user under Windows. So at install time -- assuming it's installed by an administrator -- the service gets run with admin privileges, and then later, when unprivileged Rio Music Manager runs, it can send custom USB commands via the service.
On Linux it's probably just "chmod 660
Peter
I thought we did not have patents as of yet in the EU....
They're not enforceable, because of, coincidentally, the Munich Convention on patentability. For some reason, this has not prevented many patent offices, including the UKPTO, from "awarding" them.
Peter
Perhaps there are devices that can survive having an upgrade interrupted, but I've never seen any.
Empeg car-player, Rio Karma, Rio Central -- in fact, basically the entire Empeg/Rio line, possibly even the little MP3 flash players. You have a tiny bootblock, not field-writable and only smart enough to boot or reflash the actual firmware. Many flash chips have a special block at address 0, for just this purpose, which can be made non-field-writable.
OP is right, making a device that can be doorstopped by a failed upgrade is just laziness. Of course, you'd better hope you got the bootblock right first time...
Peter
...and never followed. In particular the bit about text being either ISO Latin 1 or UTF-16 (or, in later versions of ID3v2, UTF-8), which is a very sensible idea, is always completely ignored; the overwhelming majority of tag writers, both on Windows and Linux, write text in arbitrary 8-bit encodings (shift-JIS, GBK, whatever) and then mark them as being Latin 1. There's nothing a tag reader can do about that, as there's no way to work out what the writer's locale was. Taglib can write Unicode tags correctly, but no front-end for it that I've seen does the Right Thing: use Latin 1 tags if all the characters used are available in Latin 1 (or, given the problems above, US-ASCII) and UTF-16 tags otherwise.
The problem isn't the standard. It's the implementors.
Peter
So, do you have to print those out as flipbooks or what?
Hey, don't knock it until you've tried it. I once worked on an animation package for schools in which print-as-flipbook was one of our users' favourite features.
Peter
Ugh, any chance of ZeroConf support in a future version?
Well, ZeroConf is the same as UPnP as far as IP address acquisition goes. As for the upper layers (multicast DNS and what-have-you), it's a possibility, but it's never clear how much those things are going to take off and we thought we'd test the waters with one standard at a time.
Considering that Rendezvous and UPnP are both being developed in a reasonably open-standards kind of manner, it's a bit crap IMO that there's two of them.
Peter
Karma keeps all track information in UTF-8, and the transfer software fully understands UTF-8 and UTF-16 tags. Unfortunately the very first release of the Karma firmware won't have Unicode fonts, but we're currently intending to offer a subsequent free upgrade including glyphs for Cyrillic, Greek and Kanji. The Rio Nitrus (the 1.5Gb micro-hard-disk player which we've also just announced) has UTF-8 support including Cyrillic, Greek and Kanji from the word go.
Peter
Without necessarily wishing to express an opinion on the nitwits who thought that that renaming was a good idea, Karma supports the 480Mbits/s variety of USB, or, as I'm tempted to call it, proper USB2. (That is, the wire speed is 480Mbits/s; you don't get the whole 60Mbytes/s in practice as that's more than the head rate of the winchester.)
Peter
10/100 auto-sensing. Even 100 is not as fast as USB2, though.
Peter
A definite good thing in this forum
I guess part of why it grates slightly to hear the car-player described as "hackable" is that it's like a "parkable" car or a "openable" door. It ought not to take extra words to explain that a gadget hasn't been made incapable of doing some of the things it's clearly capable of. It ought to take extra words to explain when it has.
Peter