PAM works - even using PG 7.2 (compiled with configure --with-pam)
I don't believe it was an extra patch - I don't see any PAM specific patches in my source tree.
Here's the pg_hba.conf to make it work: (for me)
local all pam postgresql host all 127.0.0.1 255.255.255.255 pam postgresql host all 10.0.0.0 255.255.0.0 pam postgresql host all 81.2.78.40 255.255.255.248 pam postgresql
and the/etc/pam.d/postgresql (which isn't perfect - and you will definately have to change this to suit your system. Taking a copy of/etc/pam.d/login is a good way to start).
Warning - if your network is insecure, then you should be using an SSL connection to the postgres server - otherwise your passwords will be wandering around in plain text.
Erm - I don't think they have 3090+ distributed computers wired up to one Tivo
relentlessly trying to pummel it's backdoor.
What they done is rip the disk out, copied the raw hash off
(which is: 96F8B204FD99534759A6C11A181EEDDFEB2DF1D4) are sending that
out to the crack pool of computers.
They took their Tivo to bits so that you don't have to (maybe).
"I've heard tales of a big country across the Atlantic, where they still measure stuff in units like feet, stones, pounds, gallons and grandfathers-length- when-he-died, but I have a hope that we someday will manage to teach them the metric standard, how to read and how to elect sane presidents."
The UK isn't that big - but we'll take that as a compliment;->
...as for President Blair, - have you seen the competition? You know - the bald dude - no,
not him, the other bald dude...
Now it's early morning and I need a half bushel of coffee.
Fiat too - in the mid 70's, my dad went to
get into our white Fiat 131 which was parked outside our house.
...which was odd as he was sure he'd left it in the garage.
He opened the door with his key and hopped in only to be confused
by some random pile of stuff on the passenger seat. Double take for a minute,
walks out, reads number plate - get's embarrassed and hopes the real
owner doesn't notice him apparantly trying to nick his car!.
Doesn't inspire too much confidence in security:-o
Hmm - getting you to buy your own modem then asking you to regulate yourself is a tad daft.
Compare to getting electricity into your house. The Electricity Board own the meter and it's sealed.
Ergo - no fiddling.
Imagine if they told you to buy your own meter, program it right, install it and don't lie about the overnight discounted hours - everyone would laugh. I'm not even sure they would get very far with a prosecution - not because it isn't wrong and against your contact not to fiddle your lekky meter - but because the company had been so utterly stupid in their approach.
Having been running the cracker client all day, it appears
two things are limited:
The character set involved is just:
ABCDEFGHIJKLMNOPQRSTUVWXYZ 0123456789
I presume that limited by what you can enter via the Tivo remote (I don't actually have a Tivo).
The experts seem to be pretty sure they are dealing with a SHA1 hash. I'll shut up now
as I'm not a crypto expert. The one thing I will say is the character set is *very* limited
and favours a brute force attack.
It could be doomed if Tivo used a long string like 20 characters because every extra
character requires 37 times as much effort to permute all combinations as was previously required.
It's taking and estimated 3 days to cover the len=9 passwords. So 100-odd days will be needed
for the len=10 case.[1][2]
But there will be a limit to the length of the string - the Tivo engineers have to type the bl**dy thing in
so I find it hard to believe it's as long as it is.
You might also think that patching the code is viable - I believe you can do that. However I did see
some mutterings on a webgroup that Series 2 Tivos are key-signing parts of the system to prevent tampering
(so the next job for someone will be hacking the firmware:-)
Best, Timbo
Note: [1] - Assuming no short cuts are used in the scan. Seems pretty linear looking at the logs on my machine.
Note: [2] - Of course, the computing pool is growing steadily.
That's a good point. IBM (I think? I forget...) still sell Big Iron mainframes where you can pay for
1 CPU, but there are actully the full number of CPU's fitted. When you upgrade, the field engineer
wanders along and enables the extra CPU's.
Now - is it reasonable to hack the box and turn on all the fitted CPUs when I paid for one?
No company would, because pop goes your maintenance contract. If I were fortunate enough to aquire such a machine, say it's being thrown out in 5 years time (take it as read I don't require a maintenance contract) - I would feel no obligation to IBM or whoever and I wouldn't have an ethical problem in enabling every feature/CPU/device on the machine. I don't really think the manufacturer would care much either.
To go back to yet another car analogy: I buy a big engined car - it probably has a computer controlled engine. Given the big engine being very understressed for the car, I could chip it for a fair bit extra power without really causing any problems (warrenty now void - I don't expect them to help me out if I blow the motor).
Now consider: the manufacturer decides to save money by using the same engine in several models of car (puts 3 litre engine in models that would have spanned 2.5-3l previously) - but they performance limit the cheaper models in the engine controller's code. So far, so good. Now suppose for much cheapness, they have all the code tables for every model in the ROM but store a model selector byte in NVRAM, setting it at the factory.
Someone finds out how to reset the code and turn the cheaper 2.5l "performance equivalent" model into the full powered 3l version. So what's the difference between that and me chipping an understressed big engine? It's just cheapness (and negligence if they actually care) on their part.
The manufacturer made my life easy by giving me the mod chip up front. I do rather feel the onus is on them not to hand it to me on a plate.
You raise an excellent point concerning cable modems. What you describe used to be possible with ADSL in the UK until recently. The main difference being, I presume that your cable co own the modem and rent it to you - so you don't own the kit. That's a clearer case. Not your box, so you definately don't have the right to fiddle with it. Even so, it's still fairly stupid of them not to throttle your bandwith at the other end.
Well, the server has 788 active IP's and it's still up and ploughing through the work.
The answer to checking for forged results is to double the workload and submit each workset to two different machines as far away from each other as possible. If the answers agree, accept, otherwise recompute the worksets on more machines and take a majority vote. I thought Seti@Home did something like this?
Looks like/. did them a favour in terms of providing more computing power.
Now, if the story were about hacking the Tivo to obtain free service without due payment,
then I would have said that the crime is in the use of deception to obtain goods or services.
Busting the Tivo unit's guts is ancilliary and does not constitute any crime in itself (ignoring the DMCA
and whatever equivalent crap we have under UK law [1] ).
I actually have a lot of time for Tivo the company. I don't even have a Tivo. I'm thinking of getting one (and subscribing) because they had a good idea and made (by all accounts) a good product, backing it up with a good service. However, I am still going to rip the lid off the box, after about 6 months, when I feel the unit has burned in and I don't mind voiding the warrenty.
They have a rather suspect economic model but that doesn't nullify their achievements.
To be honest, I'm not sure that, even if Tivo opened the spec's to the box, that they would lose much. Tivo
hackers are likely to remain a special interest minority. Most people I would wager are happy to have another
AV black box on their shelf and pay for the Tivo service without wanting to fiddle with everything because what Tivo offer is exactly what they want.
Best, Timbo
Note [1] : I confess to being largely ignorant of my own coutry's and/or the EU's recent DMCA-like rulings. I am aware that President Blair is selling us down the swanny, but I plan to display total civil disobediance towards anything which in my view is so utterly stupid and ill conceived. Besides, if they bang me up, I might be able to get Lord Archer's autograph. hehe
Good argument, mumblestheclown. But I disagree concerning the freedom to employ unstable business models.
I agree, it wouldn't be very nice
to set fire to my Tivo and throw it through your window. Conversely if I rip the silencer off my motor, it would be perfectly OK to drive it around on private land (with permission) 20 miles from the nearest inhabitant (in the UK at least).
One reason I may want to mod the box is this: consider that maybe I want to use and pay for the Tivo service but I also
want to add some random feature. That would be in the same league as installing an amp in my car or whatever. I do not expect to have to ask the manufacturer's permission to disassemble my dashboard.
The other reason I may want to crack the unit is that it's my box - I paid for it, I own it, it's on my property.
I take on board your argument supporting varying business models - but I hold that the business model is flawed. Sell the box at a profit and discount the service. In a way Tivo's business model is basically parallel to the "loss-leader" trick employed by supermarkets. They offer something at an attractive discount (actually with a negative profit margin) in the hope that I will buy other products. However, it is perfectly reasonable for me to isolate all the loss leaders and buy them and nothing else, thus making a loss for the company. That's the risk they took. On average it works out well for them (or they'd stop doing it).
I'm sorry - if Tivo want to guarantee that I will buy their service, they shouldn't sell the box on it's own. Or they shouldn't at least sell it at a loss. I can buy a phone without a phone line or rent a phone line without a phone. It would be silly, but I can do it and it doesn't cause the telco or the phone makers any problems.
I generally subscribe to the view "What I own I can take the lid off and poke around" as a starting point. I am very much against any business model which is so flimsy that it needs laws like the DMCA to support it.
All of which is why I've added 2 machines at home to the cracking pool:-)
Certainly looks most interesting.
That is quite a genius manoever, using cp -al - wish I'd thought of that:-)
Now I'll go and read the article in depth rather than just skimming(!)
All the best.
2 years ago I wrote a script to do pretty much what the linked product does - ie: maintain a duplicate set of data areas on another machine via rsync.
I use the --backup-dir option to relocate copies of the files which the current rsync run would otherwise delete or modify.
With a bit of rotation, we can have users helping themselves to a full view of their home directory as of last night and also be able to restore files effectively from each day of the week going back 7 days in our case.
Sure does cut down on the number of tape restore requests.
As mentioned it is incredibly efficient - we deal with about 900GB of data backed up in this way - but rsync actually transfers about only 10-30GB of differences each night.
Only problem is my script was a crap prototype which is why I'm not letting anyone see it;-)
But I do have a design in my head for a more professional effort (will be opensourced) - I'm might even get enough peace at work to write it one day!
Damn - no mod points. Some kind soul mod the parent up as spot on please:-)
My wife (mainland chinese) had the same problem when writing up her thesis (needed English & Chinese). CJK Latex is cool - it's the cleanest way to use latex for asian language support (other solutions required patches to the latex binary and worse). Results are good even with the standard fonts. xemacs-mule is what she used for editing and she found the IME (GB2312 - simplified chinese) quite usable and familiar.
(For anyone who wonders how this works, you just type phonetically and during each syllable entered a "tie-breaker" list appears with numbered options of all the characters with similar sounds - then hit the number - it is possible to type at a respectable speed. Several IME's exist - not all phonetic - I suppose one chooses the best according to job or taste).
Yeah - it's not ideal - but it got the job done on a popular linux distro (Mandrake 8.0) with minimum fiddling.
Just doing the next set up upgrades in our house and I'll be trying to do a better job this time round.
1st point: My Ma & Pa are 75 and 80 - so I don't expect them to install anything. They wouldn't manage a Windows *anyversion* install either. I also like clean installs - just backup/etc,/usr/local and whack back the bits I need - tweaking as I go. I usually find bits of the config which can be done better along the way. It's a personal thing really...
2nd point: The Windows ME install took nearly as long as linux by the time I had installed Office, downloaded extra drivers over a modem and tweaked all the default settings so they were nice. Being a sysadmin at one of the better Uni Comp Sci departments in the UK probably classes me as "trusted with ssh"
3rd point: Sentiment accepted. However in my view, the Mandrake 8.2 installer is close to perfection. The Sony Viao I just installed last week was fully functioning with no messing about bar the winmodem. That's an imperfection I can accept for software that cost nothing save for 3 blank CDs. 8.2 is quite a dream compared to 7.1 which took considerable messing around to get ISDN to work.
And to boot - it all stays working - so my weekend of installing and showing them around means I don't expect to have to fix it for 1-2 years when I may do the next major upgrade.
Oh - and my previous life as an NT 4 admin (before I escaped to Linux/Solaris) has developed in me the mindset that MS OS's are the work of the devil (notice I said OS - not apps - many of which are actually very nice (at least until they die mysteriously). So it's more of a religious thing now - MS - I don't really like to touch that stuff - urrghh.
While we're on the subject of parents... some time ago mine wanted a PC. I duly built one and installed Windows ME as I thought I should be nice to them (!). I am a linux diehard - but I was worried they'd find it hard to use.
To cut a long story short, months 1-3 were OK. After that I was getting phone calls every week or two about things going wrong. My mum asked if there was any alternative to Windows - preferably something that Bill hadn't been near. (Me gets heart failure)
One Mandrake 8.0 install later and much tuning of fonts and installing opera I gave them a system which looked after their phone calls for the internet (diald), selection of browsers (opera + konq + moz - choose as required) and a nice email system (kmail + local mail hub (exim) routing through the ISP or my server as seemed good.
After 1 week, I asked how they were getting on. Mum said - and I quote - "this is nice and easy to use - much easier than that Windows thing". I kid you not.
There were web problems from time to time - mostly caused by crap websites. So recently, I went down and installed Mandrake 8.2. Spent a while tuning the anti-aliased fonts in KDE, made sure moz was stuffed with every plugin I could find and improved a few system things.
Now, and truely now, they are really happy. Moz does everything right with pretty much all the websites they're interested in. Kword is a passable - but simple - wordprocessor they find meets their requirements. Kmail is pretty, clear and intuitive.
The moral of the story was too choose the few apps they actually needed as being the "best of breed", polish them and stick the icons on the taskbar. They don't want to do much - they just want those few things to work properly.
And of course, the system is rock solid and I get decent remote admin (ssh).
I don't disagree with your pointing out my simplification of it all - and it was a simplification... And it was indeed about 5 mins over breakfast. But when thinking about any problem this interesting, one has to start somewhere. Multiple hosts and multiple failures are a challenge - agreed. It would be interesting to hear some ideas instead of sarcasm
Nice points cybergibbons:-) I liked the buckyball idea.
And I wouldn't like to leave a dead brick in the middle. What if it had a serious power fault and went up with a big bang (like an ancient Sun Ultra 1 did to me the other day)? I don't like the 3D idea. It is very cool - but not practical. And as you say, 3D will get path length problems when you've got many many cubes.
So, if you had *lots* of cubes you might be wanting to connect them in a 4D, 5D or some other n-dimensional way as mentioned (still possible with 6 couplers - just a block doesn't connect to *all* it's neighbours...
...which brings back memories of transputer chips with their 4 fast-serial connectors. Then beowulf clusters. Seems these schemes often end up with each node being plugged into a super-switch and being able to achieve wire-speed connectivity to any node over 1 virtual hop.
Which would make this very practical. IBM sells you an backplane, switch modules and storage modules. Which sounds like iSCSI and gig ethernet switches - which are here today - though I don't know if anyone has made a solution exactly along these lines. Most of the big SAN boxes I've seen still seem to be a big server type box with disks that plug in.
There could be mileage in losing the automatic couplers - but doing disks - and controllers - that are stackable (in 2D) and are connected with wires to a big switch - but using black split loom/conduit aka the borg.
Having wires would be distinctly less cool - but borg tubing would make up for it in my book:-) Naturallly CISCO or Extreme would realise this is very cool and make gig switches in the same form factor - so your switch cube would sit in the middle with all those black tubes. Oooh - I feel assimilated already...
Quote: Designing software that can mask the complexity of making a collection of plug- and-play drive modules appear to a user as one cohesive file system is expected to be one of the core challenges of the project.
"Software... core challenge"??? (This sentiment is in the context that IBM aren't totally clueless about this sort of thing;-)
Starting with a simple schema:
Low level disk manager carves up disks into globally uniform chunks - say 20GB for argument's sake.
RAID manager does the usual RAID 5 stuff using chunks from different cubes.
Logical volume manager combines/carves up logical raid arrays into user required sizes.
And finally a robust resizeable filesystem presents space to the user (or go back a step to present a virtual block device to Oracle or anything else that likes to avoid filesystems.
OK - that's a simple schema from which a better system can be evolved - but the core technology exists now. 1- disk partitioning; 2- RAID; 3-Linux LVM, Veritas Volume Manager and many others exist; 4- Growable filesystems exist (reiserfs, Veritas etc etc. Need to work on the ability to shrink for a fully rounded solution.
Stage 2 needs to be careful concerning topology to avoid bad latency problems.
To make this truely plug and play (but not in the MS sense) inserting a disk-cube would see it tested, auto partitioned and put in a pool. The systems engineer would be required to create/delete/alter filesystems and/or virtual disks as they needed - and configure things like how many simulatenouse cube failures can the system tolerate, how many hot spare cubes are kept in the pool and so on.
The software to do the underlying stuff is here today - I'm using it - albeit rather manually. The automation/management software to make this polished isn't hard conceptually. Of course if you only wanted one filesystem like the article mentioned it would require even less configuration;-)
I'm actually much more impressed with the hardware here. Very cool. Not sure about the 3D and "stacking" structure. Bugger to replace a dead cube in the middle. Unless you are supposed to leave it there and throw a new cube on the top? I'd go for a 2D stacking system with overlapping layers (like a brick wall) - but with the couplers designed so you can knock a brick out sideways leaving the others undisturbed. Hmm - just a thought...
Maybe we should take to walking backwards - a favourite pastime of students caught on camera during the filming of the Oxford-set UK series Inspector Morse.
Very difficult to spot during editing apparantly;-) Wonder what it would make of that?
The parent post deserves a reply - not because I disagree (Paul is right to offer caution) but because I can add something here. Feel free to mod me offtopic.
I can confirm that linux handles TB filesystems reliably in my limited experience
Where I work (Imperial College, London) we have an online backup system consisting of 16 disks on a Chaparral RAID
controller hanging off a cheap 1U Intel box. Using some perl I wrote in conbination with rsync, we pull copies of most data (shared group, sysadmin stuff, home dirs) onto this array - keeping a full copy and a week's worth of "reverse diffs" (files which have been updated, but I got rsync to kick them sideways rather than deleting them).
Whole lot is NFS and samba exported so short term file recovery is DIY for the users (saves a *lot* of time).
Anyway - here's the df -lk listing:
Filesystem 1k-blocks Used Available Use% Mounted on /dev/vg00/lvexport1 1050132564 822295616 227836948 79%/export/recover
You'll notice that I'm using LVM - which in itself has a 1TB limit on the size of a single logical volume (well right now it does on the version I'm using).
Next version of my code will deal with pools of disk so I don't have to use LVM anymore. The filesystem is ext3 running on SuSE 7.2 with a locally patched build of 2.4.17.
Approximately 3-5GB of files are moved every night
and it has (fingers crossed) been very reliable.
My advice would be to stick to ext3. Reiserfs I used at home and it was very reliable. But on 180 odd lab PCs we have, where Reiserfs is used on/, we have found that odd system crashes *can* sometimes result in null data or even randon data (extents from other files or off the free list) have ended up inside live files, like/etc/passwd, which isn't very convenient;-)
I've seem XFS leave files full of null data under similar circumstances. So I actually think, for the hack that it is, ext3 gives a certain warm feeling on those sorts of filesystems. Ext3 even on a system which was crashing due to a mixture of (older) kernel problems and flaky firmware on the RAID never lost anything (at least as far as we've noticed).
So far so good. Your milage may vary of course. And do as Paul says - check the limits of the drives and the filesystems you are thinking of using.
Ooh - I remember those days... Our first colour TV (Sanyo) came with a full circuit diagram as did my dad's germanium-transistor tuner/amp. Used the latter's diagram to great effect - replacing the transistors when they blew and testing the circuit (tip: don't go into Radio Shack - or Tandy as they are known in the UK - and ask for an AL102 germanium transistor;-) Still working today after 30 years. So is his mono valve amp he built 45+ years ago from a magazine article.
Just aquired a load of Canon printers at work - fine machines. But getting the technical docs (beyond the basic stuff) is like getting blood out of a stone. Wouldn't mind so much if their supplied utilities did exactly what I wanted - but they don't - so I waste my time reverse-engineering everything. Educational though...
On the face of it, this scheme seems a reasonable way to apply weighted charges to different roads according to the time of day. In that sense it would be more appropriate compared to upping car road-tax or fuel duty. It also seems better that the London Mayor's flat-rate charge to enter central London.
There are a few problems though:
Privacy - this sucks bigtime. "...strict controls to protect privacy." - hmmm. They do already have the technology in place AFAIK to track vehicles by OCRing the number plate - but at least that is limited to major roads with cameras. This little black box is going to be tracking you wherever you go. I suppose it will give bored MI5 agents something to do...
The road lobby is significantly powerful in the UK and includes most of the influential personages. 5 quid says this idea dies a silent death.
Another 5 quid says the lorry drivers will go mental and blockade central London.
David Begg's quote: "... we can never road-build our way out of this or provide enough public transport." is quite interesting. Rail transport is in a pretty poor state. If the government had been in the habit of giving British Rail the 6 billion pounds a year that they are currently spending on a supposedly privatised rail system (haha) instead of the 1 billion/year that BR got in the last years of it's existence, we'd have a damn fine rail system and a whole lot less cars on the road.
Overall, the goverment needs to commit to public transport asap. Let the roads become choked. If the trains and busses get good, people will start to move over - principle of the carrot.
On an aside, Uncle Tony's New Labour Transport Department isn't having a very good time:
Inherit privatised railways from the previous idiot government.
Bail out private railway companies with lots of taxpayers money so they can squander it on shareholders and the Chairman's salary.
Watch (or help) Railtrack to go bust.
Stand behind Jo Moore (Transport Secretary's aide) when she says "Hey we can bury all our bad news just after Sept 11th".
Decide to privatise Air Traffic Control.
Air Traffic Control run out of money. Bail them out with 30million for starters.
Watch Jo Moore do it again - "Can we bury some rail bad news around Pricess Margaret's funeral on Friday?..." Hold b*lls and run for cover. Sack/require resignation from Jo Moore and Director of Communications for the Dept of Transport.
Hey - let's privatise roads... - a different story to this one involving farming out road maintenance. Pilot scheme in Scotland lead to complaints already.
I've got a personal website, www.dionic.net
(Come/. my ISDN;-)
Just a handful of holiday snaps - not professional grade at all - but OK. I put them there because some people may find them interesting.
Trouble is - if you link directly to the larger scale images, as someone else said, you eat my feeble bandwidth and no-one knows about my site (unless they can be bothered to examine and dissect the URL a bit).
What I would consider more reasonable would be:
Link to the enscapsulating page so they see the context of the image as I intended and have links to the rest of my site
By all means copy a thumbnail again with a link to the enscapsulating page
Email me and if you're a not-for-profit site and you're site isn't on my very tiny list of things I don't really want to be associated with, I'll give you the full size (2400x3600) scan for free, if you'd be kind enough to mention me in a credits list somewhere
Or link directly to the image and include a short credit and a link to my home page near your link to my picture.
On point 4 if your site get's major hits I may need to chat about my link getting slaughtered and may suggest moving to scheme 3. But upto that point, at least your site is visibly linking to mine in some form so it's good for me:-)
Sure - I could do also sorts of things to the server like traffic throttling, HTTP referer checking etc. But for this thread I'm just considering the ethics from my POV.
My idea of what's fair won't be someone else's. So if you want to use other people's stuff to enhance your site - just ask. You may be pleasantly surprised - especially if you try to do something for them by way of advertsing their site in return.
Why can't we all just try the cooperative route before banging on about rights?
Slashdot Mirror
PAM works - even using PG 7.2 (compiled with configure --with-pam)
/etc/pam.d/postgresql (which isn't perfect - and you will definately have to change this /etc/pam.d/login is a good way to start).
/lib/security/pam_unix.so likeauth nullok /lib/security/pam_deny.so
/lib/security/pam_unix.so
I don't believe it was an extra patch - I don't see any PAM specific patches in my source tree.
Here's the pg_hba.conf to make it work: (for me)
local all pam postgresql
host all 127.0.0.1 255.255.255.255 pam postgresql
host all 10.0.0.0 255.255.0.0 pam postgresql
host all 81.2.78.40 255.255.255.248 pam postgresql
and the
to suit your system. Taking a copy of
#%PAM-1.0
auth sufficient
auth required
session required
Warning - if your network is insecure, then you should be using an SSL connection
to the postgres server - otherwise your passwords will be wandering around in plain text.
What they done is rip the disk out, copied the raw hash off (which is: 96F8B204FD99534759A6C11A181EEDDFEB2DF1D4) are sending that out to the crack pool of computers.
They took their Tivo to bits so that you don't have to (maybe).
The UK isn't that big - but we'll take that as a compliment ;->
Now it's early morning and I need a half bushel of coffee.
He opened the door with his key and hopped in only to be confused by some random pile of stuff on the passenger seat. Double take for a minute, walks out, reads number plate - get's embarrassed and hopes the real owner doesn't notice him apparantly trying to nick his car!.
Doesn't inspire too much confidence in security :-o
Best, Timbo
Compare to getting electricity into your house. The Electricity Board own the meter and it's sealed. Ergo - no fiddling.
Imagine if they told you to buy your own meter, program it right, install it and don't lie about the overnight discounted hours - everyone would laugh. I'm not even sure they would get very far with a prosecution - not because it isn't wrong and against your contact not to fiddle your lekky meter - but because the company had been so utterly stupid in their approach.
Having been running the cracker client all day, it appears two things are limited:
The character set involved is just: ABCDEFGHIJKLMNOPQRSTUVWXYZ 0123456789
I presume that limited by what you can enter via the Tivo remote (I don't actually have a Tivo).
The experts seem to be pretty sure they are dealing with a SHA1 hash. I'll shut up now as I'm not a crypto expert. The one thing I will say is the character set is *very* limited and favours a brute force attack.
It could be doomed if Tivo used a long string like 20 characters because every extra character requires 37 times as much effort to permute all combinations as was previously required.
It's taking and estimated 3 days to cover the len=9 passwords. So 100-odd days will be needed for the len=10 case.[1][2]
But there will be a limit to the length of the string - the Tivo engineers have to type the bl**dy thing in so I find it hard to believe it's as long as it is.
You might also think that patching the code is viable - I believe you can do that. However I did see some mutterings on a webgroup that Series 2 Tivos are key-signing parts of the system to prevent tampering (so the next job for someone will be hacking the firmware :-)
Best, Timbo
Note: [1] - Assuming no short cuts are used in the scan. Seems pretty linear looking at the logs on my machine.
Note: [2] - Of course, the computing pool is growing steadily.
Now - is it reasonable to hack the box and turn on all the fitted CPUs when I paid for one?
No company would, because pop goes your maintenance contract. If I were fortunate enough to aquire such a machine, say it's being thrown out in 5 years time (take it as read I don't require a maintenance contract) - I would feel no obligation to IBM or whoever and I wouldn't have an ethical problem in enabling every feature/CPU/device on the machine. I don't really think the manufacturer would care much either.
To go back to yet another car analogy: I buy a big engined car - it probably has a computer controlled engine. Given the big engine being very understressed for the car, I could chip it for a fair bit extra power without really causing any problems (warrenty now void - I don't expect them to help me out if I blow the motor).
Now consider: the manufacturer decides to save money by using the same engine in several models of car (puts 3 litre engine in models that would have spanned 2.5-3l previously) - but they performance limit the cheaper models in the engine controller's code. So far, so good. Now suppose for much cheapness, they have all the code tables for every model in the ROM but store a model selector byte in NVRAM, setting it at the factory.
Someone finds out how to reset the code and turn the cheaper 2.5l "performance equivalent" model into the full powered 3l version. So what's the difference between that and me chipping an understressed big engine? It's just cheapness (and negligence if they actually care) on their part.
The manufacturer made my life easy by giving me the mod chip up front. I do rather feel the onus is on them not to hand it to me on a plate.
You raise an excellent point concerning cable modems. What you describe used to be possible with ADSL in the UK until recently. The main difference being, I presume that your cable co own the modem and rent it to you - so you don't own the kit. That's a clearer case. Not your box, so you definately don't have the right to fiddle with it. Even so, it's still fairly stupid of them not to throttle your bandwith at the other end.
Best, Timbo
The answer to checking for forged results is to double the workload and submit each workset to two different machines as far away from each other as possible. If the answers agree, accept, otherwise recompute the worksets on more machines and take a majority vote. I thought Seti@Home did something like this?
Looks like /. did them a favour in terms of providing more computing power.
Now, if the story were about hacking the Tivo to obtain free service without due payment, then I would have said that the crime is in the use of deception to obtain goods or services. Busting the Tivo unit's guts is ancilliary and does not constitute any crime in itself (ignoring the DMCA and whatever equivalent crap we have under UK law [1] ).
I actually have a lot of time for Tivo the company. I don't even have a Tivo. I'm thinking of getting one (and subscribing) because they had a good idea and made (by all accounts) a good product, backing it up with a good service. However, I am still going to rip the lid off the box, after about 6 months, when I feel the unit has burned in and I don't mind voiding the warrenty. They have a rather suspect economic model but that doesn't nullify their achievements.
To be honest, I'm not sure that, even if Tivo opened the spec's to the box, that they would lose much. Tivo hackers are likely to remain a special interest minority. Most people I would wager are happy to have another AV black box on their shelf and pay for the Tivo service without wanting to fiddle with everything because what Tivo offer is exactly what they want.
Best, Timbo
Note [1] : I confess to being largely ignorant of my own coutry's and/or the EU's recent DMCA-like rulings. I am aware that President Blair is selling us down the swanny, but I plan to display total civil disobediance towards anything which in my view is so utterly stupid and ill conceived. Besides, if they bang me up, I might be able to get Lord Archer's autograph. hehe
I agree, it wouldn't be very nice to set fire to my Tivo and throw it through your window. Conversely if I rip the silencer off my motor, it would be perfectly OK to drive it around on private land (with permission) 20 miles from the nearest inhabitant (in the UK at least).
One reason I may want to mod the box is this: consider that maybe I want to use and pay for the Tivo service but I also want to add some random feature. That would be in the same league as installing an amp in my car or whatever. I do not expect to have to ask the manufacturer's permission to disassemble my dashboard.
The other reason I may want to crack the unit is that it's my box - I paid for it, I own it, it's on my property.
I take on board your argument supporting varying business models - but I hold that the business model is flawed. Sell the box at a profit and discount the service. In a way Tivo's business model is basically parallel to the "loss-leader" trick employed by supermarkets. They offer something at an attractive discount (actually with a negative profit margin) in the hope that I will buy other products. However, it is perfectly reasonable for me to isolate all the loss leaders and buy them and nothing else, thus making a loss for the company. That's the risk they took. On average it works out well for them (or they'd stop doing it).
I'm sorry - if Tivo want to guarantee that I will buy their service, they shouldn't sell the box on it's own. Or they shouldn't at least sell it at a loss. I can buy a phone without a phone line or rent a phone line without a phone. It would be silly, but I can do it and it doesn't cause the telco or the phone makers any problems.
I generally subscribe to the view "What I own I can take the lid off and poke around" as a starting point. I am very much against any business model which is so flimsy that it needs laws like the DMCA to support it.
All of which is why I've added 2 machines at home to the cracking pool :-)
Sod the DMCA and everything like it in Europe!
Best, Timbo
Certainly looks most interesting. That is quite a genius manoever, using cp -al - wish I'd thought of that :-)
Now I'll go and read the article in depth rather than just skimming(!)
All the best.
Hi
...
;-)
rsync --backup-dir
2 years ago I wrote a script to do pretty much what the linked product does - ie: maintain a duplicate set of data areas on another machine via rsync.
I use the --backup-dir option to relocate copies of the files which the current rsync run would otherwise delete or modify.
With a bit of rotation, we can have users helping themselves to a full view of their
home directory as of last night and also be able to restore files effectively from each day of the week going back 7 days in our case.
Sure does cut down on the number of tape restore requests.
As mentioned it is incredibly efficient - we deal with about 900GB of data backed up in this way - but rsync actually transfers about only 10-30GB of differences each night.
Only problem is my script was a crap prototype which is why I'm not letting anyone see it
But I do have a design in my head for a more professional effort (will be opensourced) - I'm might even get enough peace at work to write it one day!
Damn - no mod points. Some kind soul mod the parent up as spot on please :-)
My wife (mainland chinese) had the same problem when writing up her thesis (needed English & Chinese). CJK Latex is cool - it's the cleanest way to use latex for asian language support (other solutions required patches to the latex binary and worse). Results are good even with the standard fonts.
xemacs-mule is what she used for editing and she found the IME (GB2312 - simplified chinese)
quite usable and familiar.
(For anyone who wonders how this works, you just type phonetically and during each syllable entered a "tie-breaker" list appears with numbered options of all the characters with similar sounds - then hit the number - it is possible to type at a respectable speed. Several IME's exist - not all phonetic - I suppose one chooses the best according to job or taste).
Yeah - it's not ideal - but it got the job done on a popular linux distro (Mandrake 8.0) with minimum fiddling.
Just doing the next set up upgrades in our house and I'll be trying to do a better job this time round.
Or how about a little Blues Brothers inspired enactment...
You're selling?
Your wimmen... I want your wimmen - how much for your wife?
And your children - how much for your children?
I want them haha!
Quote: ...otherwise we end up with "Smoke Ciggies - They're Good For You".
"For your throat's sake - smoke Craven 'A' "
(Slapped by the Advertising Standards Agency, UK, many years ago)
MacDonald's will be claiming their food is healthy next...
Valid points - so here's my reasoning:
/etc, /usr/local and whack back the bits I need - tweaking as I go. I usually find bits of the config which can be done better along the way. It's a personal thing really...
1st point:
My Ma & Pa are 75 and 80 - so I don't expect them to install anything. They wouldn't manage a Windows *anyversion* install either. I also like clean installs - just backup
2nd point:
The Windows ME install took nearly as long as linux by the time I had installed Office, downloaded extra drivers over a modem and tweaked all the default settings so they were nice. Being a sysadmin at one of the better Uni Comp Sci departments in the UK probably classes me as "trusted with ssh"
3rd point:
Sentiment accepted. However in my view, the Mandrake 8.2 installer is close to perfection. The Sony Viao I just installed last week was fully functioning with no messing about bar the winmodem. That's an imperfection I can accept for software that cost nothing save for 3 blank CDs. 8.2 is quite a dream compared to 7.1 which took considerable messing around to get ISDN to work.
And to boot - it all stays working - so my weekend of installing and showing them around means I don't expect to have to fix it for 1-2 years when I may do the next major upgrade.
Oh - and my previous life as an NT 4 admin (before I escaped to Linux/Solaris) has developed in me the mindset that MS OS's are the work of the devil (notice I said OS - not apps - many of which are actually very nice (at least until they die mysteriously). So it's more of a religious thing now - MS - I don't really like to touch that stuff - urrghh.
To cut a long story short, months 1-3 were OK. After that I was getting phone calls every week or two about things going wrong. My mum asked if there was any alternative to Windows - preferably something that Bill hadn't been near. (Me gets heart failure)
One Mandrake 8.0 install later and much tuning of fonts and installing opera I gave them a system which looked after their phone calls for the internet (diald), selection of browsers (opera + konq + moz - choose as required) and a nice email system (kmail + local mail hub (exim) routing through the ISP or my server as seemed good.
After 1 week, I asked how they were getting on. Mum said - and I quote - "this is nice and easy to use - much easier than that Windows thing". I kid you not.
There were web problems from time to time - mostly caused by crap websites. So recently, I went down and installed Mandrake 8.2. Spent a while tuning the anti-aliased fonts in KDE, made sure moz was stuffed with every plugin I could find and improved a few system things.
Now, and truely now, they are really happy. Moz does everything right with pretty much all the websites they're interested in. Kword is a passable - but simple - wordprocessor they find meets their requirements. Kmail is pretty, clear and intuitive.
The moral of the story was too choose the few apps they actually needed as being the "best of breed", polish them and stick the icons on the taskbar. They don't want to do much - they just want those few things to work properly.
And of course, the system is rock solid and I get decent remote admin (ssh).
A true story...
I don't disagree with your pointing out my simplification of it all - and it was a simplification... And it was indeed about 5 mins over breakfast. But when thinking about any problem this interesting, one has to start somewhere.
Multiple hosts and multiple failures are a challenge - agreed. It would be interesting to hear some ideas instead of sarcasm
So, if you had *lots* of cubes you might be wanting to connect them in a 4D, 5D or some other n-dimensional way as mentioned (still possible with 6 couplers - just a block doesn't connect to *all* it's neighbours...
Which would make this very practical. IBM sells you an backplane, switch modules and storage modules. Which sounds like iSCSI and gig ethernet switches - which are here today - though I don't know if anyone has made a solution exactly along these lines. Most of the big SAN boxes I've seen still seem to be a big server type box with disks that plug in.
There could be mileage in losing the automatic couplers - but doing disks - and controllers - that are stackable (in 2D) and are connected with wires to a big switch - but using black split loom/conduit aka the borg.
Having wires would be distinctly less cool - but borg tubing would make up for it in my book :-) Naturallly CISCO or Extreme would realise this is very cool and make gig switches in the same form factor - so your switch cube would sit in the middle with all those black tubes. Oooh - I feel assimilated already...
"Software ... core challenge"??? (This sentiment is in the context that IBM aren't totally clueless about this sort of thing ;-)
Starting with a simple schema:
- Low level disk manager carves up disks into globally uniform chunks - say 20GB for argument's sake.
- RAID manager does the usual RAID 5 stuff using chunks from different cubes.
- Logical volume manager combines/carves up logical raid arrays into user required sizes.
- And finally a robust resizeable filesystem presents space to the user (or go back a step to present a virtual block device to Oracle or anything else that likes to avoid filesystems.
OK - that's a simple schema from which a better system can be evolved - but the core technology exists now. 1- disk partitioning; 2- RAID; 3-Linux LVM, Veritas Volume Manager and many others exist; 4- Growable filesystems exist (reiserfs, Veritas etc etc. Need to work on the ability to shrink for a fully rounded solution. Stage 2 needs to be careful concerning topology to avoid bad latency problems.To make this truely plug and play (but not in the MS sense) inserting a disk-cube would see it tested, auto partitioned and put in a pool. The systems engineer would be required to create/delete/alter filesystems and/or virtual disks as they needed - and configure things like how many simulatenouse cube failures can the system tolerate, how many hot spare cubes are kept in the pool and so on.
The software to do the underlying stuff is here today - I'm using it - albeit rather manually. The automation/management software to make this polished isn't hard conceptually. Of course if you only wanted one filesystem like the article mentioned it would require even less configuration ;-)
I'm actually much more impressed with the hardware here. Very cool. Not sure about the 3D and "stacking" structure. Bugger to replace a dead cube in the middle. Unless you are supposed to leave it there and throw a new cube on the top? I'd go for a 2D stacking system with overlapping layers (like a brick wall) - but with the couplers designed so you can knock a brick out sideways leaving the others undisturbed. Hmm - just a thought...
Very difficult to spot during editing apparantly ;-) Wonder what it would make of that?
The parent post deserves a reply - not because I disagree (Paul is right to offer caution) but because I can add something here. Feel free to mod me offtopic.
I can confirm that linux handles TB filesystems reliably in my limited experience
Where I work (Imperial College, London) we have an online backup system consisting of 16 disks on a Chaparral RAID controller hanging off a cheap 1U Intel box. Using some perl I wrote in conbination with rsync, we pull copies of most data (shared group, sysadmin stuff, home dirs) onto this array - keeping a full copy and a week's worth of "reverse diffs" (files which have been updated, but I got rsync to kick them sideways rather than deleting them).
Whole lot is NFS and samba exported so short term file recovery is DIY for the users (saves a *lot* of time).
Anyway - here's the df -lk listing:
Filesystem 1k-blocks Used Available Use% Mounted on
/dev/vg00/lvexport1 1050132564 822295616 227836948 79% /export/recover
You'll notice that I'm using LVM - which in itself has a 1TB limit on the size of a single logical volume (well right now it does on the version I'm using).
Next version of my code will deal with pools of disk so I don't have to use LVM anymore. The filesystem is ext3 running on SuSE 7.2 with a locally patched build of 2.4.17.
Approximately 3-5GB of files are moved every night and it has (fingers crossed) been very reliable. /, we have found that odd system crashes *can* sometimes result in null data or even randon data (extents from other files or off the free list) have ended up inside live files, like /etc/passwd, which isn't very convenient ;-)
My advice would be to stick to ext3. Reiserfs I used at home and it was very reliable. But on 180 odd lab PCs we have, where Reiserfs is used on
I've seem XFS leave files full of null data under similar circumstances. So I actually think, for the hack that it is, ext3 gives a certain warm feeling on those sorts of filesystems. Ext3 even on a system which was crashing due to a mixture of (older) kernel problems and flaky firmware on the RAID never lost anything (at least as far as we've noticed).
So far so good. Your milage may vary of course. And do as Paul says - check the limits of the drives and the filesystems you are thinking of using.
(Mod this down -10 as random sentimental wiffle)
;-) Still working today after 30 years. So is his mono valve amp he built 45+ years ago from a magazine article.
Ooh - I remember those days... Our first colour TV (Sanyo) came with a full circuit diagram as did my dad's germanium-transistor tuner/amp. Used the latter's diagram to great effect - replacing the transistors when they blew and testing the circuit (tip: don't go into Radio Shack - or Tandy as they are known in the UK - and ask for an AL102 germanium transistor
Just aquired a load of Canon printers at work - fine machines. But getting the technical docs (beyond the basic stuff) is like getting blood out of a stone. Wouldn't mind so much if their supplied utilities did exactly what I wanted - but they don't - so I waste my time reverse-engineering everything. Educational though...
On the face of it, this scheme seems a reasonable way to apply weighted charges to different roads according to the time of day. In that sense it would be more appropriate compared to upping car road-tax or fuel duty. It also seems better that the London Mayor's flat-rate charge to enter central London.
There are a few problems though:
David Begg's quote: "... we can never road-build our way out of this or provide enough public transport." is quite interesting. Rail transport is in a pretty poor state. If the government had been in the habit of giving British Rail the 6 billion pounds a year that they are currently spending on a supposedly privatised rail system (haha) instead of the 1 billion/year that BR got in the last years of it's existence, we'd have a damn fine rail system and a whole lot less cars on the road.
Overall, the goverment needs to commit to public transport asap. Let the roads become choked. If the trains and busses get good, people will start to move over - principle of the carrot.
On an aside, Uncle Tony's New Labour Transport Department isn't having a very good time:
Time to leave the country...
Just for 1 second...
I've got a personal website, www.dionic.net (Come /. my ISDN ;-)
Just a handful of holiday snaps - not professional grade at all - but OK. I put them there because some people may find them interesting.
Trouble is - if you link directly to the larger scale images, as someone else said, you eat my feeble bandwidth and no-one knows about my site (unless they can be bothered to examine and dissect the URL a bit).
What I would consider more reasonable would be:
On point 4 if your site get's major hits I may need to chat about my link getting slaughtered and may suggest moving to scheme 3. But upto that point, at least your site is visibly linking to mine in some form so it's good for me :-)
Sure - I could do also sorts of things to the server like traffic throttling, HTTP referer checking etc. But for this thread I'm just considering the ethics from my POV.
My idea of what's fair won't be someone else's. So if you want to use other people's stuff to enhance your site - just ask. You may be pleasantly surprised - especially if you try to do something for them by way of advertsing their site in return.
Why can't we all just try the cooperative route before banging on about rights?