« first day (67 days earlier)      last day (5111 days later) » 

00:08
@Iszi This is definitely a case where "best" is going to depend on two things. 1) Exactly what they want to get certified in, 2) personal opinion
@Iszi At least it's done. What had to be done?
@packs I agree. Totally subjective, and dependent upon the requirements of the asker - that's why it's better to discuss here, and not at all suitable for the main site.
@packs Some re-working of the wires coming from the main box and stuff - pretty normal for changing providers. But I could swear it took a lot longer than usual.
Also had to set up the new set-top boxes and modem of course.
So they actually have a different line on the poles? I guess that makes sense, just not something I've considered.
I've never lived anywhere that had multiple providers.
I wouldn't have expected that long either; are the lines buried?
 
3 hours later…
03:22
@packs Yeah, but there wasn't any actual work on the buried lines themselves. Just mostly re-hooking the stuff that runs in from the outside walls, installing new splitters and stuff.
 
6 hours later…
09:50
@Iszi @AviD the reason I closed security.stackexchange.com/questions/2201/… is that it looks like spam. But as @RoryAlsop had already commented that it was a closeworthy question, I decided two arbitrary mods were better than one ;)
@AviD, the rules on changing process user/group I.D. are quite restrictive. A process has two uids/gids, the real and effective. At launch, the real UID is the user who launched the process and the effective UID is the user it's running as. Of course, they're usually the same, but a setuid process gets a different effective UID.
If a process is effectively running as root, then it can use setuid to set its real and effective UIDs to anything. Otherwise it can only use setuid to set its real ID the same as its effective ID.
In the case of a process that's launched with the setuid filesystem bit, it additionally has a saved-set-uid which is the same as the owner of the file. Such a process can use seteuid to switch between the UID of the user who launched it, and the user its running as.
So basically a process can change its UID/GID at real time, but can never use it to acquire privileges it didn't have at launch. In addition, if a process changes UID/GID to anything that isn't root, it can never change back or to anything else. The biggest trick is binaries launched with the setuid bit, where the best options are:
1. try not to. Use the platform's service management framework (if it has one) to launch privileged services on demand
2. perform all privileged tasks then completely relinquish the switched ID and run as the calling user
3. if you can't do either of the above, then you have to be careful about using the correct effective user ID for the correct purpose.
 
1 hour later…
11:28
@Graham - re the closweorthy question. Absolutely:-)
re SetUID - good explanation - wish I had written that in
(and re SetUID - amazing how many companies we find doing it totally wrong)
it was based in a very vague way on my explanation in Professional Cocoa Application Security</plug>
LOL
good plug
agreed, lots of devs don't get it
11:29
think I have seen that book - maybe Rory McCune has it on his workshelf
if you'd like to see more of it, buy yourself a copy or ten :-D
You never know - despite me expectation that I am becoming a pointy haired boss, I find myself getting to do more technical work again now I have a small company
good fun getting back into it
anyway, what I like about Solaris and OS X is that they have decent service management facilities that can launch privileged services like inetd on steroids, so devs don't need to get setuid right.
yep - OSX looks like they really have grabbed the good bits from the services front
I don't know enough about the Windows service model to know how it helps devs in that regard. Perhaps I should, but I've got through the last decade without writing any Windows code, I might be able to manage this decade too.
11:32
I was expecting more issues on that front - just based on historical issues Apple have had with enterprise grade security :-)
I think in a weird stealth way Apple have actually been addressing enterprise needs well since about 2007
:-) stealth?
Yeah, they've got consultants who do enterprise security and federal security for customers, but they don't particularly talk about what they're doing publicly
BTW, who were your "Apple security folks" you dined with? I should probably get to know them :)
I sat down with the security guy for Apple mobile on Sunday for dinner and he seems very driven to get it right... not sure how much support he gets though
They are off finding out how much they can say - so will ask if I can pass on details
great, thanks :)
11:35
Hugo Fiennes is the guy I have known for a long time. Started up Empeg, then went to head up hardware development for iPhone
Neil is the security guy
Making contact with him would be best bet
both ex-Cambridge
Oh, right. I know one of their enterprise sec bods in the States, and have tried stalking Window Snyder but to no avail.
'stalking Windows Snyder' - excellent
TBH making the right contacts has been the bit I've got most wrong in starting my company :). I've not heard anything from Google or Microsoft - haven't even tried yet with Nokia or RIM.
actually, one of the best outcomes from BSides is the number of real security people I met and got a decent chat to
got a few following me on twitter now as well
Networking is definitely the way
Yup, hoping to get a lot of good contacts out of BSLondon. And I'm going to DefCon London tomorrow, too.
11:42
That DefCon looks to be a good one! re BSides - I definitely found that blogging/tweeting throughout got me more good attention than I expected
nods I've been pretty aggressive at insinuating myself on the infosec people I find on Twitter ;)
Hahaha
needs doing these days - just so many voluble folks that you need to get louder than them until you have proven yourself. Starting to pay off for us now.
Yup. Going to DefCon has helped there, made some good friends in the industry
We shall see what sort of effect moderating an infosec Q/A site has, too ;)
BTW, meant to ask you about BSides SF, that talk on threat modeling taxonomy.
Did it have anything new in it? :)
11:58
Not really - he presented things in a slightly better layout than I have seen before, and was good at connecting with the audience, but all standard stuff
There were very few talks which were all new
Although the Qualys talk (by Cisco) aT RSA on why IPv6 has become a nightmare for them was excellent
just the changes they needed to make because of the scale of the problem - scanning 20 Nonillion IP addresses just isn't feasible :-)
Maybe that's where I'm going wrong - I spend so much effort before each con trying to write a talk I've never given or seen before
That is what always got me - I keep thinking "do I have anything new?" and being scared to present in case folks are bored but the comments back are all around the value of pulling knowledge/experience together in a way the wider audience groks
guessing most folks don't have the depth of Apple security knowledge you have so most of your stuff is likely to be of interest
yeah, I should probably just go for doing "this is how the security stuff you know relates to the iPhone". The book was basically "this is how the Mac stuff you know relates to security".
Still working on changing the world by making security a default part of requirements/software engineering, but I need to get taken seriously first :-D
...and probably ought to make some money in the meantime, too...
12:08
I hear you there!
12:18
BTW, I'm inclined to leave this question on security.se rather than migrate to server fault:
1
A: How do I apply a security baseline to 2008 R2?

Mike SouleHas been replaced by the SCM toolkit: http://www.microsoft.com/downloads/en/details.aspx?FamilyID=fb8b981f-227c-4af6-a44b-b115696a80ac&displaylang=en Over view of SCM toolkit: http://social.technet.microsoft.com/wiki/contents/articles/microsoft-security-compliance-manager-scm.aspx SCM getti...

yes, it's about systems administration, but then if we say all sysadmin qs belong on SF and all secure programming qs belong on SO, we're not left with much :)
Writing security policy...but then maybe that belongs on writers.se?
I'm okay with it here - think @Iszi was just concerned it hadn't attracted answers.
just means we need more users with Windows experience ;)
true- waiting to see what size is the tipping point for this forum
We only have around 230 users who've ever actually scored any points, it would appear.
(i.e. have more than 101 rep)
it's encouraging the remaining 1300 users who have signed up and not interacted to ask/answer Qs where things should start happening ;)
12:34
A good chunk of that 1300 were probably just other SE users who popped up and hit the login button by reflex.
It would be interesting to write up some data queries about who those users are elsewhere.
"interesting" in a stalker sense ;)
I have no idea what you're talking about.
coughs
That was a nicely done blatant ripoff of a previous work w.r.t. uid changing.
For the application to change itself, it must implement the setuid function, no?
yeah, but at least the ripoff was of my own work :)
LOL - oh hi @packs
An application doesn't need to implement setuid(), it needs to call it. Also, it needs to either be root, or have a different real and effective UID.
12:43
@RoryAlsop Good Morning Sunshine.
Sorry, yes, that's what I meant.
It's afternoon now - was just about to get the kids lunch
:-)
@RoryAlsop You and your non-canonical time zone.
pretty sure UTC is canonical
I used the phrase 'implement' meaning 'write something that uses it'. I was trained as an engineer, not as a programmer :)
12:45
or do you mean that he's Scottish, and therefore the time is always defined as "beer o'clock" ducks
Well, you know how those Scotsmen are....
whisky o'clock, surely
@GrahamLee What I wasn't sure was if the kernel had a mechanism to change the uid.
I'm feeling no, but I wasn't sure.
@packs well yes, but in response to setuid() system calls.
I don't know of anything that causes a kernel to automatically change a process's UID except a bug.
Nor any interfaces, that I've come across, such as sysfs, that would allow it.
12:48
it'd be an interesting design for an OS though. "Oh, you want to browse the web? You automatically lose root"
%s/interesting/awesome/
Well, my office opens in 10 minutes. Time to get the kid cleaned up and out the door.
time for spaghetti hoops for my lot
catch you later
time to suck some nutrients into my feeding hole myself
Jealous though, @RoryAlsop, I'd love spaghetti hoops on toast for lunch. Not sure Mrs. @GrahamLee would approve.
13:28
Spaghetti hoops? Is that like Chef Boyardee with suet?
13:39
just the local name for spaghettios
You people, what with your chips, and lifts, and milk on the counter :)
milk on the counter? o_O
I remember, granted this was '98, being down in Devon and buying milk that required no refrigeration.
ah, UHT milk
that's the stuff that goes into those little plastic things of milk you get in conferences, surely you have those in the US too
or is it all that worrying "non-dairy creamer" goop
I've not run into it, but my conference exposure has been limited.
I tend to see water, tea, and coffee at the talks I've attended. With non-dairy creamer
13:53
spaghetti downed, children esconced in front of Jeff Bridges in Starman. School holidays, meh!
now back to work
I think non-dairy creamer was only invented because some Americans complained they couldn't shovel corn syrup into their faces quickly enough if they had to drink things that didn't contain it.
@packs speaking of non-canonical timezones: bbc.co.uk/news/magazine-12523164
14:10
I have found exactly 1 use for corn syrup.
When I'm baking cookies if I substitute some of the sugar for corn syrup, then they'll stay moister for a long ass time.
And considering I'm in a house with 2 adults and 1 <2yrs, it means they'll still be edible when we're finishing off the batch
double summertime, nice.
Double summertime? At least one summer would be good this year?
LOL - now read TFA
I can just hear Wormer's voice, "Double secret summer time!"
14:14
So, a few years ago, I was in Edinburgh in the middle of May.
Looks mightily flawed for Scotland - I won't vote for it
I found myself waking up every morning at 4:30, and couldn't figure it out.
Until the 5th day when I noticed the sun outside.
I used to work in a pub in Oxford with a late license, and by the time I left to go home the sun would already be up.
Considering that, even on the solstice, my sun rise here isn't visible until a few hours later, that disturbed me a little.
I find the idea of shifting that not entirely a bad idea.
In Orkney, where I grew up, it doesn't get dark in June
but is already bad enough in the winter
where you get about 3 hours of daylight
14:16
Did you all speak English up there? :)
Did you ever visit Twatt? :)
@packs - yes, and Norn, which is a version of old Norse. No Gaelic for us, which upset the Scottish council when they tried to get us to change our roadsigns to English/Gaelic - almost a riot
@Graham - yep, used to drive through Twatt on my way to Skaill - the best surfing beach in Europe, by a long shot
I studied a little Old Norse last year, probably still couldn't get by in Norn though
Unless you can tell your roo o iper from your gite gapas, you'd mebbe have trouble :-)
Couldn't even get by in Old Norse without at least two glossaries and a grammar rulebook on hand. And preferably an English translation.
14:26
I have found Norn staggeringly useless everywhere, including Orkney, where really everyone speaks English, albeit with a particularly strange accent
Maybe you could understand a Faroese? I've no idea how closely-related the languages are.
"Øll menniskju eru fødd fræls og jøvn til virðingar og mannarættindi. Tey hava skil og samvitsku og eiga at fara hvørt um annað í bróðuranda."
They are supposed to be similar, but the only times I've been to the Faroes I was still at school and mostly chatting to other Orcadians
Hmmm - not grokking that sentence too well
(Mind you I had difficulty with Danish when I worked there, so maybe I'm just not a natural with languages)
You might want to edit the Norn language page on wikipedia which describes the language as "extinct" ;)
(related: I'm not getting much work done at the moment)
change it to -mostly extinct :-)
it's same as anywhere - just the old folks
I'm making sure I'm current on Fortify ... all good
I'm writing a book, pretty sure that as long as I finish sometime before the Summer I should be in good shape.
14:46
@GrahamLee I've been saying that about my thesis for a few years
book on norse languages? @Graham
no, book on unit testing for iPhone devs
 
1 hour later…
16:03
So while waiting for this query to execute, I've been reading the wikipedia page on Orkney. Looks like an interesting history.
 
3 hours later…
19:25
@packs - definitely. Have you read up on our annual ball game, the Ba'?
Or the wonders of the local paper - orcadian.co.uk
19:58
The name "Ba'" sounds a little...unimaginative for a ball game. Was it named by an engineer?
I do love local papers, though.
Our letters to the editor are often pretty funny
So, @RoryAlsop, do you still think of yourself as an Orcadian?
 
1 hour later…
21:20
@packs - yep, Still an Orcadian, half English, half Australian, naturalised to Edinburgh (since 1989) with my childhood in Orkney and the Falklands.
some might say "mixed up"
Falklands....that sounds vaguely familiar...
Should I take a guess at about when you last lived there?
LOL - yep, about that time
My origins are, for better or worse, much simpler. I lived in a single town in Kentucky until I graduated college, then moved to Ohio for 2 years of grad school (back in '04).
It's amazing what getting a job after finishing the coursework will do.
21:26
definitely - sadly I took a year to find one in my industry, so I know a bit about how to drive a truck, run a bar, dig for peat etc :-)
I lucked out by skipping that kind of things. One of the senior analysts was a student in a class I TAed, so my name got stuck in his head when an opening came up.
that's what it's all about - being there and being known when the time is right
Without question. However, where I came from, it was all about who you knew. Even "open postings" were mostly only in name.
So I would much rather just having good luck :)
21:43
@Iszi Just in time to head out for the evening
@packs Pretty much. Just checking in before I leave.
This is the last week I'll be wardriving, fortunately. That is, until it comes around again in six months or so.
Part of me wishes that we actually had a policy disallowing WAPs on campus so that we could justify things like that.
This one looks like too much of a "discussion" topic for SE... why is it still alive?
On the other hand, I feel like we have enough to do without it.
2
Q: Why is it difficult to catch "Anonymous" (group)?

clawsHello, I'm not a security literate. If I was. probably I wouldn't be asking this question. As a regular tech news follower. I'm really surprised by the outrage of Anonymous (hacker group). But being a critical thinker, I'm unable to control my curiosity to dig out how exactly they are doing? Fra...

@packs You can still justify wardriving in areas that shouldn't have user-installed WAPs, like classrooms and offices.
@packs Yeah, I've got a load of stuff that others have to take over during wardriving season so I can get it done.
21:46
@Iszi The problem being that without anything actually disallowing it the only thing we could really do is look for non-official APs that are broadcasting the official SSID
@packs You mean there's nothing disallowing non-official APs in the areas mentioned above? That, I would think, is bad.
We do, however, fall down like a ton(ne)? of bricks whenever one of those things actually causes problems.
@Iszi Unfortunately, we have a little over 200 years worth of policy and procedure that don't always make it conducive to change.
We are starting to get some new policies going through which should help out with some of those issues.
@packs For that, you're probably better off with adding a WIDS/WIPS to your infrastructure I think. That's kind of where I'd like to get our wireless security pointed in the long run - have a system that's ubiquitous and capable enough to monitor everything, and only actually do walkdowns when something needs to be physically identified.
We're kind of halfway there right now. Still a lot of polishing to be done to the system and procedures.
Sounds cool, but expensive. As the de facto ID guy in these parts, I'll have to dedicate some time to reading up on it.
Our network guys actually have an RFI out right now for rebuilding our wireless infrastructure.
@packs Sorry - RFI? I'm only coming up with Radio Frequency Interference.
21:56
Request For Information
@packs Ah, right.
As opposed to an RFP (request for proposals), i.e. putting out to bid.
Considering we're a 1,800-acre campus with, near enough, 100% coverage, that could get interesting.
As I understand it from our network guys, the WIDS/WIPS for us is essentially built-in to the wireless controllers. All the LWAPs basically act as APs and WIDS/WIPS nodes at the same time.
That might simplify things. And depending on the management capabilities that is an awfully big might. As it stands now we're in the thousands of APs.
@packs Oh, I'm fairly certain it can handle your network, if it can handle ours. I don't know exactly how many APs we have here, but it's definitely not a number to sneeze at.
Just checked, and it looks like we've got nearly 1500 APs.
22:02
Ok, so we're in the same ballpark
I'll have to read up a little on this stuff and ask some of the network kids about how it might fit into the impending upgrade.
@packs If you're rebuilding the whole infrastructure already, now is definitely the time to do that.
@Iszi Exactly, after that it's just a question of functionality and funding.
On the other hand, quittin time. Here's hoping I miss the traffic on my way to school.
ta
@packs I'm sure functionality won't be an issue. Good luck on the funding. :-)
@packs l8rz

« first day (67 days earlier)      last day (5111 days later) »