[00:05] scientes: Yo. Where are you at the moment -- north bay? [00:14] paulproteus, yep [00:24] paulproteus, why you ask? [00:28] scientes: Planning to go chill at Noisebridge or something. [00:30] hmm, i've never been there, but it looks cool [00:32] paulproteus, yeah I just finished my first Mozilla patch! I'd be up for it [00:32] When could you be here? Sadly I have plans to be somewhere at 8 which might make the timing tight [00:38] erggg, i'd like to come in for longer if im coming all the way to the city, i'm old fashioned like that [00:43] yeah, agreed [00:43] We'll rock it another day [00:47] that place looks cool [02:33] http://lwn.net/Articles/497106/ [02:33] we had a discussion about this at uds at the userspace namespaces session [02:34] wait, thats a shitty patch [02:34] it should just use userspace namespaces, and allow you to do files with ANY UID/GID [02:35] ahh that was mentioned later http://lwn.net/Articles/497480/ [03:11] http://makerfaire.com/ is THIS weekend [03:11] a few of us are going on Sat [03:11] MarkDude, I will be there [03:12] yep, some Ubuntu California folks are helping jack with the BerkeleyLUG booth :) [03:12] I just need to figure out how to get from napa to there via the bus [03:12] like philipballew! [03:12] like me! [03:12] On sat? [03:12] yeah. saturday morning [03:12] Cool. meet up for some fun [03:13] * MarkDude got covered in diet coke from eepybird last year [03:13] yeah, I am down for that. I'm going to probably bring a sleeping bag in case I find the bus does not go back to my temp housing. [03:14] sunday public transit in ca is hard somethimes [03:14] like buying beer in the midwest on a sunday [03:14] MarkDude, were you able to change? [03:14] philipballew: are you by any chance in auburn tomorrow? [03:15] "My problem with Kay is that he has a very strong idea of what he wants, and it is fundamentally about a desktop distro in a very specific mold. It's not at all clear how well it will work in an environment with lots of Fiber Channel attached disks, for example. Which is fine, except that companies like Red Hat and SuSE have tied their community distro to folks who only fundamentally care about the desktop, but their cus [03:15] tomers are folks who need to worry about things like full init.d backwards compatibility (including system V init levels!!!) and things like huge numbers of FC attached disks where it may not at all be scaleable to enumerate them all at system startup. [03:15] What this is going to mean for future enterprise customers at Red Hat and SuSE is going to be entertaining to watch, but I for one will be switching to Debian Testing because systemd looks like a slow motion disaster from where I sit, and even on my laptop, I don't think I want it. GNOMEos? When GNOME is actively hostile to power users? I don't think so...."-Theo Ts'o on systemd [03:15] interesting [03:16] DonkeyHotei, maybe. but not likely. why? [03:16] philipballew, like change what? [03:17] In the Buddhist sense? [03:17] i need a favor from someone in that area with bash-fu [03:17] what kind of favor? [03:17] the errand kind [03:18] I can ask people easily probably [03:18] Go meet a Columbian in the park [03:18] I do know a good 20 percent of the pop [03:18] He will be wearing a raincoat and a hat... [03:18] MarkDude, Il be wearing a shirt and pants [03:18] by favor, i mean $0.00 [03:18] sit next to his and grab his case , he will take yours [03:18] totally [03:18] i figured DonkeyHotei [03:19] they will be matching [03:19] big plus if they have ipv6-fu [03:19] MarkDude, doing sketchy activities again? [03:19] I doubt auburn does [03:19] * MarkDude was saying that is DonkeyHotei 's favor he was asking [03:20] I have a box thrown in auburn that I can let you access if you need as well [03:20] MarkDude: not too far off, except for the ipv6 [03:21] philipballew: strictly on-site [03:21] hum. Well what is it you need? [03:21] and not quite auburn; i just figured sac was on the way to auburn [03:21] also, people in auburn dont lock down their wifi [03:22] It is, I will be in Davis on Monday I know for sure [03:22] If I am not there tomorow ill be there both sunday and monday [03:22] i'm referring to tomorrow in particular [03:23] ah, well I will not be [03:23] I just confirmed [03:23] well, you did say you know 20 other people... [03:23] I do, but in Auburn, not Sacramento [03:23] ah [03:23] scientes: glad someone had the same thought [03:23] well, was worth a shot [03:23] for sure [03:24] maybe I should just write the patch. but I've never played with vfs or container implementation. [03:25] whoa, tons of traffic since then. (re mount + clone-newuser) [03:27] scientes: user namespaces aren't perfect, iirc. can you say "map 0 to UINT_MAX in this ns to the single uid 1000"? [03:44] geofft, do the ext filesystems support full 32-bit UIDs? [03:44] anyways you could always just map the first 16-bits [03:45] even 12 bits would suffice for most use cases [03:46] geofft, the rest are just mounted to the special "out-of-range" uid, and because you are root in that namespace, can be changed to something within range [03:49] geofft, we would assume that if being mounted that way, you wouldn't then be having extensive use of sub-namespaces---i.e. you wouldn't then have large amounts of uids on the device mapped to "privlidged" and globally registered UIDs, only "old-style" uids [03:50] geofft, all you have to do is have a sane management of the distinction of kuid and user-space uid [04:10] pleia2, does the people in charge of a uds meeting need to subscribe people to a blueprint? [04:10] people subscribe themselves [04:11] philipballew: not sure what you mean, can you give me an example? [04:12] the blueprints we had open in summit, those turn into launchpad pages sooner or later and people are subscribed to them. Is that automatic, or does someone do that? [04:12] pleia2, [04:12] philipballew: when you create a blueprint it becomes a launchpad page (should redirect you to the page it just created) [04:12] then people can subscribe to it [04:12] the owner has to set things like Approver, Drafter, etc [04:13] alright, the people who attended the meeting are already subscribed? [04:13] you aren't required to subscribe to attend [04:13] people who are interested in tracking the blueprint through the cycle subscribe [04:13] oh, my bad [04:13] okay. Ill need to see if people subscribed to the one I made then [04:16] Is there a list of these pages or does my summit page have the link, as I dont see the link on the page? [04:16] "these pages"? [04:16] I only have my summit page I used durring the meeting I made http://summit.ubuntu.com/uds-q/meeting/20717/community-q-ubuntu-youth/ [04:17] there is a blueprint link near the top [04:17] ... [04:18] I spent a good 10 minutes scrooling through the blueprint looking for it... [04:18] geofft, (uid_t)-1 and (gid_t)-1 are reserved as an internal error value. [04:18] Most uid/gid setting system calls treat these value specially anyway [04:18] so attempting to use -1 as a uid would likely cause entertaining [04:18] failures in userspace. [04:18] * philipballew shoots himself in the foot [04:18] wait, that means that uids are 31-bit not 32 i guess [04:33] scientes, should still be 32-bit. it's presumably the same deal as IS_ERR [04:33] This maker fair thing looks sick [04:34] geofft, well if they allow -1 then it has to be 31 bit, unless they are doing it 64-bit in the kernel, which doesn't seem wise [04:34] scientes: pretty sure ext supports 32-bit UIDs; I use several systems with more than 64k accounts [04:34] Does anyone know if the cal train lets off near there? [04:35] scientes: but in any case, mapping overflowuid in the host would work mostly fine [04:35] philipballew: It does ... near enough, anyway. [04:35] geofft, yeah but do you have multiple detachable user-controlled devices that have more than 64k accounts? [04:35] philipballew: The maker faire site should give directions from the caltrain station. [04:36] alright, alright, that sounds good. I have mapped out how to get there from the Richmond station. [04:36] and a dennys 1.8 miles from it. [04:37] thanks akk [04:43] scientes, no :) [04:43] (I'm also vaguely unsettled by the concept of different namespaces mapping to the same overflowuid, but that's a separate question) [04:44] geofft, yes, that was an issue at the uds conf i attended to as well [04:44] it means that ALL unix-like systems have to pre-allocate that UID as a user, to prevent it from being auto-allocated [04:45] seems to me that just having uid's be signed, and the lower half reserved by the system would be far saner [04:45] but i guess that decision is already over ;) [04:46] then we could cap the overflowuid to a well-known negative number, and have it linked to a capability (now that capabilities are usable inside user namespaces) [04:47] djb has this funky proposal that's that except backwards -- all odd uids are system-reserved, or something like that [04:47] oh geeze, thats madness [04:47] aka big-endian signed-ness [04:49] i can see that it would be sane to attach the sign to the little-end of the number, instead of the more traditional big-end [04:49] that way when you ask for a user name-space you could even concievably decide how many bits of uids you need [04:50] and then the kernel can just do a simple truncation of the number---IF the sign is at the little end [04:51] but that really is just a quibble, nothing from with using 7 or E in a bitmask [04:51] geofft, link? [04:52] gos, im rambling [04:52] http://cr.yp.to/unixaccount.html [04:52] "I suggest using odd numbers for all site-specific account IDs, so that even numbers can be safely allocated to global accounts." [04:53] and every djb-allocated uid is some even number greater than 2^30 [04:53] if they were in equal quantities, that wouldn't be so ridiculous, but they're not [04:58] jtatum: Emailed. [04:58] DonkeyHotei, its rediculous if only for the reason that it is super confusing [04:59] people expect numbers to be *sequential* [04:59] right [05:15] geofft, see, this would be a great use of those negative UID's, cause we could map the entire user-drive to negative numbers behind the scenes [05:15] so that it wouldn't get in the way of any possible UIDs elsewhere in the system [05:20] bkerensa, around? [05:57] philipballew: now I am [06:12] bkerensa, so were you givin any details about the hp cloud service? [06:13] philipballew: I'm waiting for Mr. Ash to hit me up [06:13] philipballew: why? I thought he set you up [06:13] He said he did, I am just wanting to make sure my beta never ends so they dont start charging me [06:19] philipballew: uhh do you have a card associated with ur account? [06:21] bkerensa, well yeah, its still there since i had it in the privitee beta [06:21] I was going to email the guy this weekend [06:21] ahh [06:21] yeah I would e-mail him in advanced [06:21] beta is over and unless he did something on their side then you will get charged [06:21] before i do any serious hacking [06:22] I think he did, because he said he did. but I have not seen any thing on my webpage [06:41] Mapping public transit is annoying [06:47] heh [06:49] Saturday means I have to ride my bike 15 miles at 6 in the morning [06:49] what do that think I am, a morning person? [06:51] First World Problems [06:51] seriously === izdubar is now known as MarkDude [17:54] Anyone know anything about the developer.ubuntu.com process and Ubuntu App Center? I'm trying to figure out what repo contributed apps go into. [17:54] #ubuntu-app-devel is just crickets. [18:01] akk: pretty sure it's extras.ubuntu.com [18:05] Ah, interesting, I don't have that enabled -- I'll try that. [18:05] There's apparently only one contributed app anyway (only one that shows up in the software center), leds [18:06] it's very early in the cycle and they don't persist between versions automatically [18:07] you have to go through a process to get your app in precise after precise is released, there are more in oneiric [18:07] oh, interesting. I assumed that the most apps would be for the newest release. [18:08] http://extras.ubuntu.com precise Release: The following signatures couldn't be verified because the public key is not available: NO_PUBKEY 16126D3A3E5C1192 -- where would I get keys for it? [18:08] I also need to figure out how to get my own PPA's key, the PPA docs don't explain that. [18:09] should just be able to use the typical gpg --keyserver keyserver.ubuntu.com --rec 3E5C1192 [18:09] er recv [18:11] Just the last 8 digits of the key it lists in the error message? [18:11] yeah, just like personal gpg fingerprints [18:12] the key id is always the last 8 digits [18:12] Oh, does it matter if I do that as me or as root? [18:13] Maybe that's why it happens for my PPA, because I have my key but root doesn't? [18:13] you can do it as a normal user, it just gets you the key, then you use sudo to put it in to apt, lemme find the command... [18:13] gpg --export --armor 3E5C1192 | sudo apt-key add - [18:15] Thanks! [18:15] sure :) [18:17] It looks like I have to do the keyserver step even for my own ppa -- does ubuntu make a separate key different from my own key that I used to sign it? [18:19] not sure [18:29] for ppas, if you use apt-add-repository, it does it for you [18:35] It didn't for me, jtatum [18:35] That's how I added the repo, but it still complained about the key. [18:35] But I added the repo early on, before it had finished accepting and building the package -- maybe it doesn't set up the key until it has packages. [18:37] hmmm [18:37] strange [18:37] via the "ppa:xyz" type command line? [18:38] yep [18:45] what the heck! very odd. usually the output of that says something about the keys its adding. [19:03] philipballew: I just sent me info over to HP for the lifetime comp cloud acct ;0 [19:03] philipballew: Ash CC'ed me when he sent a e-mail to all the cloud people at Canonical too [19:04] so what does that mean for me [19:04] and you? [19:04] bkerensa, [19:05] How do I know I have it like he said? [19:05] philipballew: idk what it means for you? You should probably e-mail him.... I just know he is comp'ing all developers hacking in the cloud [19:05] do you have his email? [19:05] you can pm it to me if thats better [19:07] if not bkerensa thats cool [19:07] I just pm'ed it [19:08] I mean if he said you are good to go at UDS then you should be assuming you provided him a acct id and tenant id? [19:09] i gave him my business card that had my email associated with my account. This is the hp guy at the booth [19:09] yeah you better get in touch because they need an acct # and tenant id which only you can provide [19:11] bkerensa, the guys name is ash that emailed me. do you knoow where to find the guy on irc? [19:11] ? [19:11] I dont think he uses irc [19:11] oh, well I do have his phone number. [19:11] * philipballew is sad that everybody does not use irc... [19:12] I think I will call him [19:13] most people do not use irc [19:13] the fraction of the population that does is very insignificant in size [19:14] It's amazing how many tech/OSS projects don't have IRC channels. [19:15] well I called him, maybe I will email as well. Thats usually polite [19:17] bkerensa, did you still have to enter your credit card? [19:18] I hate "free" things that require a CC number. [19:19] * akk wants to get a $1 gift card to use on such sites [19:20] I agree, I always feel unsure. [19:20] hints the=is email and all [19:20] *this email [19:20] akk, all ready for the eclipse? [19:22] Yep, I think so! [19:22] philipballew: I have my card on file because my account pre-dated the offer for comp [19:22] yeah, me to [19:23] i called and emailed him, gave him two methods to contact me. Figure he should get back to me in the next few days [19:24] generally not a good idea to call and e-mail someone =/ might irritate them [19:24] hum, I was taught to do that at my last job. [19:24] hmm [19:24] call/e-mail same day = urgency [19:25] I usually figure it allows people two methods to contact me. easier for them [19:25] if he asks ill say I am not in a hurry [19:29] maybe its a so cal thing [19:39] How safe of an area is San Mateo where this fair is this weekend? [19:45] It's not super snazzy, kinda low-rent, but I don't think it's dangerous to walk around. [19:46] Didn't feel at risk walking back to the car (way across town) last year. [19:47] And the Maker Faire area itself is super crowded, unbelievably crowded. [19:47] alright. Thats the impression I get. I need to find a good spot to put my bike. Maybe I can keep it at the booth or lock it somewhere [19:48] Its apparently a big event [19:48] Ah, bike safety can be an issue even in fairly nice areas (I had a bike stolen in palo alto). [19:48] But if you find a crowded area to lock it, hopefully bike thieves won't be doing anything obvious like breaking locks with lots of people around. [19:49] If you can lock it somewhere near the faire gates (not somewhere like the train station) I'd think that would be safe. [19:50] true. I need the bike to ride 15 miles tomorow morning to catch the bus so I kinda need it [19:50] stupid taxi in Oakland added an extra $10 to my card >.< [19:51] ive never riden a taxi. They seem like nice things though [19:52] http://makerfaire.com/bayarea/2012/alternative.csp : FREE, secure valet bicycle parking is located on S. Delaware Street (south of the West Gate near 25th Avenue). Bicycle locks are encouraged. Valet service provided by the Silicon Valley Bicycle Coalition. [19:52] I googled "maker faire" "san mateo" bikes [19:53] Also you can get discount discount tickets if you bike (same page). [19:53] er, only one discount :) [19:55] What if I ride two bikes at once akk :) [19:56] philipballew: This being maker faire, they might give you credit for it if you looked entertaining. :) [19:56] lol [19:56] but thats good to hear, I think im going to leave here at 530 to get to sf by 8 and get to the maker fair by 9 30 [19:59] mapped it out, what could go wrong. [20:00] besides a flat tire, get lost, bike stolen or dont make the bus [20:02] "Restart lightdm? Y/n" NO WAIT JFLKDSDF [20:02] sigh :) [20:03] I have to run service lightdm restart every few days [20:03] that's not good :\ [20:03] this is just from an update [20:04] yeah, I have run the unity daily. might be related [20:05] though I purged that a few weeks back [20:08] pleia2, how often do you update? [20:11] But at least it's not apache. :) /me currently fighting with apache yet again to turn on php for another vhost [20:16] After I write anything in PHP, I always feel the need to wash my hands really well... [20:16] it never seems to do what it is told [20:17] I only use it for simple stuff -- adding sidebars/footers, or a simple image gallery with prev/next buttons. [20:17] It's pretty good for that. [20:17] I'd hate to write a real app in it. [20:17] Especially since apache keeps changing the rules for how to enable it, sigh. [20:18] yeah, you can write apps in it iirc, but you used to not have as much functionally correct? [20:22] Just got a call from a friend from high school who is coming to the fair. Offered to pick me up in Vallejo. [20:23] Everything seems to work out [20:24] Hopefully that means not starting at 5:30 too. [20:27] yeah. hes gonna meet me at like 830 in Vallejo. So ill leave at 8 on my bike to meet him near the Marine World area. [20:27] Or whatever they call that place these days... [21:57] philipballew: whenever I feel like it or there is a particularly unpleasant security alert I want to update my system for [22:04] hum, seems logical. I usually uninstall the update manager as its a nuisance to be reminded to update so I sometimes forget. [22:04] but I update servers not very often [22:05] my debian server is only say every couple months [22:47] apt-unattended-upgrades [22:53] jtatum, I like to see what I am upgrading so I know whats going on. Helps me keep control in some ways [23:14] philipballew, makerfaire? [23:41] Hi [23:41] Hello Everybody [23:42] Is anyone ready to help me out??... [23:42] heh [23:53] Im an American damnit. I dont have all minute.... [23:53] :D