[02:21] ddecator: you up for more SM2 testing? [02:31] micahg: sure [02:31] ddecator: k, I think I almost added all the missing components [02:31] micahg: ok. should i downgrade first? [02:32] ddecator: nah [02:33] micahg: alright, just let me know when [02:53] wow, flash is crashing like crazy in chromium [02:56] ddecator:reason to use Firefox :) [02:56] micahg: i know, i want to, but only FF 3.6 is working right now and that's really outdated as far as add-ons and such, so i am waiting for 3.7 to work again =p [02:56] ddecator: outdated? [02:57] micahg: i have add-ons installed on 3.6 i don't use on 3.7 and i'm missing some that i use on 3.7 so rather than update it, i'm just using chromium to hold me over [02:57] it's the lazy way.. [02:58] ddecator: k, 3.7 should be fixed tonihgt [02:58] micahg: good to hear =) [02:58] let's see if we can get the rest fixed tonight :) [03:00] sounds good to me [03:24] micahg: any idea around what time SM2 will be ready? [03:26] Hi [03:26] ddecator: another hour at least [03:26] still working out dependencies, then 1 hr to build [03:26] hi Milos_SD [03:27] micahg: alright, i'm gonna get off for a bit and relax, i've been working on homework and other things all day. i'll get on a little later [03:27] after xulrunner 1.9.3 update, Firefox 3.7 doesn't work (it starts, but can't open any page), I can only open Help -> About and nothing else :S [03:27] ddecator: don't worry about it [03:27] Milos_SD: we know =) [03:27] Milos_SD: should be ok after tonight [03:27] Milos_SD: will be pushed in another hour [03:27] great :) [03:28] micahg: i'll still be on later, theres some other things i want to do, so i can test SM2, i just need a short break, haha [03:28] I thought it was something with my configuration ... becouse it happend after I uninstalled some addon... [04:34] micahg: i'm back, so just let me know when SM2 is ready [05:43] ddecator: at this stage of the game, you should probably make sure bugs are filed for your gwibber issues [05:44] micahg: i've been meaning to, haven't had time to get around to it, but while i'm waiting for you i'll take a look ;) [05:45] * micahg is fading fast and needs time [05:46] micahg: np [05:58] * micahg doesn't know how this is supposed to be finished tonight :( [05:58] micahg: you seemed pretty confident earlier =p [06:00] * micahg didn't realize how long this takes [06:00] ddecator: it's manually assigning the proper isntalled files to each of the .install files [06:00] micahg: ick.. [06:08] ddecator: k, I think I only have about 20 more files... [06:08] micahg: sure thing. i'll be up at least 2 more hours anyway [06:14] ddecator: ok, I think I got this part finally [06:14] micahg: alright, SM2 or FF3.7? [06:16] ddecator: SM2 :) [06:16] micahg: alright. you said it needs an hour to build? [06:16] I'm testing local first [06:16] k [06:17] ff37 i386 is ready [06:18] amd64 still building? [06:19] ddecator: yep [06:19] darn [06:25] ddecator: I think I got it [06:25] micahg: good deal =) [06:25] ddecator: do you want me to just upload hte binaries I have to p.u.c rather than build in PPA? [06:26] micahg: doesn't matter to me. if you want others to test it then you can do the PPA, but if i'm the only one then p.u.c is fine [06:28] ddecator: ? [06:28] micahg: do you want multiple people to test it, or do you just want me to? [06:32] ddecator: you at the moment [06:32] micahg: p.u.c is fine if that will be faster [06:32] I hope so :) [06:33] ddecator: I have limited upload from here [06:33] but it'll be less than an hour [06:34] ddecator: I just uploaded the build as well [06:34] micahg: alright [06:34] ddecator: p.u.c still going, should be another 15 min [06:34] micahg: sure thing [06:36] i spy some xulrunner updates.. [06:39] and that did nothing [06:39] still broke? [06:39] you have ff37 yet? [06:40] ah, nope, that update hasn't reached me yet [06:40] just checked [06:45] there it is [06:46] FF3.7 is a go =) [06:47] well, kinda...OOPP still isn't working for me [06:48] ddecator: you have it enabled? [06:49] micahg: i just enabled it and it froze when i went to youtube. i don't see a separate listing just for flash though, which i've seen people talk about in reports for 3.6.x [06:50] * micahg is upgrading [06:52] ddecator: apparmor :) [06:53] micahg: great..so what do i need to do? [06:54] i'm also trying to reproduce a bug i keep having.. [06:54] not going too well though, maybe it was fixed =p [06:55] ddecator: sudo aa-complain /etc/apparmor.d/usr.bin.firefox-3.7 [06:55] micahg: do i need to restart for it to take effect? [06:55] ddecator: that'll put the profile in complain mode until it's fixed [06:55] ddecator: no [06:56] good deal [06:57] micahg: what is the name of the flash process? [06:57] oh, plugin-containe? [06:57] well would you look at that =D [06:58] ddecator: yeah [06:59] i'll keep my eye out for the other bug (before, my bookmarks icon wasn't showing up half the time, can't get it to happen now) [06:59] jdstrand: if you get a chance can you look into allowing plugin-container in xulrunner-1.9.3 for the firefox-3.7 apparmor profile? [07:00] micahg: what's the status on SM2? [07:00] ddecator: uploading to p.u.c seems to not work :( [07:00] micahg: darn. so ppa then? [07:00] * micahg will upload to my server [07:01] actually, I can probably use Ubuntu One, you have an account? [07:01] micahg: yup [07:08] ddecator: k, I think it's uploading [07:10] micahg: great. haha, i was in FF withdrawl.. [07:33] ddecator: ok, uploading the last deb to uBuntu one now [07:34] micahg: k [07:35] ddecator: sent share invite, go for it [07:35] micahg: alright, one sec [07:35] * micahg is going to bed, will try to finish the bzr part in the morning === almaisan-away is now known as al-maisan [07:41] ddecator: do you have it? [07:42] micahg: something went wrong (site issue) let me try again a sec [07:43] micahg: ok, there they are [07:45] ddecator: k, PPA build should be done in about a half hour anyways [07:45] ddecator: I'm off will be back in 5 or so hours [07:46] micahg: alright, i'm installing now [07:46] micahg: night [07:47] micahg: everything works =) [07:48] great news, I'll finish packaging in the morning [07:48] night [08:47] The Sun is High, the Vampires down 0/ [08:48] ... and ff 3.7 is broken [08:48] again [08:48] lololol [08:48] ajax silently fails [08:48] wth are you guys doing :) [08:48] i'm no longer touching it [08:48] not my fault [08:49] the same code works fine with 3.6 [08:49] and chrmomium [08:49] -m [08:50] FIXIT [08:50] no, i don't care enough to loose my time with it [08:51] i mean, i'm already short in time so i spend it where it's needed [08:53] !info js2-mode [08:56] FF 3.7 broke again? [08:57] i'm using it right now without any issues [09:25] it's either an ajax issue, or eval() [09:26] most probably eval as firebug shows the expected result from my ajax requests [09:27] but nothing in the console to give a clue [09:54] fta ch snapping on ssl, again [09:54] new profile doesn't reproduce it [09:54] starting now in debug mode [10:04] wfm, as usual [10:11] still working fine in debug [10:11] I'll let it run for a bit more to see if it blows while in debug [12:43] asac - i see that FF3.6 still fails to build on hardy (it's using some gtk functions that don't exist in hardy) [12:43] are you ok with me adding a patch to make it compatible with gtk on hardy? [12:43] it doesn't look too difficult [12:48] oh, it's actually the mozilla-kde patch which makes it depend on a newer gtk [12:48] that's easy then [12:55] chrisccoulson: yes, but do it in a "stable" branch [12:55] we need to update mozilla on us changing the patch for this sake etc. [12:56] before rolling it [12:57] asac - the mozilla-kde patch comes from mozilla? [12:59] chrisccoulson: no ... thats why we need to keep them updated. [12:59] chrisccoulson: we committed that you would upstream it after release afaik [13:00] ;) [13:00] or something like that [14:03] hi chrisccoulson [14:10] hi micahg [14:13] chrisccoulson: ok, so I got SM2 working, now to get the packaging back into bzr [14:14] chrisccoulson: how long do you need before you can upload once I do that? [14:14] i'm not sure, i shouldn't need too long though [14:14] chrisccoulson: k [14:41] chrisccoulson: should we follow the convention for TB/FF in teh SM changelog? [14:55] chrisccoulson: also, due to time constraints, if I'm just adding bzr commits as needed if I notice an error previously [15:40] chrisccoulson: should I install the gnome-support files in the main package for Seamonkey like we do for Firefox/TB? [15:41] micahg - yeah, that makes sense [15:41] chrisccoulson: k, I think that's it then, there are a whole bunch of lintian warnings/errors though [15:43] chrisccoulson: do I leave a blank .install file then? [15:43] i don't think you need to leave a blank file [15:48] chrisccoulson: so, I'm going to do one more test build, do you want me to push up my bzr changes for you to review? [15:48] micahg - yes please [15:48] i will try and look at those later on then [15:48] chrisccoulson: should I use the team branch or a personal branch? [15:49] the team one is fine [15:50] chrisccoulson: k, pushed [15:51] thanks [15:51] chrisccoulson: I guess I'll do lightning on my way into $WORK [15:56] oh my! 3.7 is so broken [15:58] fta: ? [15:58] fta: flash? [15:58] no, everything [15:59] * micahg only knows of the apparmor/flash issue [15:59] the context menu has a hundred entries or so [15:59] micahg: what exactly did you need me to change? [15:59] my ajax are all broken [15:59] jdstrand: plugin-container in xulrunner193 package is blocked by apparmor [16:00] * micahg thinks ff37 should mirror ff36 but that's a project for later [16:01] * micahg doesn't have a huge context menu [16:06] whats the status of seamonkey? anyone knows [16:06] ? [16:06] chrisccoulson: do you know if the package is done? [16:07] asac - micahg has been working on that, and just pushed the changes to bzr [16:07] chrisccoulson: are the changes good? [16:10] micahg: can you give me the dmesg? [16:10] asac - i've not reviewed them just yet [16:10] jdstrand: http://pastebin.ubuntu.com/422815/ [16:11] micahg: and fyi, ff37 does have (effectively) the same profile as ff36 [16:11] jdstrand: right, but it's a xulrunner build, not all-in-one [16:12] micahg: what types of things are in plugin-container? [16:13] micahg: specifically, why does it need 'x'? [16:13] jdstrand: it's a wrapper for flash and other plugins so it doesn't kill the browser === nikolam_ is now known as nikolam [16:13] s/wrapper/sandbox [16:14] micahg: can you test it with: [16:14] /usr/lib/xulrunner-*/plugin-container ixr, [16:17] jdstrand: seems to wfm [16:18] micahg: cool. also-- should we add a similar rule for the all-in-one build? if so, what is its path? [16:19] jdstrand: it's in the firefox dir [16:19] that seems to work [16:19] * micahg forgot he's not running enforce on the profile though [16:21] jdstrand: seems to work on the ff36 profile w/out anything extra [16:21] that seems odd [16:21] /usr/lib/firefox-3.6.4/** ixr, [16:22] that's why it works [16:22] yes, I see that now [16:22] ok good [16:24] micahg: committed [16:24] jdstrand: thanks [16:25] I'm going to put it in ff36.head too [16:27] jdstrand: if you think it's necessary [16:27] chrisccoulson: are any of these lintian warnings an issue: http://pastebin.ubuntu.com/422823/ [16:27] mostly I want things to stay in sync [16:32] micahg - i'll have a proper look at those in a bit, but we can add overrides for some [16:32] (like the image-file-in-usr-lib) [16:37] micahg, are you aware of any change related to ajax in ff trunk? === vish is now known as Vish [16:37] fta: well, I think they're reinventing their JS engine [16:38] micahg, i use http://www.prototypejs.org/api/ajax/request it used to work fine with ff 3.5/3.6/3.7 and chromium, now it's broken with 3.7 === Vish is now known as vish [16:39] i just get the 'onCreate' but nothing else [16:39] no error, no nothing === vish is now known as Vish [16:47] micahg, ?? [16:47] micahg, at a5? [16:53] rickspencer3: hola, do you happen to know a contact person (here in irc) for ubuntu partner repo questions? === |eagles05138785| is now known as eagles0513875 [17:02] sebner, hmmm [17:03] I guess langasek in #ubuntu-devel [17:04] rickspencer3: nah, I don't mean archive admin stuff, more like application/company stuff [17:04] hmmm [17:04] sebner, what's your question, specifically? [17:05] (can use that to direct you) [17:06] rickspencer3: to specific about the valve steam linux client rumors, canonical could offer help/whatever or make a deal like they did with dell and both sides would benefit [17:06] oh, right [17:07] sebner, that seems like a business <-> business thing, so not sure how we can help there from the engineering side [17:09] rickspencer3: well, some people think valve is not gonna offer something like that officially because it's too much work for such few people (just a fun hacker project from some valve employees), I would just be interested to know if canonical already thinks about helping valve or if not how we can talk to the specific people responsible so they start thinking about it [17:14] fta: what about a5? [17:14] sebner, oh, yeah I don't know what Canonical is thinking there, if anything [17:15] rickspencer3: that's why I'm trying to get in contact with someone there and I thought you might have contacts or know some responsible for such stuff [17:15] sebner, hmmm, I don't know how to do that [17:15] micahg, changing the js engine [17:16] rickspencer3: oh, ok. np then :) [17:16] fta: yeah, idk if they'll have it complete for 3.7, but that's one of the major goals is to try to make a faster engine than v8 [17:16] fta: I think the early alphas were mainly for the OOPP stuff [17:18] asac: I thought we weren't going to backport the kde-patch to the stable 3.6 branches [17:22] micahg: well. if thats easier, then yes. [17:22] otherwise no [17:22] i leave it to chrisccoulson to decide [17:24] asac: k, I'll discuss with him [17:40] chrisccoulson: I'll bbi 75 min === al-maisan is now known as almaisan-away [17:52] ccheney - i had a look at epiphany today [17:53] chrisccoulson: great :) [17:53] i don't know how much you unserstand about the crashes, so i don't know if i'm duplicating work here [17:54] chrisccoulson: all i understood was that asac told me that the callbacks where going to the wrong location and that if i updated some of the linkage in the class init it would probably fix it [17:54] chrisccoulson: though i was not clear on how to do that properly [17:54] yeah, i'm not entirely sure it's possible to do that, as there aren't really any class methods to override [17:55] asac: ping ^ [17:55] the issue is that GtkEntryCompletion is using the GtkEntry calls from the real GTK, rather than the local EphyGtkEntry, and the text storage is different between the 2 versions [17:56] chrisccoulson: perhaps asking asac what he meant might help, i just pinged him to see :) [17:56] chrisccoulson: oh that might be a bug that I didn't fix when trying to restore what i had done [17:56] chrisccoulson: the GtkEntryCompletion copy in epiphany-browser should be using EphyGtkEntry, unless i forgot to switch it over [17:57] the old system gtk is just using a char** to store the data in the entry box, whereas the new one (local to epiphany) is using a GtkEntryBuffer instead [17:57] chrisccoulson: i had at one time gutted that part out so I might have messed up when trying to restore it [17:57] ah [17:57] ccheney - yeah, i think the issue is that not all of GtkEntryCompletion is included [17:58] so that whole class would need to be copied over as well as modifying it to use EphyGtkEntry? [17:59] ccheney, yeah, i think so [18:01] ok will try that out when i get a chance, working with mvo to fix another OOo issue, this one appears to be exposing a dpkg bug (not apt this time) :-\ [18:02] chrisccoulson: there are class methods to override in GtkEntryClass .. no? [18:03] that was an attempt to break the loop ... otherwise we can see if porting GtkEntryCompletion as EphyGtkEntryCompletion helps to unbreak the tie [18:05] mvo is going to be looking at the OOo/dpkg issue in more detail later today so i can work on the epiphany stuff right after lunch (30m or so) [18:05] asac - yeah, i think we'll have to copy over the whole GtkEntryCompletion class [18:06] chrisccoulson: GtkEntryClass doesnt exist? [18:06] asac: i have the most up to date version of the package in my ppa if you want to take a look [18:06] maybe we can include the private header? [18:06] ccheney: i doubt things changed sinc we did it the last time ;) [18:07] its possible i somehow botched something when restoring the changes [18:07] well. if its still crashing the same way its probably ok [18:07] asac: last we talked i had made some changes you thought were bad so i went back and tried to fix them, i think i did it correctly but maybe not [18:07] ok [18:08] yea i have EphyGtkEntryClass [18:08] asac - GtkEntryClass does exist, but there aren't any class methods we can override (unless i'm missing something) [18:08] chrisccoulson: i see them in gtkentry.h [18:08] its just internal API ... but thats ok given what we are doing [18:08] struct _GtkEntryClass [18:08] { [18:08] GtkWidgetClass parent_class; [18:08] /* Hook to customize right-click popup */ [18:08] void (* populate_popup) (GtkEntry *entry, [18:08] GtkMenu *menu); [18:08] /* Action signals [18:08] */ [18:08] void (* activate) (GtkEntry *entry); [18:08] void (* move_cursor) (GtkEntry *entry, [18:08] GtkMovementStep step, [18:09] gint count, [18:09] gboolean extend_selection); [18:09] so we want to overload those ... and i think there was a new func in the new API [18:09] we want to properly use the reserved1 slot in the class definition for that [18:09] right there were 3 reserved in hardy and now just 2 ... so we just use the gtk_reserved1 or something from the old GtkEntryClass for that [18:10] and write the wrapper func that calls klass->gtk_reserved1 ... or something [18:14] chrisccoulson: also use override for all properties rather than install_property ... and similar for signals [18:14] except for new properties of course [18:18] asac - ok, i will look at that. the reason for the crash though is that the old system GtkEntryCompletion class uses gtk_entry_get_text (and passes the EphyGtkEntry class), but that does the wrong thing in hardy, and doesn't seem to do anything which we can override [18:20] chrisccoulson: err. if we overload that it should call the right thing, not? [18:21] chrisccoulson: could be that something instanciates the GtkEntry inside the gtk ... but afaik the instantiation happens in our code, so instantiating EphyGtkEntry should be ok [18:22] gtk_entry_get_text calls klass->get_text most likely. so if we ovelroad it, we get called [18:23] asac - the issue though is that gtk_entry_get_text doesn't call klass->get_text. it just returns return entry->text [18:24] but entry->text is not used to store the contents with EphyGtkEntry (the contents are stored in a GtkEntryBuffer instead) [18:26] interesting [18:26] overloading gtk_entry_get_text ;) [18:26] try that [18:28] let me update my git gtk tree [18:29] i am looking at something half a year old ... that still has entry->text [18:31] asac - the use of GtkEntryBuffer is since 2.18 [18:32] so, from karmic onwards [18:32] * ccheney back [18:32] asac: btw are you still in charge of NM? i have a question for later if you are [18:35] * ccheney pings cyphermox about it instead :) seems he may have taken over it [18:36] chrisccoulson: so you could overload gtk_entry_get_text ... which tests if its a EphyGtkEntry and call the buffer and if its not return entry->text ;) [18:36] you can also try to pull in the CompletionClass [18:36] or maybe both [18:37] pulling in too many classes leads to madness ;-) [18:37] right. thats why i tried to break the loop here [18:38] as they seem to have near infinite dependencies :-\ [18:38] yea [18:38] or we really register to the inserted/deleted text signals and update entry->text in the super struct [18:58] asac: are sunbird/lightning exposed to insecure content? [18:59] ccheney, sorry I was out to lunch [18:59] cyphermox: np, will take it to privmsg to avoid clogging the channel [19:09] chrisccoulson: I forgot the replaces/conflicts for seamonkey-gnome-support [19:23] chrisccoulson: how does it work with shlibs:Depends on -gnome-support since nothing is actually in there [19:43] micahg - the shlibs:Depends is populated before the so file is moved [19:43] so it gets the correct dependencies [19:44] chrisccoulson: ok, but if I removed it from .installl, how does it know? [19:45] micahg - oh, i've not looked at it yet. as long as the files are installed to debian/seamonkey-gnome-support (or whereever) when dh_shlibdeps is run, then it will be ok [19:46] chrisccoulson: so, that won't happen without the files, right? [19:46] micahg - not unless there are some dh_install calls in debian/rules installing them to -gnome-support first [19:47] chrisccoulson: k, so are they supposed to be installed and then removed? [19:47] micahg - yeah, that's the idea [19:49] * micahg sees what you did [19:55] micahg: chrisccoulson: if -gnome-support is kept empty the idea would be to hack the depends out of the main package nd put them manually there [19:55] even if there are no .so's [19:55] asac: I'm doing what chrisccoulson did for thunderbird [19:55] looks sane [19:55] and that was copied from firefox ;) [19:55] chrisccoulson: what are you doing for tbird? [19:56] i thnk there is a bug in firefox and we might want to do it different (though the idea was great ;)) [19:56] asac: http://pastebin.com/bBqdtqkw [19:56] asac - for thunderbird - bug 543060 [19:56] Launchpad bug 543060 in thunderbird "thunderbird - gnome integration should work even without -gnome-support" [High,Fix released] https://launchpad.net/bugs/543060 [19:57] ok cool. i think there was a problem with predeb if that was the stage i used in firefox [19:57] e.g. those binaries dont get stripped properly or sometihng [19:57] but cant remember [19:57] ok on a call again :) [19:57] * micahg is testing the install for it now [19:57] asac - yeah, that's the stage used in firefox [19:58] i'll look at the issue with stripping then, but it's not a major issue [19:58] no its not a major issue. just some symbols in the .so that should be in the dbgsym ;) [19:59] asac: are sunbird/lightning exposed to insecure content? [19:59] micahg: somewhat. [19:59] you click on ics links etc. [19:59] but not that badly [19:59] asac: :(, idk if I can get sunbird done [20:00] asac: ui ui ui, not that long ago I heard that you left the desktop/mozilla team. Such news passed me -.- [20:00] asac: was thinking to leave it and backport newest version [20:00] sunbird has more patches than anything I've seen [20:00] sebner: i am still here, aint i? [20:01] asac: well, present in an irc chan != part of the team [20:02] asac: chrisccoulson: so, can we leave sunbird and SRU it or should we drop it and backport just lightning, or another option? [20:02] micahg: i dont know what the state is [20:02] we should try to get it up today if any [20:02] or drop it [20:02] a full new upstream version etc. in a SRU is tough [20:03] asac: I wouldn't be able to work on it till this evening (which I think is too late) [20:03] asac: I think we're pushing it with Seamonkey right now [20:03] micahg: dont we have a lightning only solution? [20:03] asac: not ATM [20:03] i remember that scottk wanted me to comment on something [20:03] thought there was a lightning only package [20:03] dropping sunbird from the sunbird package, that's what I wanted to do [20:06] asac: that's what I want to do, but it doesn't look like I can do it in time (I wanted to have lightning-sunbird build just lightning) [20:09] ccheney - do you have time to work on the epiphany issues, or do you want me to do it? [20:09] chrisccoulson: i can but i wasn't clear on what needed to be done [20:10] i'm learning as i'm doing it so its more poking around in the dark looking for a light switch ;-) [20:12] in any case i am interested to at least see what was needed to fix it and understand how it works so i can be more helpful in the future on issues such as this [20:12] with the deadline for completion looming it might be better if someone more knowledgable fix it quickly, if the deadline is as close as i understand it to be [20:13] but i am more than willing to continue to work on it if there is still enough time [20:16] chrisccoulson: I seem to be failing at the replaces/breaks her [20:16] *here [20:17] micahg - whats failing? [20:17] ccheney - ok, i can try and look at that too then [20:18] chrisccoulson: if you get to a place where its beginning to be grunt work just let me know what to do and how to continue and i will take over :) [20:18] chrisccoulson: I added a breaks on < 2.0.4+nobinonly-0ubuntu1 and replaces on < 2.0.4+nobinonly-0ubuntu1, but a few packages won't configure [20:19] * ccheney is intending to be more of a help than a hindrance in getting this done :-) [20:19] are we planning on backporting mozilla-devscripts to hardy? (or do we need to make FF3.6 build with the older version?) [20:19] asac^^ [20:19] ccheney: are you familiar with cairo? [20:19] chrisccoulson: I hope backport... [20:20] micahg - which packages fail to configure? and how are you installing the new packages? [20:20] (you have to be careful with dpkg, as it doesn't resolve things like this) [20:20] chrisccoulson: dpkg :) [20:21] chrisccoulson: how else can I install local packages? [20:21] micahg: no [20:21] ccheney: oh well... [20:23] micahg - i think there's a flag you can pass to dpkg to make it go smoother, but i ended up testing thunderbird by hosting it in my PPA and using apt [20:23] chrisccoulson: I can push to PPA, but it'll take at least an hour to build [20:24] you can use apt-ftparchive to setup a local apt repo [20:24] eg: apt-ftparchive packages . > Packages then put in your sources.list file:///foo/path ./ [20:24] or something similar to that, then you can use apt to resolve for you without needing to push to ppa to test it out [20:24]