<eldios>
well I'll check the answer in a few hours tomorrow morning. night all.
natea has joined #sandstorm
<zarvox>
eldios: is your kernel version between 4.0.2 and 4.0.4, inclusive?
<kentonv>
eldios: If you're between 4.0.2-4.0.4, there's a serious kernel bug caused by a botched backport (apparently). I would recommend not using those kernel version for anything (sandstorm or otherwise). Should be fixed in the next kernel point release, whenever that is. If you aren't on one of those versions, please file a bug against Sandstorm so we can investigate.
sasattack has joined #sandstorm
joshbuddy has quit [Quit: joshbuddy]
GeorgeHahn_ has joined #sandstorm
GeorgeHahn has quit [Ping timeout: 264 seconds]
sasattack has quit [Ping timeout: 256 seconds]
isd has quit [Quit: Leaving.]
rhapsodhy has quit [Ping timeout: 246 seconds]
paroneayea has quit [Read error: Connection reset by peer]
paroneayea has joined #sandstorm
mquandalle has quit [Quit: Connection closed for inactivity]
sasattack has joined #sandstorm
sasattack has quit [Ping timeout: 272 seconds]
mort___ has joined #sandstorm
joshbuddy has joined #sandstorm
natea has quit [Quit: natea]
gopar has quit [Quit: Leaving]
jeffmendoza has quit [*.net *.split]
mort___ has left #sandstorm [#sandstorm]
sasattack has joined #sandstorm
jeffmendoza has joined #sandstorm
paroneayea has quit [Read error: Connection reset by peer]
paroneayea has joined #sandstorm
itscassa|away is now known as itscassa
mort___ has joined #sandstorm
joshbuddy has quit [Quit: joshbuddy]
nwf has quit [Ping timeout: 252 seconds]
nwf has joined #sandstorm
jadewang has quit [Remote host closed the connection]
GeorgeHahn_ has quit [Read error: Connection reset by peer]
landspite has quit [Ping timeout: 250 seconds]
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 264 seconds]
sasattack has quit [Ping timeout: 256 seconds]
jadewang has joined #sandstorm
pixelport has joined #sandstorm
pixelport_ has joined #sandstorm
jadewang has quit [Ping timeout: 265 seconds]
pixelport has quit [Ping timeout: 252 seconds]
paroneayea has quit [Read error: Connection reset by peer]
paroneayea has joined #sandstorm
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 265 seconds]
pixelport_ has quit [Quit: Leaving]
jadewang has joined #sandstorm
natea has joined #sandstorm
jadewang has quit [Ping timeout: 276 seconds]
<eldios>
thanks zarvox and kentonv
<eldios>
checking
<eldios>
do you have anu link of the mentioned bug?
landspite has joined #sandstorm
ripdog has quit [Ping timeout: 272 seconds]
ripdog has joined #sandstorm
natea has quit [Quit: natea]
<eldios>
confirmed 4.0.4
<eldios>
to be honest though we have many servers using this same kernel and that's the first time I see this behavior
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 258 seconds]
natea has joined #sandstorm
natea has quit [Client Quit]
natea has joined #sandstorm
<eldios>
zarvox kentonv do you have a link to a description for that kernel bug?
<dwrensha>
eldios: it'll probably be another hour or two before they wake up
<eldios>
ah okay, thanks dwrensha. No hurry :)
<dwrensha>
it sounded to me like they just discover the bug a few days ago and were corresponding privately with some kernel developers
<ocdtrekkie>
It appears during initial run, EtherDraw simply copies this file to settings.json
<ocdtrekkie>
And it does not work.
jadewang has joined #sandstorm
jadewang has quit [Ping timeout: 258 seconds]
sasattack has joined #sandstorm
<ocdtrekkie>
I have fixed EtherDraw and submitted a PR. And a second PR for funsies.
gopar has joined #sandstorm
<dwrensha>
that comma is what was causing the problem?
<ocdtrekkie>
Seems so. There's one more JSON structure below that line, but it's commented out since it's an "example".
<ocdtrekkie>
So, my understanding is, that ending JSON with , } is bad, and that's what EtherDraw's template does by default.
<ocdtrekkie>
The comments thing misled me, because EtherDraw's setup does keep the comments, but the program itself parses them out at runtime.
natea has quit [Quit: natea]
<ocdtrekkie>
And I was asked if the JSON file validates by the maintainer, which it does not. But it isn't supposed to validate properly, because the comments are removed at runtime.
<ocdtrekkie>
And when I removed all the comment blocks (which confuse highlighting programs since they're invalid JSON), the extraneous comma was much easier to find.
mort___ has quit [Ping timeout: 255 seconds]
sasattack has quit [Ping timeout: 264 seconds]
jadewang has joined #sandstorm
natea has joined #sandstorm
mort___ has joined #sandstorm
mort___ has quit [Quit: Leaving.]
<zarvox>
eldios: I'm in touch with some folks; there's no tracking ticket yet until we get a CVE assigned, but I'll ping back once we have one. Are you in contact with the folks at Linode?
jadewang has quit [Remote host closed the connection]
<eldios>
zarvox: well I guess at least through regular support, being a customer with a lot of VMs with them :)
<eldios>
it was more to pass on to the rest of my team and let them know that something bad is happening with the latest CentOS7 kernel
<zarvox>
hmmm, at Linode, are they even running the standard CentOS7 kernel? I was under the impression they used their own that knows how to run under paravirt: https://www.linode.com/kernels/
<zarvox>
My guess would be that the kernel shipping with CentOS7 is actually fine - too old to have the bug introduced
<zarvox>
I'd guess that it affects all new PV instances on Linode
<eldios>
I'm definitely using ---> 4.0.4-x86-linode75 - 2015-05-21
jadewang has joined #sandstorm
<zarvox>
Yeah, that's definitely got the bug.
<eldios>
I can try with PVGRUB and the standard kernel
<eldios>
but I'm not sure if that can help
<zarvox>
If you have a kernel older that 4.0.2, then that one should be unaffected by this bug
jadewang has quit [Read error: No route to host]
jadewang_ has joined #sandstorm
dosch has quit [Remote host closed the connection]
ocdtrekkie has quit [Remote host closed the connection]
aldeka has quit [Remote host closed the connection]
ocd_cant_connect has joined #sandstorm
<ocd_cant_connect>
phildini did you kill the Quassel?
<ocd_cant_connect>
Preliminary version of EtherDraw port done. kentonv
<phildini>
huh... I didn't think I killed the quassel. let me check...
<ocd_cant_connect>
dwrensha: They work for me. The users thingy doesn't work right though in master either.
<dwrensha>
or, at least it's weird that the users button pops up on the left
<ocd_cant_connect>
Connected users just brings up the color picker. Which, on the bright side, means you won't notice I haven't implemented X-Sandstorm-Username yet.
<ocd_cant_connect>
But yeah, that's an upstream problem.
<ocd_cant_connect>
I've already submitted fixes for two upstream problems today, figured I'd see how that goes before I go too far into it.
<zarvox>
how did everything zoom in suddenly? O_o
aldeka has joined #sandstorm
ocdtrekkie has joined #sandstorm
<phildini>
ocd_cant_connect: kicked the quassel server, try again.
<ocd_cant_connect>
Someone screenshot it zarvox
<ocd_cant_connect>
And uploaded the screenshot
<ocd_cant_connect>
:P
<zarvox>
aha, I see :P
<ocdtrekkie>
Better :D
<ocdtrekkie>
Thanks phildini
ocd_cant_connect has quit [Quit: Page closed]
<phildini>
ocdtrekkie: np!
<ocdtrekkie>
So my EtherDraw todos: X-Sandstorm-Username isn't implemented (and won't be unless/until upstream is fixed), and read-only sharing, if EtherDraw supports that as easily as Etherpad does.
<dwrensha>
yay
<ocdtrekkie>
Even though the UI at the top is near identical, the code is all arranged differently. I was hoping the code would have more commonality between the two.
<dwrensha>
I'm having too much fun
<ocdtrekkie>
It works though. \o/
<dwrensha>
\o/
<zarvox>
dwrensha: we can tell, your name is in the screenshots :)
<ocdtrekkie>
lol
<zarvox>
ocdtrekkie: nice work!
<kentonv>
indeed
gopar has quit [Remote host closed the connection]
<kentonv>
you should probably remove the in-app share button (for Etherpad I think this could be done in settings.json)
<ocdtrekkie>
settings.json is a bit lighter in this app. But yeah, that's a good note.
<ocdtrekkie>
I'll get this GitHub'd and available probably sometime today, and put an email out to the dev list with the SPK for further testing.
<kentonv>
cool!
itscassa is now known as itscassa|away
<ocdtrekkie>
I hate that you can't code search forks.
<ocdtrekkie>
Code searching the master is usually workable, though for some reason EtherDraw lives in a fork.
<kentonv>
ocdtrekkie: you could re-upload the code to a newly-created repo that github won't see as a fork, I think
<ocdtrekkie>
kentonv: Yeah, though for Sandstorm ports, I'm fine with it being clear as a fork. It feels a little dirty to upload separately for that reason.
<ocdtrekkie>
But it's just weird when the upstream source is a fork.
<kentonv>
yes I agree your official repo for the port should be a fork, but if you just wanted something that you could search on the side...
<kentonv>
OTOH, you have the code locally so you can use grep
<ocdtrekkie>
Yeah, I just find the GitHub website browsing experience generally pleasant. Except when it fails hard.
<ocdtrekkie>
Which is admittedly often.
<kentonv>
I have noticed that github search doesn't always find all matches
<ocdtrekkie>
Yeah, I've experienced that.
<zarvox>
elasticsearch considered best effort
<zarvox>
(I suppose you could say the same of most search engines. I guess when the haystack to be searched is smaller, we have higher expectations of our search tools.)
<ocdtrekkie>
FYI guys: spk init does not create the appMarketingVersion field.
<ocdtrekkie>
Unless my spk tool is old or something.
<ocdtrekkie>
I just realized my EtherDraw port doesn't have it. (It'll be in the next release.)
<kentonv>
ocdtrekkie: hmm I thought we made it error out if it didn't have that field
<ocdtrekkie>
It didn't, so maybe my spk tool didn't update with the rest of Sandstorm?
<kentonv>
spk --version
<ocdtrekkie>
78
<kentonv>
wat
<kentonv>
ls -l $(which spk)
<kentonv>
or is your server still on build 0.78?
<kentonv>
we're at 0.81 now
<ocdtrekkie>
My server is still 78.
<ocdtrekkie>
Musta been changed after that.
<ocdtrekkie>
Will update.
<kentonv>
ah, do you not have auto-updates enabled?
<ocdtrekkie>
Not having it enabled has saved me once or twice.
<ocdtrekkie>
But sometimes I forget if I updated it, I thought I updated it like a week ago or something.
<kentonv>
we generally push every week
<kentonv>
and this week I think we pushed a couple times because of minor issues
<ocdtrekkie>
Cuz I thought the PR that changed appMarketingVersion was a while ago.
<ocdtrekkie>
So I thought it was weird it still wasn't creating it. Even in 78.
<kentonv>
I think we've only ever had one bug that caused auto-update to break itself, and that bug didn't apply to debian-derived distros.
<ocdtrekkie>
SPK is on GitHub, email is on dev list.
<ocdtrekkie>
kentonv: WTB liberated Ripples in Sandstorm so I can see who got my test grain from IRC vs. who got it from my dev list, and who shared to who and such and such.
<kentonv>
ocdtrekkie: we intend to do all that
<ocdtrekkie>
Trying to get in the habit of posting different public links to different sources so the data is meaningful.
<kentonv>
data is in the database; not sure when we'll be able to implement a visualization, though
<ocdtrekkie>
(Queue a spammer using a different share link to each email they spam, so that they can tell which email addresses were valid and which weren't.)
<kentonv>
well, they can do that already
<kentonv>
and in fact they *do* do that already
<ocdtrekkie>
Yeah, I know.
gopar has joined #sandstorm
<ocdtrekkie>
Should piwik be listed as "in progress", zarvox?
<zarvox>
Possibly. I'm currently shaving yaks to make it possible.