Hi all, Enjoy today App Engine IRC Office Hours transcript below, next session should happen Wednesday, Aug 31, 2011 7PM PST on irc://irc.freenode.net/#appengine.
-------------------------------------------------------------------------------------------- 2011-07-20 5:59 PM CEST App Engine IRC Office Hours starts -------------------------------------------------------------------------------------------- [5:59 PM] <proppy> App Engine Office Hours are starting now [5:59 PM] <robertk> are you the only googler proppy? [5:59 PM] <proppy> most of the eng team is away today, so I will mostly record questions and forward to them [5:59 PM] <johns1> Will my site have extended validation enabled with SSL? I see that Google Plus is "Verified by Google". Will my appengine site be as well? [5:59 PM] <proppy> if I can't answer to them myself [5:59 PM] <robertk> where are your comrades? [6:00 PM] <johnlockwood> I've noticed a weird issue with the dev server and Chrome Beta Profiles [6:00 PM] <proppy> robertk: away paying a visit to another office I guess [6:01 PM] <case__> robertk, i think they're getting some motorola droids :) [6:01 PM] <robertk> ha ;) [6:01 PM] <johnlockwood> it I access the dev site from an alternate profile, it will have long pauses while loading resources [6:01 PM] <proppy> johns1: are you referring to appspot.com ssl cerfificate ? [6:01 PM] <johns1> well, there will be a custom domain name [6:01 PM] <johns1> (naturally) [6:02 PM] <johns1> but yes, i access the management panel via an appspot.com domain name [6:02 PM] ⇐ tid-wave (51c49c5d@gateway/web/freenode/ip.81.196.156.93) quit: Ping timeout: 252 seconds [6:02 PM] <proppy> johnlockwood: I think this bug is similar to the one already filled on the public issue tracker for incognito [6:02 PM] <proppy> devappserver hanging [6:02 PM] → aa12a_ ([email protected]) joined [6:02 PM] <proppy> http://code.google.com/p/googleappengine/issues/detail?id=4716 [6:03 PM] <proppy> yeah public issue tracker is back ! [6:03 PM] <proppy> \o/ [6:03 PM] → ikai_google (voiced) and guy__ joined ⇐ aa12a_ quit [6:04 PM] <proppy> johns1: ssl on custom domains is not yet available [6:04 PM] <proppy> but it is on the roadmap [6:04 PM] <proppy> http://code.google.com/appengine/docs/roadmap.html [6:04 PM] <johns1> okay, do you have an unofficial ballpark figure on the TTA? [6:04 PM] <sinclair> proppy: do you know if there are any plans to allow for naked domains to be mapped? [6:05 PM] <sinclair> proppy: i.e. mydomain.com not, www.mydomain.com [6:05 PM] <robertk> johnlockwood: have you tried disabling that pre-fetch (or whatever it is called) in chrome? [6:06 PM] <johnlockwood> robertk: no, did not try that [6:06 PM] → wesley_google (43a0c178@gateway/web/freenode/ip.67.160.193.120) joined [6:06 PM] <robertk> yeah maybe that would help [6:06 PM] <johnlockwood> robertk: I actually went back to using the stable chrome because of another issue [6:06 PM] <robertk> i usually just flip to the terminal and do a ctrl-c [6:06 PM] <robertk> ah [6:06 PM] <proppy> sinclair: dunno, it doesn't seems on the roadmap [6:07 PM] <wesley_google> oh, it's you two... :-) [6:07 PM] <johnlockwood> hi wesley_google [6:07 PM] <proppy> sinclair: I know that people are using custom redirection to handle those [6:07 PM] → jgn (~jgn@unaffiliated/jgn) joined [6:07 PM] <robertk> johnlockwood: yeah i'm on stable chrome today too. what was your issue? [6:07 PM] <proppy> or tiny reverse proxy [6:07 PM] <robertk> hey wesley_google ;) [6:07 PM] <robertk> johnlockwood: just curios if i should not bother flipping back to dev channel for a bit ;) [6:08 PM] <ikai_google> given that every single ETA we've given has been horribly, horribly wrong [6:08 PM] <ikai_google> why do you guys still ask for them? [6:08 PM] <robertk> *curious [6:08 PM] <ikai_google> (re: think about all the stuff we've announced at IO that we said was coming soon .. and took way longer) [6:08 PM] <johns1> well, just do make sure we can be secure soon [6:08 PM] <johns1> well, not to be sure [6:08 PM] <johns1> but to feel reassured [6:08 PM] <ikai_google> well, if you asked me 8 months ago [6:08 PM] <johns1> more psychological than anything [6:08 PM] <ikai_google> I would have said, 2 months [6:08 PM] → derfj ([email protected]) joined [6:08 PM] <johns1> haha [6:09 PM] <ikai_google> and I would have thought that was right [6:09 PM] <johns1> so 2 months? [6:09 PM] <johns1> ;) [6:09 PM] <ikai_google> sure [6:09 PM] <johns1> yaqy [6:09 PM] → max-oizo ([email protected]) joined [6:09 PM] <ikai_google> have you heard the phrase [6:09 PM] <ikai_google> the fastest algorithms are the ones that don't run? [6:09 PM] <mbw> howdy googlers [6:09 PM] <robertk> so for all their match skillz, googlers suck at estimating stuff too ;) [6:09 PM] <johns1> nope... i like it [6:09 PM] ⇐ rwilhelm23 ([email protected]) quit: Remote host closed the connection [6:10 PM] <johnlockwood> robertk: it is not totally compatible with google maps [6:10 PM] <ikai_google> if your script takes 0ms to run, you probably didn't do anything, and it probably has a correctness of 0 [6:10 PM] → rwilhelm23 ([email protected]) joined [6:10 PM] <robertk> johnlockwood: ah, yeah i saw your comment on that. k, will skip the switch too. ty :) [6:11 PM] <johns1> I'm designing a system which utilizes channels to avoid periodic GET update requests. Assuming correct implementation, do you appengine masters foresee any potential issues (i.e. cost/performance) which may arise from using channels for these day-to-day, standardized updates? [6:11 PM] <ikai_google> johns1: not really, except # of channels [6:12 PM] <ikai_google> johns1: make sure you take that into account. if you are doing broadcast type stuff, I know we mentioned we're looking at it, but for instance 1:1mm user broadcast isn't something you get enough quota to do [6:12 PM] <ikai_google> johns1: and you might be better off looking at a 3rd party solution like beaconpush.com [6:13 PM] <wesley_google> johns1> also, channels expire like within a few hours, so they're not for long-term [6:13 PM] → starship (6161c39e@gateway/web/freenode/ip.97.97.195.158) joined [6:13 PM] <rwilhelm23> wesley_google: Do channels expire if not used, or do they naturally just die after X time? [6:13 PM] → magicbronson joined ⇐ max-oizo quit [6:14 PM] <johns1> ah, so they're generally meant for goal-oriented interaction? [6:15 PM] <magicbronson> Hey guys, any App Engine devs here for the chat? [6:15 PM] <starship> Ok I have an app that it's data usage has exploded twice from .5 GB to 1GB over night normally it only increases .01GB a week or so [6:15 PM] <wesley_google> rwilhelm23> they expire naturally but you need to manually scrub the channel IDs if you're saving them. [6:16 PM] ⇐ ristoh ([email protected]) quit: Ping timeout: 252 seconds [6:16 PM] <wesley_google> see the 2nd para in this section of the docs to tell you about the 2-hour timeout: http://code.google.com/appengine/docs/python/channel/overview.html#Tokens_and_Security [6:16 PM] <ikai_google> starship: could be anything honestly. task queues, session data. I'm not so worried about it since its relatively small amounts of data [6:16 PM] <robertk> starship: did you have a corresponding increase in traffic? [6:17 PM] <starship> This is the second time it has happened to solve it I have wiped one type of entity. I took a look at some of the data and it did not seem to be abnormal or even doubling of the number of enites [6:17 PM] <proppy> magicbronson: only developer relatios so far, dev are currently travelling [6:17 PM] <proppy> but we can still help you :) [6:17 PM] → jsurfer ([email protected]) joined [6:17 PM] <magicbronson> proppy: thanks! would now be an okay time to ask about some issues in the tracker? [6:18 PM] <ikai_google> usually the tracker is the best place to ask about issues in the tracker, but we're at least open to hearing about them [6:18 PM] <starship> ikai_google: well it started getting quota error as it was set for just 1GB [6:18 PM] <proppy> magicbronson: yes sure [6:19 PM] <johns1> ikai_google: 1:1mm is a no go. got it. how about 1mm X 1:10 broadcast? [6:19 PM] <starship> robertk: there was a increase of traffic once 500 errors started other wise it was just a normal [6:19 PM] <johns1> broadcasts* [6:20 PM] <robertk> starship: sounds like maybe you've got some tasks? [6:20 PM] → dbslusser ([email protected]) joined [6:20 PM] <ikai_google> johns1: you can't get enough channel quota for that [6:20 PM] <ikai_google> johns1: and you need to manage broadcast lists in memory [6:20 PM] <johns1> oooooooooooo [6:21 PM] <ikai_google> johns1: check out that other service and let me know what you think [6:21 PM] <johns1> can't datastore the tokens? [6:21 PM] <ikai_google> oh you can certainly do that [6:21 PM] <ikai_google> you just have to do it yourself [6:21 PM] <wesley_google> johns1> yes [6:21 PM] <ikai_google> but the problem is getting 10 million channels [6:21 PM] <starship> robertk: it is not tasks data storage and we do not use much [6:21 PM] <starship> use the tasks* [6:21 PM] <wesley_google> that's what i meant by after expiration, you need to scrub out the old channel IDs yourself... [6:21 PM] <vmil86> I'm experiencing a similar issue as starship is, the HRD and TDS quotas no logger count the data deleted or replaced, thus continuously grow (150MB per day for HRD for me) [6:21 PM] ⇐ dnjaramba ([email protected]) quit: Ping timeout: 259 seconds [6:21 PM] <johns1> would an hourly CRON be a smart way to scrub? [6:22 PM] <starship> That would make sense as the data gets updated pretty often [6:22 PM] <magicbronson> thanks, proppy, ikai_google... haven't seen any recent updates in the tracker on a few particularly concerning issues: #4560, #4878, #4277, #2258, for starters [6:22 PM] ⇐ jsurfer ([email protected]) quit: Quit: jsurfer [6:22 PM] <magicbronson> want the full links? [6:22 PM] <starship> once every 30 min [6:23 PM] → TomS24 and jmotta joined [6:24 PM] <ikai_google> magicbronson: we prioritize bugs against all other bugs [6:24 PM] <wesley_google> johns1> that sounds good [6:24 PM] <proppy> magicbronson: there is currently issue with the issue tracker [6:24 PM] <proppy> magicbronson: so I can't check them out right now [6:24 PM] <ikai_google> magicbronson: so I'll take a look at these 4, but if we've marked them as acknowledge, i means they've been triaged [6:24 PM] <proppy> I will make sure to check them out when it is fixed [6:25 PM] <johns1> Can channels be used within native, mobile apps? [6:25 PM] <ikai_google> johns1: nope [6:25 PM] <ikai_google> johns1: it's one of the issues in the issues tracker [6:25 PM] <johns1> ah, i'll find it [6:26 PM] <proppy> johns1: the only implementation I'm aware of is [6:26 PM] <proppy> https://bitbucket.org/lohre/gae_channel [6:26 PM] <proppy> and it is python [6:26 PM] <magicbronson> proppy, ikai_google: thanks. at least one of those was marked as high priority but there's been no update on the issue since May, so I just thought I'd try asking here. [6:26 PM] <starship> Now I can open an issue but I am not sure what all to say [6:26 PM] ⇐ soyapi ([email protected]) quit: Ping timeout: 250 seconds [6:26 PM] <ikai_google> johns1: but for those situations you probably want to implement push notifications [6:26 PM] <johns1> from the server? [6:26 PM] <proppy> johns1: it could be used as a base for the protocol documentation for a Java or Obj-C imp [6:26 PM] → dsarmientos__ joined ↔ magicbronson nipped out [6:27 PM] <TomS24> What is the recommended way to get a blob from blobstore to Google Docs? [6:28 PM] → SegFaultAX joined ⇐ vbabiy quit [6:28 PM] <johns1> It would be nice to extend my client side javascript app to other advices. I realize that this isn't 'mobile' or 'V8' channel, but given your google expertise and vision, is it starting to look like V8 may make an appearance on mobile devices? [6:28 PM] <proppy> use the Docs API ? [6:28 PM] <proppy> TomS24: ^ [6:28 PM] <vmil86> starship, there is a similar, but not the exact issue already - http://code.google.com/p/googleappengine/issues/detail?id=2568 [6:28 PM] <robertk> TomS24: you probably want to look at resumable upload, but otherwise read the data and send it? [6:28 PM] ⇐ lmatteis (53675e32@gateway/web/freenode/ip.83.103.94.50) quit: Quit: Page closed [6:28 PM] <aa_> TomS24: use the ResumableUpload functionality in docs [6:28 PM] <TomS24> Docs API gets errors on large blobs [6:28 PM] <TomS24> the insert gets the error [6:29 PM] → dnjaramba ([email protected]) joined [6:29 PM] <aa_> TomS24: how large is your blob? [6:29 PM] <TomS24> 750K [6:29 PM] <aa_> that's not too large at all [6:29 PM] <proppy> starship: did you add any new indexes ? [6:29 PM] <TomS24> no [6:29 PM] <TomS24> It's a HTML doc in blobstore [6:30 PM] → hungryghost (d8ef2d82@gateway/web/freenode/ip.216.239.45.130) joined [6:30 PM] <starship> proppy: no new indexes it was running fine for months but this is the second time it has happened [6:30 PM] → soyapi joined ⇐ dsarmientos quit [6:30 PM] <aa_> TomS24: if you are having errors, please raise them here for docs: http://code.google.com/apis/documents/forum.html [6:31 PM] <ikai_google> johns1: aren't they already using v8 in tablets? [6:31 PM] <johns1> oOo [6:31 PM] <ikai_google> johns1: anyway, what javascript engine they're using is irrelevant [6:31 PM] <itsPhilip> Whats the reason behind Google not officially supporting edge caching? [6:31 PM] <ikai_google> itsPhilip: it's complicated. [6:31 PM] <starship> ikai_google: is there a way to get an accounting of db usage? [6:32 PM] <proppy> starship: datastore stats [6:32 PM] <proppy> starship: but it does not show indexes [6:32 PM] <starship> also the database statics had not been updated snice the explotion [6:32 PM] ← magicbronson ([email protected]) left [6:32 PM] <johns1> tring to figure out how much this will cost per month based on usage is a bit of a glamorous mystery to me. Are there any templates/use cases + costs which I can match to in order to find a decent ballpark figure? [6:32 PM] <proppy> yes, it is only daily IIRC [6:32 PM] <proppy> starship: I would suggest to wait again for 1 day [6:32 PM] <starship> it is at 1 day 4 hours [6:33 PM] <itsPhilip> ikai_google: Great, that sounds like a couple that is in trouble :-) [6:33 PM] <ikai_google> johns1: http://googleappengine.blogspot.com/2010/06/how-app-engine-served-humble-indie.html [6:33 PM] <johnlockwood> TomS24: you could use Apps Script to pull from the blobstore [6:33 PM] <proppy> and to open a production issue if you still dont understand why it growed [6:33 PM] <ikai_google> johns1: with the new pricing he would have probably paid $90 [6:33 PM] <starship> normally updated much faster [6:33 PM] <ikai_google> johns1: just a guess. [6:33 PM] <johns1> :) [6:34 PM] <robertk> yeah, estimating data use is very hard based on the info we've got. [6:34 PM] <robertk> i compute stuff and get really close, then i've always got a fuzz-factor multiplier i have to use [6:34 PM] <vmil86> proppy, my issue of the non-stop grow is already 5 days old [6:34 PM] → sirdancealot joined ⇐ pr0ton quit [6:35 PM] <robertk> for those that haven't: http://code.google.com/p/googleappengine/issues/detail?id=2740 [6:35 PM] <proppy> vmil86: fill a production issue once the tracker is back [6:35 PM] <vmil86> proppy, I already mentioned it here - http://code.google.com/p/googleappengine/issues/detail?id=2568 [6:35 PM] <starship> proppy: it is up for me [6:35 PM] <proppy> starship: 1 out of 2 requests hangs for me [6:36 PM] → jfmontesdeoca_GO (480ee551@gateway/web/freenode/ip.72.14.229.81) joined [6:37 PM] <ikai_google> it's slow for me, but I've been blaming that on AT&T [6:37 PM] <proppy> ikai_google: nop there is a code.google.com outage going on [6:37 PM] <starship> also is there a way to get a graph of storage usage like you can of cpu? [6:38 PM] <proppy> vmil86: acknowledged the issue [6:38 PM] ↔ derfj_ popped in [6:39 PM] <proppy> starship: sound like a candidate for a feature request [6:39 PM] <proppy> starship: we only have pie graph for now [6:39 PM] ↔ derfj_ popped in [6:39 PM] <starship> well it would make seeing spikes in usage easer [6:40 PM] <starship> and more detailed compared to gradual increase [6:40 PM] → derfj_ ([email protected]) joined [6:41 PM] <proppy> starship: theoritacly you should be able to graph it yourself [6:41 PM] <proppy> using the upcoming monitoring api [6:41 PM] ⇐ derfj ([email protected]) quit: Ping timeout: 250 seconds [6:41 PM] <proppy> that will expose the dashboard datas [6:41 PM] <starship> nice [6:43 PM] → littleq ([email protected]) joined [6:43 PM] <proppy> starship: http://www.youtube.com/watch?v=rgQm1KEIIuc at 45:00 [6:43 PM] littleq → LittleQNCCU [6:43 PM] <proppy> this was the I/O talk for the announcement this year [6:43 PM] <robertk> ikai_google: any eta on the monitoring api :D [6:44 PM] <mbw> robertk: did you look at the docs for it? #fail [6:44 PM] <itsPhilip> I predict the answer will contain "soon" [6:44 PM] <LittleQNCCU> ikai_google: hi~ [6:44 PM] <ikai_google> robertk: for you and mbw, today [6:44 PM] <robertk> ha ha [6:44 PM] <mbw> it doesnt do much of anything and the precision is pretty much worthless, int [6:44 PM] <robertk> oh is it in tt? [6:44 PM] * starship looks for docs [6:44 PM] <ikai_google> the docs aren't public [6:44 PM] ⇐ derfj_ ([email protected]) quit: Ping timeout: 258 seconds [6:45 PM] <ikai_google> robert, I'll send you what I have [6:45 PM] ↔ derfj popped in [6:45 PM] <mbw> i really hope its just not done yet, because in its current state I am not sure it's very useful [6:45 PM] <robertk> ikai_google: awesome! [6:45 PM] <ikai_google> mbw: it's definitely not done yet [6:46 PM] ⇐ jfmontesdeoca_GO (480ee551@gateway/web/freenode/ip.72.14.229.81) quit: Quit: Page closed [6:46 PM] <proppy> mbw: do you have a contact to provide feedback to already ? [6:46 PM] <ikai_google> mbw: but we need your feedback as to why it's not useful [6:46 PM] <mbw> ya, ill get to that [6:46 PM] <wesley_google> see you guys! i need to leave for the airport now. [6:46 PM] <mbw> I just got access last night [6:47 PM] <johnlockwood> wesley_google: you get to have all the fun [6:47 PM] <robertk> later wesley_google [6:47 PM] <LittleQNCCU> where is the most effective way to give feedback to app engine team? [6:48 PM] → pr0ton ([email protected]) joined [6:48 PM] <robertk> you can try here.. depending on what it is [6:49 PM] <robertk> or the groups... or the issue tracker [6:49 PM] ⇐ wesley_google (43a0c178@gateway/web/freenode/ip.67.160.193.120) quit: Quit: Page closed [6:50 PM] <starship> here you go [6:50 PM] <starship> http://code.google.com/p/googleappengine/issues/detail?id=5609 [6:53 PM] <LittleQNCCU> thanks [6:56 PM] ⇐ hungryghost (d8ef2d82@gateway/web/freenode/ip.216.239.45.130) quit: Quit: Page closed [6:56 PM] <starship> ok anythign else on dek [6:59 PM] ⇐ guy__ ([email protected]) quit: Quit: Ex-Chat [6:59 PM] <proppy> starship: ack'ed you issue [6:59 PM] <proppy> will take a look tomorrow [6:59 PM] <LittleQNCCU> hey, guys, sorry for bothering, we are ready to form a gae study group in Taiwan, if you are interesting in it, you can follow up www.taipei-gtug.com, and if you have something to share, please contace me [email protected] [6:59 PM] <starship> ok thank you [6:59 PM] → zemanel ([email protected]) joined [7:00 PM] <LittleQNCCU> we will have live streaming. [7:00 PM] <robertk> proppy: any news back from the eng team on my moderator question? [7:00 PM] <ikai_google> LittleQNCCU: =) [7:00 PM] <ikai_google> LittleQNCCU: good job. [7:00 PM] ← johns1 ([email protected]) left [7:00 PM] <robertk> LittleQNCCU: you should post that announcement to the groups too [7:01 PM] <LittleQNCCU> robertk: i will [7:01 PM] <robertk> cool [7:01 PM] <proppy> robertk: got a few feedback from ikai on it [7:01 PM] → johnlockwood_ ([email protected]) joined [7:01 PM] <LittleQNCCU> ikai_google: maybe you can come and give us some talks, if you have free time [7:01 PM] <proppy> robertk: but I'm working on getting also feedback from other member of the team [7:02 PM] <proppy> to compute a comprehensive answer [7:02 PM] ⇐ dave_google (~smithd@nat/google/x-xgqkphjqizfgwqqp) quit: Quit: Ex-Chat [7:02 PM] <ikai_google> LittleQNCCU: you know I'd love to. not planning on traveling to Taipei anytime soon. I should also really visit my aunt in Kaosiung [7:02 PM] <proppy> does someone has the full log of the IRC office hours ? [7:02 PM] <starship> LittleQNCCU I do not get a webpage for http://www.taipei-gtug.com/ [7:03 PM] <LittleQNCCU> sorry, it's .org [7:03 PM] <ikai_google> I do [7:03 PM] <ikai_google> anyway, thanks for coming to IRC office hours [7:03 PM] <ikai_google> when I feel better I'll start running hangouts on Google+ again: http://plus.ikailan.com [7:03 PM] <LittleQNCCU> ikai_google:Kaosiung has GTUG, too. If it is possible, i will follow up :) [7:03 PM] <robertk> cool [7:04 PM] <robertk> thanks *_google ;) [7:04 PM] <ikai_google> bye guys [7:04 PM] ← yarrumretep left ⇐ ikai_google quit [7:04 PM] <LittleQNCCU> thank *_google [7:04 PM] <robertk> (and proppy) ;) [7:04 PM] <johnlockwood> thanks -------------------------------------------------------------------------------------------- 2011-07-20 7:04 PM CEST App Engine IRC Office Hours closes -------------------------------------------------------------------------------------------- -- Johan Euphrosine (proppy) Developer Programs Engineer Google Developer Relations -- You received this message because you are subscribed to the Google Groups "Google App Engine" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.
