[7:00pm] nor3: marzia_google, dan_google: are you guys going to drop any info about the next release at this thing? [7:00pm] jeverling: wouldn't something like this would be a good task for the remote_api? [7:01pm] jeverling: 200_000 entities? [7:01pm] nurey: remote_api? [7:01pm] jeverling: any news on this? it was mentioned in the last chat session [7:01pm] marzia_google: well i'll officially call a start to the app engine chat time [7:01pm] • pr3d4t0r 's ears perk up. [7:01pm] marzia_google: we will be doing a release in the next few days, which will have more information about the remote_api [7:01pm] jeverling: great! [7:02pm] cying left the chat room. [7:02pm] nurey: marzia_google: is it a major release? [7:02pm] nurey: like 1.2? [7:02pm] sumit joined the chat room. [7:02pm] marzia_google: no, not a 1.2 [7:02pm] marzia_google: it's the 1.1.9 release [7:02pm] nor3: what will be in it? [7:03pm] oktopus: My app at http://simplifyconnections.appspot.com/ is throwing warnings [7:03pm] oktopus: about high cpu usage. How can I diagnose the problem? [7:03pm] Wooble left the chat room. (Remote closed the connection) [7:03pm] marzia_google: a revamped bulk upload tool [7:03pm] Wooble joined the chat room. [7:03pm] marzia_google: that goes along with the remote API [7:03pm] nurey: marzia_google: woohoo [7:03pm] marzia_google: and the usually issue fixes, etc that go along with a release [7:03pm] nurey: marzia_google: how about a function to delete all data for an app? [7:04pm] marzia_google: no [7:04pm] marzia_google: oktopus: a good way to diagnose high cpu usage is the python profiler [7:04pm] nor3: so "# Datastore import and export utility for large datasets" is the only roadmap item 1.1.9 will address? [7:04pm] oktopus: is the python profiler part of the sdk dashboard? [7:05pm] marzia_google: well, it will provide a better data import mechanism [7:05pm] marzia_google: profile is a built in python module [7:06pm] nurey: marzia_google: how will it be better? the data import mechanism? [7:06pm] oktopus: I see, so I "import profiler"? [7:06pm] oktopus: Then what? [7:07pm] jeffr: i am going to be creating an app, where the whole purpose is for people to send their friends referral URLs, e.g. example.com/<your-user-id> where each click on the referral url increments your number [7:07pm] marzia_google: http://code.google.com/appengine/kb/commontasks.html#profiling [7:07pm] jeffr: so theoretically, i will have a table with huge amounts of entities [7:07pm] oktopus: Got it. [7:07pm] oktopus: Thanks. [7:07pm] jeffr: is datastore fine for this application raw? [7:07pm] marzia_google: it will generally have a better way to manage the upload options and flow, better recovery [7:08pm] dan_google: jeffr: You'll want to store the counts in the datastore, yes. If they're read often, you can cache them in memcache. [7:08pm] nurey: is it ok to use datastore.Entity directly. i couldn't see if it's a private API [7:09pm] dan_google: nurey: datastore.Entity is safe to use. We document ext.db instead because we expect most people to want to use it. [7:09pm] oktopus: Are there folks who do code reviews of appspot apps, or is code review only done among associates? [7:09pm] nurey: oktopus: that's a good Q [7:10pm] jeffr: let's say for this app i want to make it so that duplicate IPs cannot be referred -- in this case it would be conceivable that i could have a million entities in the datastore [7:10pm] jeffr: if i store them there [7:10pm] dan_google: oktopus: There's no review of code. [7:10pm] nor3: how does google foresee quotas scaling over the years? like, 10 years from now, i wonder what reasonable "free" quotas will be [7:10pm] marzia_google: there is no 'official' application codereview, flow, though i'd encourage you to set one up if you are interested [7:11pm] marzia_google: and publicize it on the group [7:11pm] dstory: Will the inserts into the datastore scale? [7:11pm] dstory: Each time I add a device to my service, I need the ability to insert another entity every 15 seconds. [7:11pm] dstory: The entities are very small (160 bytes), and the inserts are indpendant, I don't need transactions. [7:11pm] dstory: 1000 devices needs to handle 66 inserts per second. [7:11pm] dstory: 5000 devices will need 330 per second. [7:11pm] dstory: 100000 devices 6600 per second. [7:11pm] dstory: Is GAE for me? [7:11pm] dstory: Will it scale with me? [7:11pm] marzia_google: nor3: I don't know if we have a '10 year' plan per se [7:11pm] marzia_google: i would say we always intend for it to be free to get started [7:12pm] nor3: yeah but you guys are the cutting edge of cloud platform development [7:12pm] nor3: you must have some sense of how things will scale [7:12pm] Cyndre: ooo is it talk to google time? [7:12pm] nor3: is there a moores law for cloud resources? [7:12pm] dan_google: dstory: App Engine will work for that, at least once we launch billing and you can pay to raise quotas. [7:12pm] nurey: nor3: i think prices will go down if that's what u're asking [7:12pm] marzia_google: we've planned for system resource scaling [7:13pm] marzia_google: but not what will be free, or what costs will be 10+ [7:13pm] dan_google: dstory: Especially if each web request only has to do one or two datastore writes, you should be fine. [7:13pm] marzia_google: which I don't think any one can really forsee [7:14pm] Cyndre: dan, would I be better to use 1700 http requests with 1 set of data in each, or 56 requests with 30 sets of data? [7:14pm] nor3: ok [7:14pm] pr3d4t0r: nor3: Let's talk later about that. I've been doing cloud infrastructure research on AWS, App Engine, Intel, etc. and how they scale; this was for both work and for InfoWeek. [7:14pm] nor3: can you give us some sense of what long running processes will look liek? [7:15pm] jeffr: how is the cron coming? [7:15pm] dan_google: Cyndre: Do you mean each set of data would go into its own entity group? [7:15pm] jeffr: i saw some rumors on the groups about cron [7:15pm] nor3: pr3d4t0r: ok. InfoWeek? are you writing an article for them? [7:15pm] pr3d4t0r: nor3: Later. Google time. [7:15pm] nor3: ok [7:15pm] nor3: yeah, how is the cron thing coming? [7:15pm] nurey: jeffr: afaik, cron is different from long running process [7:15pm] oktopus: I have an application that is sort of a social interaction facilitator. Is anyone here associated with that group at google? [7:15pm] Cyndre: dan, no the data is stored all in the same entity [7:16pm] marzia_google: all feature development mentioned is continuing apace, i don't really have specific details to release [7:16pm] dan_google: Cyndre: So a single request would update a single entity with a batch of data? [7:16pm] Cyndre: no, just inserts [7:16pm] Cyndre: each piece of data is a new insert [7:16pm] marzia_google: other than to say we are working very hard [7:16pm] marzia_google: on everything on the roadmap and more [7:17pm] dan_google: Cyndre: Right, I guess I'm asking, is your question about creating a bunch of entities in one request vs. creating one entity in each of a bunch of requests? [7:17pm] nurey: any idea on what the next language will be [7:17pm] pr3d4t0r: marzia_google: Can you please post the URL to the road map here so that I can log it for later use, especially when someone asks? [7:17pm] marzia_google: cross your fingers for FORTRAN [7:17pm] kixx: I'd like to store RDF triples in datastore. Any ideas on a good way to do this. It is certainly possible to use the high level API, but has anyone considered hacking at a lower level for RDF? [7:17pm] marzia_google: it's what i've asked for [7:17pm] Cyndre: dan: yes - is it better to use 1 http request and insert 30 diferent items, or 1 http request per item? [7:17pm] marzia_google: http://code.google.com/appengine/docs/roadmap.html [7:17pm] nurey: marzia_google: r u kidding me [7:18pm] nor3: yes she's kidding you [7:18pm] pr3d4t0r: marzia_google: RATFOR! [7:18pm] marzia_google: though i will say there is a strong LOLCODE faction on the team [7:18pm] nurey: marzia is female? [7:19pm] dstory: Cyndre: must keep below your quota... each request only gets a few seconds. [7:19pm] marzia_google: yes, I am female [7:19pm] marzia_google: let's clear that up once and for all [7:19pm] nurey: cool [7:19pm] dstory: no, she's geek [7:19pm] dstory: complement [7:19pm] Cyndre: well, the funny thing is that each request contains same ammount of data, but some take a long time and some dont..... not very consistant [7:19pm] nor3: what about larger datastore sizes [7:19pm] pr3d4t0r: marzia_google: http://is.gd/irj9 [7:20pm] dan_google: Cyndre: Multiple requests lets to parallelize the app logic a bit. If the datastore entities are all in their own entity groups (none of them share a group), the datastore writes are already parallelized to a certain extent, so that might not save you much. The best approach is to try something that's simple to implement and profile it. [7:20pm] marzia_google: [7:20pm] nor3: does "# Service for storing and serving large files" mean larger datastore entities, or static files? [7:20pm] jeffr: has anyone seen DownloadError: ApplicationError: 5 before [7:20pm] jeffr: i get it in my logs pretty often [7:20pm] marzia_google: that specifically refers to dynamic storage [7:20pm] marzia_google: not static files [7:21pm] nor3: ah ok [7:21pm] Cyndre: dan: ty [7:21pm] nor3: marzia_google: great [7:21pm] marzia_google: so to begin with things like uploading and storing large files to your applicaiton [7:21pm] marzia_google: etc [7:21pm] oktopus: I am looking for code that trims the gql query when a user zooms in on a map. Has anyone seen that? [7:21pm] marzia_google: rather uploading, storing, serving [7:21pm] nor3: what about the image library. is there any chance of that being extended in the foreseeable? [7:21pm] Cyndre: also, Im running out of datastore storage, upto .24 gigs..... how do I get more? [7:21pm] marzia_google: fill out the the quota increase requests! [7:22pm] marzia_google: oktopus: are you talking about geolocation? [7:22pm] marzia_google: or what specifically [7:22pm] marzia_google: or, rather, i'm just looking to plug this: [7:22pm] marzia_google: http://code.google.com/appengine/articles/geosearch.html [7:22pm] marzia_google: [7:22pm] marzia_google: which basically explains how to do geosearches at varying levels of granularity [7:23pm] dstory: nice, ty [7:23pm] oktopus: Maybe but I don't know what that means. I have lots of markers in the datastore and want to limit how many are rettrieved by what area the user is looking at. [7:23pm] Cyndre: where do I fill out the quota increase request at? (also if your intrested in what Im doing http://docsis-monitor.google.com/test2 things to come are maping of devices based on the address of the device) [7:23pm] marzia_google: markers, as in GeoPts? [7:23pm] Cyndre: er http://docsis-monitor.appspot.com/test2 lol [7:24pm] oktopus: markers are google map overlays [7:24pm] marzia_google: http://code.google.com/support/bin/request.py?contact_type=AppEngineContact [7:24pm] Cyndre: oktopus, do you have the markers gps coords stored? [7:24pm] Cyndre: you could select based on the gps of center of screen + or - 10 [7:24pm] Cyndre: type of thing [7:24pm] oktopus: I just have each marker stored with lots of info and latlng, too [7:25pm] marzia_google: yes, the issue with using just lat/lng is that you can only do inequality filtering on 1 property in the datastore [7:25pm] marzia_google: the only two good methods for doing geosearch is using geohash [7:26pm] marzia_google: or the method described in the article i linked above [7:26pm] oktopus: Great, so geosearch works with the datastore. [7:26pm] stumpy: Are there any plans to improve the speed that queries are handled? Serving a small 2KB static file seems to take a minimum 140ms of server side processing, though strangely 0ms is reported in the request log. [7:26pm] dwsBeta2_ is now known as dwsBeta2. [7:27pm] nor3: wait [7:27pm] nor3: does the database behind google maps resemble datastore? [7:27pm] taaz: my critical question of the day: what does _ah stand for? [7:27pm] nor3: haha, that's a good one [7:27pm] nurey: admin hole? [7:28pm] marzia_google: lol, more theories? [7:28pm] nor3: administrative hijinks [7:28pm] pr3d4t0r: marzia_google: App something. [7:28pm] kixx: Any suggestions for storing RDF triples? [7:28pm] marzia_google: not awesome as heck? [7:28pm] kixx: lots of them. [7:29pm] taaz: i think the leading theories are appengine handler or admin handler [7:29pm] dstory: I have my app up and running on GAE. [7:29pm] dstory: I'm trying to run a benchmark to create a cost model. [7:29pm] dstory: When I load test with only 500 devices (33 inserts per second) [7:29pm] dstory: I hit my quota limits too quickly. [7:29pm] dstory: Any suggestions? [7:30pm] Cyndre: dstory, what are you polling? [7:30pm] dstory: My devices are sending requests. bits of data I need to store. [7:30pm] dstory: not polling [7:30pm] Cyndre: what type of devices is what I meant [7:30pm] marzia_google: actually _ah doesn't really stand for anything anymore, it's just an old handler name that stuck around [7:31pm] dstory: cellular devices [7:31pm] marzia_google: but i think i'll go with awesome as heck from now on [7:31pm] nurey: i was wondering if a delayed put makes any sense...similar to insert delayed in mysql [7:31pm] Cyndre: dstory: cool, working on polling data from docsis cable modems [7:32pm] dstory: Cyndre: we are in a similar situation. [7:32pm] nor3: marzia_google, dan_google: is there anything coming out of the community you guys find particularly interesting? [7:32pm] nor3: like, interesting usage cases [7:32pm] marzia_google: hmmm, lots of things, i would say GAEBAR definitely caught our attention [7:32pm] marzia_google: oh like apps, you mean? [7:33pm] Cyndre: dstory: but Im only polling hourly, so I dont hit my quota yet, but when I get my first real customer I will be hitting it [7:33pm] oktopus: I noticed in the maps docs that map pinching, or somelthing like that, which makes maps resize on iphones, is now a default part of the application. When I was trying my app on an iphone about 3 weeks ago I didn't notice it. Is it very new? [7:33pm] marzia_google: i don't know if i'd classify the use cases as 'interesting' per se, but i've definitely noticed the number of mobile based apps that I hear of being developed [7:33pm] Adhir joined the chat room. [7:34pm] marzia_google: oktopus: i'm not that familiar with the particulars of maps + mobile honestly [7:34pm] nor3: i did mean apps [7:34pm] Cyndre: marzia: can you say thanks to google for giving us the oppurtunity to dream big and not worry about hardware/bandwidth? [7:34pm] marzia_google: there are all these great mashup type apps that i personally really like, such as hype tape [7:35pm] marzia_google: dream big or go home, i say [7:35pm] dstory: hehe, nice [7:35pm] marzia_google: there are so many great services being exposed by all these different platforms we see being brought together on App Engine, like the Best Buy remix [7:35pm] Cyndre: and the best part about gae is the only thing Im out is my time, until it takes off [7:36pm] oktopus: Yes, google has really made me happy of late. Thanks to you guys. [7:36pm] marzia_google: those are generally just great for ease and accessability of information [7:36pm] marzia_google: well, i'm glad people are putting AE to good use, it makes our job much more fun [7:36pm] Cyndre: did you expect buisness apps to be developed on it? [7:37pm] marzia_google: absolutely [7:37pm] oktopus: Are you in California, Marzia? [7:37pm] owen1 joined the chat room. [7:37pm] dstory: I guess I'll search for the info on how to request increases in your quotas. [7:37pm] oktopus: I am in Atlanta [7:37pm] Adhir: marzia:db.put(200 obj) is taking ~2 seconds,is it normal?is there any way to reduce this [7:37pm] marzia_google: As my flip-flops will attest to, I am in california [7:37pm] dstory: I need to create a cost model for my company. [7:37pm] marzia_google: put on 200 objects? [7:37pm] Adhir: yes [7:38pm] marzia_google: i would say, that's not surprising, and aside from changing the look of your data, no [7:38pm] marzia_google: puts, in general, are much much much much slower than reads [7:38pm] marzia_google: by design [7:39pm] marzia_google: *plug* http://sites.google.com/site/io/under-the-covers-of-the-google-app-engine-datastore [7:39pm] pr3d4t0r: marzia_google: Would making all those objects part of the same entity group help, though? [7:39pm] marzia_google: then you would definitely run in to contention issues [7:39pm] marzia_google: or, most probably [7:39pm] marzia_google: i would say it's unlikely you would want an entity group with 200 entities [7:40pm] Cyndre: marzia, is allowing infinite runtime maybe in the future (of course with cpu utilization quotas) [7:40pm] marzia_google: you mean long running processes? [7:41pm] Cyndre: yea, like say a bot [7:41pm] marzia_google: this is something we are working on supporting [7:41pm] Adhir: yes we are using same entity group to put all this object [7:41pm] dan_google: Yeah, you won't see significant any write performance improvement when updating multiple entities in the same group. In fact, you lose the opportunity for parallelization that way. A batch of puts of entities in different groups can be handled by multiple datastore servers simultaneously. [7:42pm] pr3d4t0r: dan_google: Awesome -- that's great info. [7:43pm] dstory: When will GAE be ready for prime time? [7:43pm] Adhir: i understand,if we use different groups,then how we can achieve run_in_transaction [7:43pm] stumpy: in the future might gae apps be distributed across multiple geographical locations to improve response times? kind of like amazons CloudFront [7:44pm] dan_google: Adhir: Your understanding is correct that you can only use run_in_transaction with entities in the same group. [7:44pm] marzia_google: i think it's worth pointing out that there are definitely business currently running on App Engine [7:44pm] oktopus: In one situation my app is just stopping prematurely without warning. Is that a symptom of exceeding my CPU quota? [7:44pm] marzia_google: though of course, you will be able to pay for more resource usage by the end of the quarter [7:44pm] marzia_google: which perhaps, is more what youa re driving at [7:44pm] dw: has any thought been given to providing an ajax-compatible messaging service? i.e. long-lived http requests [7:45pm] marzia_google: if you exceed CPU quota, your app will start returning over quota 403s [7:45pm] marzia_google: i guess i'd need to know the nature of the 'stopping prematurely' [7:45pm] marzia_google: to say anyting definitively [7:45pm] dan_google: Adhir: The reason why we recommend keeping groups small is that when the app updates *any* entity in the group, no other user (simultaneous request) can update *any* entity in the group. [7:45pm] oktopus: I am getting warnings in my request Logs, Are those 403's? [7:46pm] marzia_google: no, warnings are just... warnings [7:46pm] jamesv joined the chat room. [7:46pm] marzia_google: you can look at the quota details page to see if you are having quota issues [7:46pm] dstory: Can I pay for increased quotas now? [7:46pm] oktopus: It stops at the point it should be presenting a map. But no map and no feedback. [7:46pm] dw: currently i have a nasty hack of a component which runs on just a single machine, and will cost significant time and money to make scaleable and redundant. it only exists to serve messages asynchronously [7:46pm] Adhir: dan:is there any way to put all this 200 objects in separate thread(parallel) [7:46pm] Adhir: with in same transaction [7:46pm] oktopus: Where would I see the numbers 403? [7:47pm] Cyndre: do you guys have a pricing plan yet so we can see what we want to accomplish will cost based on our test runs? [7:47pm] oktopus: I am seeing little yellow triangles in the logs [7:47pm] marzia_google: you can look in your request logs, or on the dashboard at quota denials [7:47pm] marzia_google: sure the pricing plan on the blog: [7:47pm] marzia_google: http://googleappengine.blogspot.com/2008/05/announcing-open-signups-expected.html [7:48pm] Cyndre: and those prices apply to the datastore quotas as well? [7:48pm] dan_google: Adhir: I'm not sure I understand the question. A single request cannot create multiple threads. Depending on your client, you can achieve some parallelization with multiple simultaneous requests. There's no way for multiple requests to contribute to a single transaction. [7:49pm] oktopus: I see: 68.223.18.10 - - [04/Feb/2009:15:22:27 -0800] "GET /?place=contraUSA HTTP/1.1" 500 1022 - - [7:49pm] marzia_google: yes, that includes storage prices [7:49pm] dan_google: Adhir: Could you describe your use case a bit? What aspect of your app implies that 200+ entities need to be potentially modifiable with each other in a single transaction? It sounds like you need to shard your data a bit. [7:49pm] dstory: I need to run a small load test to create a cost model. I'm currently hitting my quotas too soon. [7:50pm] Cyndre: Thanks marzia [7:50pm] oktopus: This too: 500 3102ms 6172ms-cpu 1kb [7:50pm] nurey: is there an accessor for the key name of a entity [7:50pm] marzia_google: all quota increases are handled by filling out the application form [7:50pm] dw: nurey: key().name() [7:51pm] oktopus: So it looks like my error is a 500, not a 403, right? [7:51pm] marzia_google: yep, 500 the dreaded server error [7:51pm] dw: anyone for the long-lived/blocking http question? [7:51pm] marzia_google: which means [7:51pm] marzia_google: well, anything [7:51pm] marzia_google: something is going wrong, and nothing is being returned [7:51pm] Cyndre: okay, /test 500 410ms 8512ms-cpu and /test 500 18ms 198ms-cpu both of those are results from almost identical http requests (same ammount of data, but from 30 different devices) why such a difference? [7:52pm] oktopus: marzio: are you talking to me? [7:52pm] oktopus: marzia, I mean. [7:52pm] marzia_google: oktopus: yes, [7:52pm] marzia_google: the 500s could mean many things [7:52pm] marzia_google: so, again, it's hard to diagnose just from one request line [7:53pm] Adhir: dan:we are building a rule engine over GAE. [7:53pm] Adhir: we have two phases in our app,dev and prod [7:53pm] Cyndre: marzia, I have 6 of those out 20 requests [7:53pm] marzia_google: cyndre: one thing that that could be is start up cost for the interpreter, [7:54pm] Cyndre: marzia: but why is it so inconsistant? my data is submitted in chunks of 30, and 7/10 times no issues, but 3/10 times putting the data fails [7:54pm] oktopus: I see. So it may be causing the slowness, but is not the result, right? [7:55pm] marzia_google: cyndre: because we launch an instance, it serves some requests, and then we may launch a new one [7:55pm] nurey: are there plans to do something more sophisticated than SearchableModel [7:55pm] marzia_google: of the python interpreter [7:55pm] marzia_google: so there is no garunteed amount of time [7:55pm] dstory: Cyndre: put try: except db.Error, e around your put(), and log the error. [7:55pm] Cyndre: it takes 8200 ms of cpu time to launch an interperter? [7:56pm] marzia_google: nurey: SearchableModel, probably (?) not, search yes [7:56pm] marzia_google: we would like to have more sophisticated search support [7:56pm] nurey: full text search basically [7:56pm] marzia_google: it could, certainly especially if you were using something like Django [7:56pm] marzia_google: it depends on what modules you are loading [7:57pm] marzia_google: full text search is definitely something that is being explored [7:57pm] oktopus: Can someone tell me how to do this better in django. [7:57pm] oktopus: {% for t in color %} [7:57pm] oktopus: {% for s in category %} [7:57pm] oktopus: {% ifequal forloop.counter forloop.parentloop.counter %} <td align="center" bgcolor="#{{ t }}"> {{s}} </td> {% endifequal %} [7:57pm] oktopus: {% endfor %} [7:57pm] oktopus: {% endfor %} [7:57pm] moraes: oktopus. don't. [7:57pm] moraes: [7:57pm] dw: nurey: it is simple to do FTS on an external host, by incrementally exporting your datastore [7:57pm] Cyndre: http://pastebin.com/m31f74a1d is what Im including [7:58pm] nurey: yes but this is google... google is all about search [7:58pm] dw: nurey: have you seen google base api? [7:59pm] pr3d4t0r: marzia_google: I think there's only a minute left. [7:59pm] nurey: sorta [7:59pm] dw: perhaps you could abuse it somehow [7:59pm] • dw is still wondering about long-lived/blocking http requests, or something competitive to amazon simple queueing service [7:59pm] oktopus: There are the same number of items in category and in color, btw. [8:00pm] pr3d4t0r: marzia_google: As a side-thing: I never heard back from Mike R. Should nudge him directly, or just wait for a couple more days? Email you separately? [8:00pm] marzia_google: cyndre: my guess is these costs are associated with launching a new python interperter, again, it's a hard problem to diagnose via chat, but i would say it's best not to dwell on any specific CPU reading, instead look at the handler's performance as a whole [8:01pm] marzia_google: pr4d4t0r: nudging never hurt [8:01pm] pr3d4t0r: marzia_google: OKi. [8:01pm] oktopus: g'nite and much thanks to all. [8:01pm] Cyndre: marzia: its not about performance, its about losing 30% of submitted data..... [8:01pm] marzia_google: well that's unrelated to CPU [8:01pm] nurey: any hint on this error: can't update the same entity twice in a transaction or operation [8:02pm] marzia_google: it sounds more like then you are trying to process too much data in one request [8:02pm] Cyndre: The API call datastore_v3.Put() required more quota than is available. [8:02pm] Adhir: RPC call time is getting varied from morning to evening?is this because of the traffic(any bottle neck) [8:02pm] marzia_google: right, so, you are exhausting quota [8:02pm] marzia_google: the only thing you can do is apply for more [8:02pm] Cyndre: but its the exact same amount of data as the 70% that succeed [8:03pm] taaz: if you google folks have any spare time it would be nice if you went through the cookbook and left comments about the code left there. it's often hard to tell if some of those snipets are efficient and scalable or just quick hacks that shouldn't really be used. [8:03pm] Cyndre: and this is a new issue that just started up [8:03pm] owen1: can anyone recommend a blog engine? [8:03pm] dw: it would also be nice if the cookbook properly escaped html, assuming that hasn't been fixed yet. [8:03pm] moraes: ow btw the cookbook needs to be fixed. [8:03pm] dw: owen1: bloog, http://lmgtfy.com/?q=bloog [8:03pm] moraes: yeah, see dw ^ [8:04pm] JamesMowery is now known as JamesMowery|away. [8:04pm] marzia_google: right, but it's the number of puts() not the amount of data, apply for a quota increase [8:04pm] owen1: dw: thanks dw, i'll check it out [8:04pm] JamesMowery|away is now known as JamesMowery. [8:04pm] dw: owen1: wasn't aiming to be sarcastic, went to the wrong host [8:04pm] owen1: dw: that was cool! [8:04pm] marzia_google: yes, the cookbook could use some bug fixing [8:04pm] marzia_google: i can not deny that [8:05pm] cresloyd: +1 on the cookbook fix - I just entered a new entry, messed it up badly, and couldn't fix [8:05pm] marzia_google: you can edit your entries [8:05pm] cresloyd: how? [8:05pm] marzia_google: did this not for some reason work/ [8:05pm] dw: so, before i try and get 2 hours sleep before my 9am appointment, any comment on long lived http, or client synchronization or efficient messages of any kind that won't complete exhaust my request quota? [8:05pm] marzia_google: it's in the upper right hand side [8:06pm] moraes: with unescaped html, it gets messed sometimes [8:06pm] pr3d4t0r: marzia_google, dan_google: Thanks for stopping by, guys. [8:06pm] marzia_google: we would like to support long running processes, but i have no impending announcements [8:06pm] dw: the cookbook does not like "<", whereas it handled ">" just fine (IIRC) [8:06pm] cresloyd: thanks; I'll go look for that edit button [8:06pm] dan_google: pr3d4t0r: Always a pleasure, thank you! [8:07pm] marzia_google: yeah thanks for stopping by
--~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Google App Engine" group. To post to this group, send email to google-appengine@googlegroups.com To unsubscribe from this group, send email to google-appengine+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en -~----------~----~----~----~------~----~------~--~---
