| 00:03.08 | *** join/#storm seiflotfy2 (n=seif@ce1.hrz.tu-darmstadt.de) |
| 00:43.46 | *** join/#storm drudi (n=drudi@187.46.58.203) |
| 00:45.55 | *** join/#storm artista-frustrad (n=artista_@201-24-233-220.ctame704.dsl.brasiltelecom.net.br) |
| 01:00.13 | *** join/#storm seiflotfy1 (n=seif@ce3.net.hrz.tu-darmstadt.de) |
| 02:11.09 | thumper | anyone awake? |
| 02:11.15 | thumper | I have a frustrating storm issue |
| 02:11.26 | thumper | jkakar, jamesh? |
| 02:11.41 | jamesh | hi thumper |
| 02:12.10 | thumper | jamesh: I'm wanting to union two result sets together in code that doesn't know the "guts" |
| 02:12.15 | thumper | I want one query |
| 02:12.22 | thumper | so among other things I can do a count |
| 02:12.35 | thumper | a.union(b) doesn't work |
| 02:12.40 | thumper | neither does Untion(a,b) |
| 02:12.57 | jamesh | when you say it doesn't work, what happens exactly? |
| 02:13.15 | thumper | Union(a,b) does not return a result set |
| 02:13.21 | thumper | but something to use inside a select |
| 02:13.31 | jamesh | well, Union(whatever) will return a Union instance |
| 02:13.45 | jamesh | what about a.union(b)? |
| 02:14.00 | thumper | I don't think that exists |
| 02:14.06 | thumper | I'm pretty sure I tried that |
| 02:14.22 | jamesh | what are a and b? |
| 02:14.38 | jamesh | ResultSet.union() certainly does exist. |
| 02:14.40 | thumper | queries (complex) that return result sets of branch merge proposals |
| 02:14.47 | thumper | in which version of storm? |
| 02:14.58 | jamesh | every version that Launchpad has used |
| 02:15.10 | thumper | fires up make harness |
| 02:15.29 | jamesh | it isn't exposed through the security proxy though |
| 02:15.42 | jamesh | (if you're getting a forbidden error) |
| 02:16.20 | thumper | ah |
| 02:21.46 | jamesh | you should have access to ResultSet.find() for LP development now too, btw. |
| 02:22.00 | jamesh | if you want a way to narrow result sets |
| 02:22.35 | thumper | ah |
| 02:22.37 | thumper | what does that do? |
| 02:23.04 | jamesh | create a new result set from an existing one, but apply a few more constraints |
| 02:23.10 | thumper | cool |
| 02:23.34 | jamesh | e.g. store.find(Foo, a=1, b=2) is equivalent to store.find(Foo, a=1).find(b=2) |
| 02:25.13 | thumper | ForbiddenAttribute: ('_find_spec', <storm.store.ResultSet object at 0xbbb31ec>) |
| 02:25.22 | thumper | drat |
| 02:25.29 | thumper | that is removeSecurityProxy(a).union(b) |
| 02:25.38 | jamesh | remove the security proxy from b? |
| 02:25.42 | thumper | yeah |
| 02:25.55 | thumper | wonders about those |
| 02:26.42 | jamesh | with the security model of Launchpad, this is the sort of thing you should be doing on the model side. |
| 02:36.43 | thumper | hmm... |
| 02:37.25 | thumper | I'll have to deal with the security proxy anyway due to how we get the resultset in the first place |
| 02:40.04 | thumper | jamesh: IResultSet doesn't have a union |
| 02:40.19 | *** join/#storm bigdog1 (n=scmikes@72-197-8-8-arpa.cust.cinci.current.net) |
| 02:40.27 | thumper | jamesh: should it? |
| 02:40.33 | jamesh | thumper: right. IResultSet is the interface exposed through the security proxy |
| 02:40.42 | jamesh | and union() isn't exposed through the security proxy |
| 02:40.50 | thumper | why? |
| 02:41.58 | jamesh | because the usual model is for view code to just display a result set passed to it by the model code |
| 02:42.14 | thumper | frustrates |
| 02:42.21 | jamesh | there are a lot of public methods not exposed through the security proxy |
| 02:42.30 | thumper | thinks |
| 04:11.59 | *** join/#storm stub (n=stub@ppp-58-8-13-223.revip2.asianet.co.th) |
| 04:46.53 | *** join/#storm artista-frustrad (n=artista_@201-24-224-19.ctame704.dsl.brasiltelecom.net.br) |
| 05:50.02 | *** join/#storm jukart (i=lovely@81.189.156.94) |
| 06:50.02 | *** join/#storm elmom_ (n=elmom@hoasnet-ff05dd00-13.dhcp.inet.fi) |
| 07:34.11 | *** join/#storm zmijunkie (n=chatzill@i538732CF.versanet.de) |
| 07:57.08 | *** join/#storm seiflotfy1 (n=seif@ce1.hrz.tu-darmstadt.de) |
| 08:45.35 | *** join/#storm artista-frustrad (n=artista_@201-24-234-114.ctame704.dsl.brasiltelecom.net.br) |
| 08:56.15 | *** join/#storm wgrant (n=wgrant@ubuntu/member/wgrant) [NETSPLIT VICTIM] |
| 08:56.35 | *** join/#storm bigdog (n=scmikes@72-197-8-8-arpa.cust.cinci.current.net) |
| 09:25.51 | *** join/#storm bigdog (n=scmikes@72-197-8-8-arpa.cust.cinci.current.net) |
| 09:45.56 | *** join/#storm niemeyer (n=niemeyer@91.189.88.12) |
| 10:08.09 | *** join/#storm seiflotfy3 (n=seif@ce3.net.hrz.tu-darmstadt.de) |
| 10:47.17 | *** join/#storm keppla (n=keppla@i577B3F4A.versanet.de) |
| 11:13.45 | *** join/#storm seiflotfy (n=seif@ce1.hrz.tu-darmstadt.de) |
| 11:51.52 | *** join/#storm andrea-bs (n=andrea@ubuntu/member/beeseek.developer.andrea-bs) |
| 12:14.10 | *** join/#storm elmom_ (n=elmom@80.221.5.13) |
| 12:14.24 | *** join/#storm seiflotfy (n=seif@ce1.hrz.tu-darmstadt.de) |
| 12:22.44 | *** join/#storm fcorrea (n=fcorrea@187.3.194.219) |
| 12:23.51 | *** join/#storm fcorrea (n=fcorrea@187.3.194.219) |
| 12:31.39 | *** join/#storm jml (n=jml@mumak.net) |
| 12:45.47 | *** join/#storm artista-frustrad (n=artista_@201-24-229-193.ctame704.dsl.brasiltelecom.net.br) |
| 12:53.43 | *** join/#storm stub (n=stub@ppp-58-8-13-223.revip2.asianet.co.th) |
| 13:16.58 | *** join/#storm seiflotfy (n=seif@ce3.net.hrz.tu-darmstadt.de) |
| 13:20.15 | *** join/#storm drudi (n=drudi@189.117.31.160) |
| 14:05.39 | *** join/#storm hazmat (n=kapil@plone/hazmat) |
| 14:06.11 | hazmat | is there anything preventing inclusion of the twisted support in 0.16? |
| 14:12.05 | *** join/#storm hazmat (n=kapil@plone/hazmat) |
| 14:15.07 | therve | hazmat: yes, a lot |
| 14:15.14 | therve | like someone actively working on it :) |
| 14:26.59 | hazmat | therve, i'm about to put it into play on a project.. afaics there aren't any outstanding issues? |
| 14:28.09 | *** join/#storm seiflotfy (n=seif@ce3.net.hrz.tu-darmstadt.de) |
| 14:28.18 | therve | hazmat: there are. One major problem is that you can use an object in the main thread transparently |
| 14:28.25 | therve | but stuff will break sometimes |
| 14:28.46 | therve | and of course it will make blocking DB queries to populate attributes |
| 14:31.35 | hazmat | hmm.. i'll have to play with it then.. and see how it behaves.. the using an object in a main thread.. is a documentation issue imo.. the blocking db query on atttribute is troublesome.. but afaik this will only happen on a reference attr that hasn't been marked as a deferredreference |
| 14:31.58 | *** join/#storm drudi (n=drudi@189.116.183.247) |
| 14:33.33 | therve | hazmat: it's more than a documentation issue. It should break loudly if you do that |
| 14:40.46 | sidnei | therve: it should be safe enough for hazmat *wink* |
| 14:41.44 | sidnei | therve: he's done stuff way more dangerous than that with zodb anyway :P |
| 14:43.05 | therve | hehe |
| 14:43.08 | therve | well I'm just saying |
| 14:43.23 | therve | it would really cool if someone was using it |
| 14:43.35 | therve | (some people already do, actually) |
| 14:44.00 | therve | there have some discussions to merge it to trunk and mark it as unstable, but I'm not sure it's a good idea |
| 15:08.56 | hazmat | the concern is understandable.. but its also hard to get people using something that's going to be relegated to a bzr branch grave. |
| 15:27.48 | *** join/#storm jukart (n=jukart@d86-32-163-51.cust.tele2.at) |
| 15:32.02 | *** join/#storm seiflotfy (n=seif@ce1.hrz.tu-darmstadt.de) |
| 16:04.17 | *** join/#storm artista-frustrad (n=artista_@201-25-167-222.ctame704.dsl.brasiltelecom.net.br) |
| 17:03.41 | *** join/#storm jukart (n=jukart@d86-32-163-51.cust.tele2.at) |
| 17:14.50 | *** join/#storm bigdog (n=scmikes@72-197-8-8-arpa.cust.cinci.current.net) |
| 17:18.12 | *** join/#storm drudi (n=drudi@189.116.183.247) |
| 17:31.51 | *** join/#storm hazmat (n=kapil@plone/hazmat) |
| 17:33.59 | hazmat | is it possible to query with raw string? i've got a simple domain model.. mostly just a bag of strings.. if i model it as a RawStr then attempts to query out the value get no results.. if i switch it to unicode, then the query works fine.. |
| 17:36.47 | therve | weird |
| 17:37.13 | therve | hazmat: what's your database field? and the db engine? |
| 17:38.31 | hazmat | sqlite |
| 17:39.16 | hazmat | varchar field in the db |
| 17:40.09 | hazmat | works on example test case |
| 17:51.51 | hazmat | http://paste.pocoo.org/show/136403/ |
| 17:51.59 | hazmat | so if you execute that you get no results.. |
| 17:52.19 | hazmat | if you delete the fo.db ( sqlite) .. and use a Unicode field for host, then it works.. |
| 17:53.00 | *** join/#storm artista-frustrad (n=artista_@201-24-207-185.ctame704.dsl.brasiltelecom.net.br) |
| 17:53.31 | hazmat | worth a ticket? |
| 17:58.43 | therve | it doesn't look like a pb in storm itself |
| 18:01.14 | therve | oh |
| 18:09.42 | therve | hazmat: yeah you need to use BLOB type in the db if you want to use RawStr |
| 18:10.08 | therve | otherwise sqlite doesn't recognize the buffer type pass as parameter in the query |
| 18:14.38 | hazmat | so you have to use a blob for any non unicode str !? |
| 18:15.02 | hazmat | sigh.. i guess the assumption is ascii maps to unicode.. |
| 18:15.13 | hazmat | instead of an explicit str type |
| 18:27.34 | therve | ascii maps to unicode? I don't get it |
| 18:27.47 | therve | varchar and text maps to unicode |
| 18:28.08 | therve | because you can put unicode characters in those columns |
| 18:28.16 | therve | sqlite has a pretty bad type system |
| 18:35.51 | hazmat | i'm just saying its not obvious if i'm storing some ascii strings, that i need to use a unicode property |
| 18:37.32 | therve | no, if you're storing some ascii strings, you need to use a RawStr property, but a BLOB type in the db |
| 18:51.03 | hazmat | there's no reason for that to be a blob type in the db |
| 18:56.17 | therve | hazmat: there is, because otherwise it can store unicode characters |
| 20:34.59 | *** join/#storm hazmat (n=kapil@plone/hazmat) [NETSPLIT VICTIM] |
| 20:34.59 | *** join/#storm jamesh (n=james@canonical/launchpad/jamesh) [NETSPLIT VICTIM] |
| 20:34.59 | *** join/#storm radix (n=r@wordeology.com) [NETSPLIT VICTIM] |
| 20:40.13 | *** join/#storm oubiwann (n=oubiwann@97-119-5-224.omah.qwest.net) |
| 20:49.32 | *** join/#storm artista-frustrad (n=artista_@201-14-24-114.ctame704.dsl.brasiltelecom.net.br) |
| 21:06.22 | *** join/#storm jkakar (n=jkakar@S0106001ee57a879c.vc.shawcable.net) |
| 22:00.13 | *** join/#storm fcorrea (n=fcorrea@187.3.194.219) |
| 22:30.59 | *** join/#storm mac (i=52b655b7@gateway/web/freenode/x-vcrfcdwmiggkmxtx) |
| 22:31.32 | mac | hi everybody. Is this the right place to ask for help if I don't get something to work? |
| 22:31.36 | *** join/#storm artista-frustrad (n=artista_@201-41-183-159.ctame704.dsl.brasiltelecom.net.br) |
| 22:32.56 | mac | well... I will give it a crank anyway. |
| 22:33.04 | jml | mac, if you don't get something to work with storm, then yes |
| 22:33.16 | mac | thank you, so this is my problem: |
| 22:33.52 | mac | I have a legacy DB and I need to have a table that practicall add fields to another table (one-to-one) |
| 22:34.19 | mac | I tried to implement it by defining two classes and including in this line in one of the two: |
| 22:34.28 | mac | dati = Reference(id, ToOne.id) |
| 22:34.44 | mac | where id is a field of the class this line belongs to, and ToOne is another class. |
| 22:35.03 | mac | however, at runtime, I get an error telling me: |
| 22:35.41 | mac | NoneType object has no attribute nome, where nome is one of the fields of the ToOne class. |
| 22:36.33 | mac | Here is the code: http://pastebin.com/d7cb9f258 |
| 22:39.38 | mac | With the main body as well: http://pastebin.com/m5c3c294c |
| 22:46.16 | mac | I am using storm 0.14 ubuntu2 on Jaunty 64bit, if this matters. |
| 22:48.41 | *** join/#storm seiflotfy (n=seif@ce1.hrz.tu-darmstadt.de) |
| 22:51.00 | mac | anybody with a pointer? I look at the official tutorial and can't understand what i get different from the example there... |
| 22:57.10 | *** join/#storm hazmat (n=kapil@plone/hazmat) |
| 22:57.27 | mac | well, seems everybody is too busy to help me right now. Thank you anyhow, bye! :) |