| *** grantc has quit IRC | 00:05 | |
| *** grantc has joined #ingres | 00:06 | |
| *** ChanServ sets mode: +o grantc | 00:06 | |
| *** grantc changes topic to "Ingres Chat | Community - http://community.ingres.com | Channel Archives http://irc.planetingres.org" | 00:06 | |
| *** magnusg| has joined #ingres | 00:15 | |
| *** grantc has quit IRC | 00:33 | |
| *** grantc has joined #ingres | 00:33 | |
| *** ChanServ sets mode: +o grantc | 00:33 | |
| *** dyki has joined #ingres | 01:36 | |
| dyki | morning all | 01:36 |
|---|---|---|
| *** Deyan has joined #ingres | 01:43 | |
| Deyan | meep | 01:43 |
| *** Deyan has quit IRC | 02:27 | |
| *** Gerhard has joined #ingres | 03:07 | |
| Alex| | Hi grantc | 03:33 |
| grantc | morning bloke | 03:34 |
| Alex| | is it possible for us to offer an rpm of the php driver? | 03:34 |
| Alex| | I guess this would only work for a specific php version then? | 03:35 |
| grantc | anything is possible you know ;) - i would have to look in to it but i guess | 03:35 |
| grantc | more strict than php version, i will be tied to the vendor of the distro | 03:35 |
| grantc | be it RedHat / SuSE / Mandrake... | 03:35 |
| grantc | s/i will/it will | 03:35 |
| Alex| | we're about to create an appliance template for partners - one distribution, one arch | 03:36 |
| grantc | studio appliance? | 03:36 |
| Alex| | yup | 03:36 |
| Alex| | and as they work best with rpms, this would be the easiest solution to get the driver in | 03:36 |
| grantc | yup... i am just thinking | 03:37 |
| grantc | iirc novell have a single src RPM that builds each extension | 03:37 |
| grantc | i guess there are multiple spec files within that src rm | 03:37 |
| grantc | rpm | 03:37 |
| Alex| | you want to compile the extension? For a known php version, providing the so would be enough, wouldn't it? | 03:38 |
| grantc | perhaps | 03:39 |
| Alex| | because I don't think that the base template will include a toolchain | 03:39 |
| grantc | drop teresa an email so she is aware of the requirement | 03:39 |
| grantc | and i will take a look at it | 03:39 |
| *** Deyan has joined #ingres | 03:39 | |
| Deyan | hello guys | 03:40 |
| grantc | hi Deyan | 03:40 |
| Gerhard | Hi all, are there any plans for Ingres to add a "drop table if exists" like there is in MySQL? | 03:40 |
| Deyan | Gerhard: i need that too... | 03:40 |
| grantc | Gerhard: not that i am aware of (which is not a good benchmark) | 03:41 |
| grantc | Alex|: i would need a spec file to create the rpm, the toolchain needed for building php5 should be enough | 03:42 |
| Alex| | no chance to deploy the pure .so file? :( | 03:43 |
| grantc | afaik the extensions require the same toolchain to build as does the mother language | 03:43 |
| grantc | Perhaps - I would need to chat with Jay | 03:43 |
| Deyan | grantc: http://ingres.pastebin.com/d123e5d0c | 03:44 |
| Alex| | I'm pretty sure the base system is pure binary, no gcc, etc | 03:44 |
| Gerhard | copydb's -add_drop switch is really nice but would be even nicer if there was a "drop table if exists" in the resulting copy.in | 03:44 |
| grantc | sure but there will be a dev environment | 03:44 |
| grantc | Gerhard: log it as a SIR/Enhancement Request | 03:45 |
| Deyan | grantc: and http://ingres.pastebin.com/d204ba021 | 03:45 |
| Deyan | the second URI contains output | 03:45 |
| Alex| | well, yes, there would be a suse to build the thing | 03:45 |
| Deyan | ingres_close() simply does not return | 03:45 |
| Deyan | it blocks | 03:45 |
| grantc | Deyan: i would need a test case to debug it | 03:46 |
| grantc | i.e. a copy of the data and tables | 03:46 |
| grantc | Alex|: exactly - like with rPath | 03:46 |
| Deyan | ipm says there are no locks | 03:46 |
| Deyan | :( | 03:46 |
| Alex| | do you know who maintains the rpath update channel? | 03:46 |
| Alex| | I guess chris and his team? | 03:47 |
| grantc | Deyan: if you build it in debug then attach a debugger | 03:47 |
| grantc | Alex|: Jay or Ray | 03:47 |
| Alex| | ok | 03:47 |
| grantc | Ray knocked up a recipe for the driver in conary | 03:47 |
| *** Gerhard has quit IRC | 04:25 | |
| Deyan | i just executed an instert query on a table which has 15 foreign keys | 05:05 |
| Deyan | and it was running for few minutes | 05:05 |
| Deyan | a simple insertion of one, single row | 05:06 |
| Deyan | is this common in Ingres? | 05:06 |
| Alex| | maybe there was a lock on one of the related tables/indexes | 05:08 |
| Deyan | nope | 05:08 |
| Deyan | i examined | 05:09 |
| Deyan | no active locks | 05:09 |
| Deyan | (9.3 version btw) | 05:09 |
| Alex| | well, maybe it was a very complicated insert........ no idea. Even when indexes need to be updated, Inserts shouldn't take minutes | 05:10 |
| *** rossand has joined #ingres | 05:19 | |
| *** ChanServ sets mode: +o rossand | 05:19 | |
| magnusg| | Deyan, I don't think tables with 15 foreign keys are very common but I agree with Alex| that it absolutely should not take minutes if indexes and tables are OK | 05:21 |
| *** atrofast has quit IRC | 05:24 | |
| *** grantc has quit IRC | 05:30 | |
| *** atrofast has joined #ingres | 05:31 | |
| *** grantc has joined #ingres | 05:31 | |
| *** ChanServ sets mode: +o grantc | 05:31 | |
| *** grantc has quit IRC | 05:31 | |
| *** grantc has joined #ingres | 05:32 | |
| *** ChanServ sets mode: +o grantc | 05:32 | |
| *** grantc has quit IRC | 05:33 | |
| *** atrofast has quit IRC | 05:38 | |
| *** atrofast has joined #ingres | 05:38 | |
| *** atrofast has joined #ingres | 05:39 | |
| *** grantc has joined #ingres | 05:41 | |
| *** ChanServ sets mode: +o grantc | 05:41 | |
| *** atrofast1 has joined #ingres | 05:46 | |
| *** atrofast has quit IRC | 05:47 | |
| *** thiagomz has joined #ingres | 05:59 | |
| atrofast1 | Anyone here familiar with the DM_DATA struct? What format is the data stored in it anyway? | 06:32 |
| toumi01 | atrofast1: as I understand it DM_DATA is just an all-purpose header for variable length data that has no single format - it depends on what is being passed | 06:37 |
| atrofast1 | Does one instance of DM_DATA define a whole row? | 06:38 |
| toumi01 | I thought it was more general purpose than that. What's triggering the question? | 06:42 |
| atrofast1 | Well I'm working on a new (separate) system catalog that is updated when a new table is created (one example)... If I go through dmf_call I have to use the DM_DATA struct to pass the data... | 06:43 |
| atrofast1 | Maybe I'm totally off here, maybe I'm supposed to use another facility | 06:43 |
| toumi01 | Interesting. What project is this for? | 06:44 |
| atrofast1 | geospatial :) | 06:45 |
| toumi01 | of course! | 06:45 |
| toumi01 | unless you're doing something radically different with your catalog entries I'd think that looking at what is currently done would be the best "documentation" - but I have to admit that reading tea leaves seems easier than reading dmf sometimes | 06:47 |
| atrofast1 | Hehe, thanks toumi01! | 06:49 |
| *** mull has joined #ingres | 06:50 | |
| *** Alex| is now known as Alex|off | 06:54 | |
| *** dyki has quit IRC | 07:00 | |
| *** dyki has joined #ingres | 07:17 | |
| *** magnusg| has left #ingres | 07:42 | |
| Deyan | grantc: something is 100% wrong with ingres_close() | 07:56 |
| Deyan | it blocks and takes 100% of CPU | 07:56 |
| pboro | Deyan, can you provide a test case? | 07:57 |
| Deyan | i will try, tomorrow | 07:57 |
| pboro | great | 07:57 |
| Deyan | but i really have no clue what to put in the test case | 07:57 |
| Deyan | because sometimes everything works well | 07:57 |
| Deyan | sometimes not | 07:57 |
| pboro | ah, it is not easily reproducible? have you tried debugging it from both sides, ie. from the server and from the client side? | 07:58 |
| Deyan | see, i am not an Ingres expert | 07:59 |
| Deyan | i just just compared my experience with MySQL, Oracle and SQLite | 07:59 |
| Deyan | with what I have here | 08:00 |
| Deyan | something like this never happened with other extensions | 08:00 |
| Deyan | there must be something in ingres_close() | 08:00 |
| Deyan | it simply should not block forever | 08:00 |
| pboro | since it's blocking it should be pretty easy to check it using gdb or strace | 08:01 |
| pboro | and you don't have to be ingres expert to debug a software :) you just need to be a software expert... | 08:01 |
| Deyan | we are going to use JDBC anyway | 08:02 |
| Deyan | we have wasted too much time using php extension | 08:02 |
| Deyan | ty for your help | 08:03 |
| pboro | I use JDBC a lot, it has worked fine | 08:07 |
| Deyan | good | 08:07 |
| Deyan | :) | 08:07 |
| pboro | in fact, it's the only thing I really use, since I develop web apps using Java EE | 08:08 |
| pboro | ah, gotta run -> | 08:08 |
| Deyan | which version of ingres are you using, pboro | 08:08 |
| Deyan | ? | 08:08 |
| pboro | Deyan, Ingres 9.2 with patch 13457 | 08:35 |
| *** Alex|off is now known as Alex| | 08:51 | |
| pboro | I think that one of the problems with community versions is that they are simply a snapshot from the past which are not anymore maintained by anyone | 08:55 |
| pboro | I hope that when 9.3 releases there will be a branch & tag for community which the community can then work further (bug fixing etc. stuff) | 08:56 |
| pboro | the current community version of 9.2 is nice but the build system is utterly outdated and there are pretty many bugs which would need fixing before it would be usable, and since the fix history for 9.2 is a bit absent, fixing the community version is a bit boo-hoo | 08:57 |
| Deyan | does that mean basically that if i want an Ingres server for production use, i should buy it from Ingres? | 09:07 |
| pboro | no, but without support you're on your own (well that's how it's actually with any free open source dbms...) | 09:09 |
| pboro | we have a support subscription from Ingres, the service has been really good | 09:10 |
| Deyan | yes, but when i do "yum install mysql*" on my Fedora box, it gives me a rock-stable server | 09:10 |
| Deyan | here i am managing to crash server few times per day | 09:10 |
| Deyan | with quite simple PHP script(s) | 09:10 |
| pboro | yeah, that's probably true since 9.3 and 9.4 are still in development and 9.2 is pretty old and have not been updated since it was released | 09:10 |
| pboro | I guess 9.2 should be better, but I wouldn't even consider trying 9.3 or 9.4 for production use... but that's just my opinion | 09:11 |
| *** grantc has quit IRC | 09:20 | |
| *** DerMeister has joined #ingres | 09:20 | |
| Alex| | Hi rossand | 09:42 |
| rossand | Hi Alex| | 09:42 |
| Alex| | I've just created a blog post in Drupal... does the preview ignore <p>s ? There are no paragraphs | 09:43 |
| rossand | Ah, set the input format to html. That should clear that up. | 09:44 |
| rossand | I'll see if we can make html the default. Should be able to. | 09:44 |
| Alex| | hum, I now added linebreaks (not <br> but real breaks) and they are ok | 09:44 |
| rossand | the editor is wysiwyg. You can flip it to html mode as well. | 09:45 |
| Alex| | yup, I added them in html mode | 09:45 |
| thiagomz | guys, what is the command to change a table page size ? | 09:45 |
| Alex| | linebreak in wysiwyg will add <p> but at least the preview doesn't render the <p> | 09:45 |
| Alex| | modify table to btree|hash|heap... with page_size = 8192 | 09:46 |
| Alex| | rossand - do you publish blog posts or do I need to do something to make it visible? | 09:46 |
| rossand | Alex|: Under publishing options, select published and promoted to the front page (if you want it to show up there). Please avoid sticky at the top of lists. | 09:49 |
| rossand | You have rights to do so. | 09:50 |
| Alex| | hm, where do I find that? Don't see that option | 09:50 |
| rossand | Nice web page! You got the theme perfect. | 09:50 |
| Alex| | wasn't that hard with c'n'p ;) | 09:50 |
| rossand | checking to make sure you're in the right group. few seconds. | 09:51 |
| Alex| | wanted to keep it as near to the .org as possible since we only want to add a few german pages that contain the necessary files such as becoming member, etc... the rest will stay at .org. We don't want to drive traffic away from there | 09:52 |
| rossand | makes sense. | 09:52 |
| rossand | I'll debug in a sec, someone popped in for a moment. | 09:53 |
| thiagomz | Alex|: tkz | 09:54 |
| *** dyki has quit IRC | 10:04 | |
| Alex| | rossand - I'm off now. If you can publish the post that would be great. We can take a look at the rights later | 10:07 |
| *** Alex| is now known as Alex|off | 10:08 | |
| *** Deyan has quit IRC | 10:11 | |
| rossand | Alex|off: Sure, no problem. | 10:11 |
| *** atrofast1 is now known as atrofast | 10:17 | |
| pboro | Hmm... is Ingres 10.0 ex-9.4? | 11:15 |
| * thiagomz x | 11:16 | |
| * thiagomz www.thiagomz.com ;-) | 11:16 | |
| pboro | Oracle certified? Nice. I wonder if it's possible to get Ingres certification :) | 11:17 |
| *** siman01 has joined #ingres | 11:42 | |
| siman01 | anyone got any idea why embeding python in ome functions work on win32 but not linux - I'm worried it is because ingres was compiled with a different compiler to python | 11:53 |
| siman01 | it segfaults on the line: pModule = PyImport_ImportModule("iiudf"); | 11:57 |
| siman01 | for anyone with too much time on there hands gdb stacktrace is here http://pastebin.com/f4ceb4d0b | 12:17 |
| siman01 | includes debug symbols | 12:18 |
| *** siman01 has quit IRC | 12:36 | |
| thiagomz | anyone has a script to verify overflow in my tables ? | 12:39 |
| thiagomz | ;-) | 12:39 |
| thiagomz | i did one | 12:46 |
| thiagomz | :D | 12:46 |
| *** rossand has quit IRC | 13:36 | |
| *** rossand has joined #ingres | 13:38 | |
| *** ChanServ sets mode: +o rossand | 13:38 | |
| *** rossand has quit IRC | 13:59 | |
| *** DerMeister has quit IRC | 14:16 | |
| *** FrankW has joined #ingres | 14:30 | |
| *** rossand has joined #ingres | 14:57 | |
| *** ChanServ sets mode: +o rossand | 14:57 | |
| FrankW | I seek to build ingres on 64bit linux. Can anyone point me to a good doc to walk me through the process? I'll eventually want to add geospatial support if that is not in the core. It would likely be best to build from development head I think. | 15:02 |
| *** rossand1 has joined #ingres | 15:07 | |
| *** rossand has quit IRC | 15:08 | |
| *** rossand1 has quit IRC | 15:47 | |
| *** rossand has joined #ingres | 15:49 | |
| *** ChanServ sets mode: +o rossand | 15:49 | |
| *** mull has quit IRC | 16:01 | |
| *** rossand has quit IRC | 16:10 | |
| *** rossand has joined #ingres | 16:12 | |
| *** ChanServ sets mode: +o rossand | 16:12 | |
| *** rossand has quit IRC | 16:12 | |
| FrankW | Hmpf, svn checkout on ingres/src/tools/techpub/pdf takes a while. | 16:53 |
| * FrankW is defeated by the need for libpam0g-dev which I'm not able to install since my gutsy ubuntu distribution is no longer supported by the ubuntu people. | 17:11 | |
| FrankW | I'm going to be driven to upgrade yet! | 17:11 |
| *** FrankW has quit IRC | 17:14 | |
| *** Alex|off is now known as Alex| | 22:36 | |
| *** Alex| is now known as Alex|off | 22:40 | |
| *** magnusg| has joined #ingres | 23:46 | |
| *** Alex|off is now known as Alex| | 23:48 | |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!