*** Deyan has joined #ingres | 01:16 | |
Deyan | good morning | 01:16 |
---|---|---|
pboro | morning :) | 01:17 |
dyki | morning | 01:17 |
*** grantc has joined #ingres | 02:13 | |
*** ChanServ sets mode: +o grantc | 02:13 | |
*** thiagomz has joined #ingres | 03:28 | |
thiagomz | morning all | 03:29 |
Deyan | morning | 03:52 |
thiagomz | guys, i changed my isolation level to RC (read commited), but... if I execute a update on session S1 and a select on S2, I see the new value... why ? | 04:34 |
thiagomz | I not commited yet | 04:34 |
grantc | auto commit enabled? | 04:36 |
thiagomz | hummm | 04:50 |
thiagomz | how to disable autocommit ? | 04:50 |
thiagomz | im using isql | 04:50 |
grantc | set autocommit off | 04:52 |
thiagomz | could be it ==> system_readlock │nolock │string | 04:54 |
*** Alex| is now known as Alex|off | 05:04 | |
*** rossand has joined #ingres | 05:45 | |
*** ChanServ sets mode: +o rossand | 05:45 | |
thiagomz | grantc: how to do to ingres work as oracle, i mean... i changed the isolation level do read_commited... but.. if a update a row on session 1, and on session 2 i execute a select, it get still in wait... why ? | 05:55 |
thiagomz | i would like to see the old value .... | 05:56 |
thiagomz | on S2 | 05:56 |
thiagomz | i have a fresh install of ingres 9.2 | 05:56 |
thiagomz | i just changed ==> system_isolation │read_committed | 05:56 |
*** Deyan has quit IRC | 06:59 | |
*** mull has joined #ingres | 07:15 | |
pboro | thiagomz, did you restart after changing system_isolation? | 07:26 |
thiagomz | pboro: http://groups.google.com.br/group/comp.databases.ingres/browse_thread/thread/3369a8353199955a?hl=en# | 07:26 |
thiagomz | ingres doesnt support MVCC | 07:26 |
pboro | yeah I know it doesn't, it's a shame for web apps | 07:27 |
*** Alex|off is now known as Alex| | 07:27 | |
thiagomz | now.. I understood ! | 07:27 |
pboro | but that should not affect system_isolation, you should still get the old value | 07:28 |
pboro | or at least the second session should wait until the other succeeds, since the other session has exclusive lock (or at least it should afaik) | 07:28 |
pboro | have to try that on our installation :) | 07:29 |
*** Alex| is now known as Alex|off | 07:30 | |
*** Alex|off is now known as Alex| | 07:30 | |
pboro | thiagomz, for me, the other session hangs because the other session is keeping exclusive lock on the page | 07:34 |
pboro | try putting "set lockmode session where readlock = shared;" in beginning of your sessions | 07:34 |
thiagomz | pboro: yeah... | 07:34 |
pboro | maybe your readlock is nolock at the moment | 07:35 |
thiagomz | pboro: that works, but you get a dirty read | 07:35 |
thiagomz | you see ta NEW data... | 07:35 |
pboro | yeah, that's how it should work | 07:36 |
pboro | it's not MVCC, you don't get a snapshot, you only have classical ANSI isolation levels | 07:36 |
pboro | well, dirty reads depend on isolation too, but anyway | 07:36 |
pboro | read uncommitted allows dirty reads (in principle) | 07:37 |
*** DarylM has joined #ingres | 07:37 | |
*** Alex| is now known as Alex|off | 07:38 | |
*** atrofast has left #ingres | 08:02 | |
*** atrofast has joined #ingres | 08:08 | |
*** rossand has quit IRC | 08:49 | |
*** rossand has joined #ingres | 08:52 | |
*** irc.freenode.net sets mode: +o rossand | 08:52 | |
*** rossand has quit IRC | 08:57 | |
*** rossand has joined #ingres | 08:59 | |
*** irc.freenode.net sets mode: +o rossand | 08:59 | |
*** mc_ has joined #ingres | 09:22 | |
mc_ | Hi I have a .NET Data Provider question. | 09:22 |
pboro | shoot | 09:22 |
mc_ | I'm trying to insert some unicode into a nvarchar field but am getting garbage | 09:23 |
mc_ | I'm specifying character set as UTF-8 | 09:23 |
mc_ | is that correct? | 09:23 |
pboro | where do you specify the character set? | 09:23 |
mc_ | Connection string: ;BlankDate=NULL;TZ=UNITED-KINGDOM;" + "CharacterEncoding=UTF-8 | 09:24 |
mc_ | If I don't specify the CharacterEncoding it just replaces my unicode characters with ???? | 09:24 |
mc_ | Is there a unicode example anywhere? Can't find much on the web. :-( | 09:25 |
pboro | not that I know of :/ | 09:25 |
grantc | is the data in your .net app already in utf-8? | 09:25 |
mc_ | The data is in a standard VB.NET String data type and it looks okay on screen. I think that's UTF-16? | 09:26 |
mc_ | But if I choose an encoding of UTF-16 the driver doesn't connect to the database | 09:26 |
mc_ | it errors with some japanese / chinese error message, but when I trussed the back end it said invalid username / password so I don't think it understood what I sent | 09:27 |
grantc | ingres's nchar types are also UTF-16 | 09:27 |
grantc | so i don't believe you need to add any extra charset encoding to the url | 09:27 |
*** dyki has quit IRC | 09:28 | |
grantc | one test would be to get the hex value of what you are putting in and compare that with select hex(ncharcol) from .... | 09:28 |
grantc | the ???? could be a font issue or garbage data | 09:28 |
mc_ | If I don't add the character set encoding it just replaces my unicode characters with '?' characters, I stepped through the driver code and watched it do it! | 09:28 |
grantc | when you say unicode chars, which codepoint/language is the data? | 09:28 |
mc_ | I "trussed" the outgoing stream from the driver and it does actually change them to ? ascii characters. | 09:29 |
mc_ | It's some sample chinese text, but it does it for all values >255 | 09:29 |
grantc | ok - another test - you should have a database called demo db | 09:29 |
mc_ | Oracle worked perfectly first time. ;-) But I want to use Ingres | 09:29 |
grantc | sorry "demodb" | 09:29 |
grantc | mc_, :) | 09:29 |
grantc | in there there is a table which has some chinese data | 09:30 |
grantc | try the query "select up_first, up_last from user_profile" | 09:30 |
mc_ | No demodb on my installation, is there a script to build it. I suspect if I could get the unicode in, that getting it out again wouldn't be a problem. | 09:30 |
grantc | i'll have to check to see how the db gets created | 09:31 |
grantc | which version of Ingres is this? | 09:31 |
mc_ | Just about to say this is a fairly old installation: II 9.0.4 (su9.us5/105) | 09:32 |
mc_ | Solaris 8 | 09:32 |
grantc | hmm, and the Ingres client? | 09:32 |
mc_ | But I took the latest .net data provider to try things with. | 09:32 |
mc_ | I'm thinking about writing a Unix ESQLC program and inserting some wchar * stuff, hopefully that will work. | 09:33 |
grantc | let me see - perhaps there is a better way | 09:33 |
grantc | i have the data for demodb but its for an intel macine not sparc | 09:33 |
grantc | machine... | 09:34 |
mc_ | If it's UTF-16 I should be able to create a file and do a copy in? | 09:34 |
grantc | yup | 09:34 |
mc_ | I'm mostly concerned with getting the data out and I'm hoping the DataAdapter / DataSet will be okay, but we'll see. | 09:35 |
mc_ | QBF doesn't like the table at all. ;-) | 09:35 |
grantc | :) | 09:35 |
mc_ | Okay, thanks for the responses. I'll try a few things (tomorrow now, it's 17:36 here) and let you know how I get on. | 09:36 |
grantc | mc_ no problem - can you /msg me your email address, i might have a zip file with the demodb data | 09:37 |
mc_ | it's spam dot mc at gmail dot com (my spam address!) | 09:37 |
grantc | :) - i'll mail it out in the next 20 mins or so | 09:38 |
mc_ | thanks very much | 09:39 |
*** mc_ has left #ingres | 09:39 | |
*** Alex|off is now known as Alex| | 09:43 | |
*** Alex| is now known as Alex|off | 10:12 | |
*** rossand has quit IRC | 10:13 | |
*** Deyan has joined #ingres | 10:19 | |
Deyan | hello | 10:19 |
Deyan | wassup fellows? | 10:19 |
*** Deyan has quit IRC | 10:37 | |
thiagomz | wasaaaauppp | 11:16 |
*** DerMeister has joined #ingres | 13:44 | |
*** DerMeister has quit IRC | 14:43 | |
*** grantc has quit IRC | 15:07 | |
*** DarylM has quit IRC | 15:46 | |
*** Nornagest has quit IRC | 17:10 | |
*** bonro01 has quit IRC | 17:10 | |
*** atrofast has quit IRC | 17:10 | |
*** thiagomz has quit IRC | 17:10 | |
*** Alex|off has quit IRC | 17:10 | |
*** Vroomfondle has quit IRC | 17:10 | |
*** mull has quit IRC | 17:10 | |
*** toumi01 has quit IRC | 17:10 | |
*** clach04 has quit IRC | 17:10 | |
*** Alex|off has joined #ingres | 17:11 | |
*** atrofast has joined #ingres | 17:11 | |
*** mull has joined #ingres | 17:11 | |
*** thiagomz has joined #ingres | 17:11 | |
*** toumi01 has joined #ingres | 17:11 | |
*** Nornagest has joined #ingres | 17:11 | |
*** bonro01 has joined #ingres | 17:11 | |
*** clach04 has joined #ingres | 17:11 | |
*** Vroomfondle has joined #ingres | 17:11 | |
*** irc.freenode.net sets mode: +o Alex|off | 17:11 | |
*** Nornagest has quit IRC | 17:57 | |
*** Nornagest has joined #ingres | 17:57 | |
*** _Alex|off has joined #ingres | 17:58 | |
*** Alex|off has quit IRC | 18:00 | |
*** mull has quit IRC | 18:00 | |
*** rossand has joined #ingres | 19:33 | |
*** ChanServ sets mode: +o rossand | 19:33 | |
*** rossand has quit IRC | 20:18 | |
*** grantc has joined #ingres | 22:08 | |
*** ChanServ sets mode: +o grantc | 22:08 | |
*** _Alex|off is now known as Alex| | 22:35 | |
*** ChanServ sets mode: +o Alex| | 22:35 | |
*** Alex| is now known as Alex|off | 23:01 | |
*** Alex|off is now known as Alex| | 23:22 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!