> Why is this discussion in the private list at all? That's good thing you have moved to the discussion to the dev list.
-- With best regards / с наилучшими пожеланиями, Alexei Fedotov / Алексей Федотов, http://dataved.ru/ +7 916 562 8095 On Fri, Jun 7, 2013 at 12:14 AM, [email protected] <[email protected]> wrote: > Hi folks, > > I agree on point no 5 with Maxim: > *auto > reconnect feature stopped to work long ago* > => There was never such a feature. It is also rather complicated, cause > reconnect means that you basically out of sync, and the application has no > idea what has happened in between, so in theory you could have a reconnect > after 2 hours and the entire list of participants, screensharing, > whiteboards, et cetera _everything_ has changed. So the only chance I see > to realize a *reconnect* is to actually re-login the user into the room and > load everything from scratch. > > But about this entire discussion: > Why is this discussion in the private list at all? > There is no *privileged* channel to request or discuss features and bugs. > Please, lets use the dev list and give everyone that would like to be > involved the chance to at least read what you have to say. > @Alexei: For instance anybody reading the Analytics (or error > infrastructure :)) question now, is missing the context as how this email > that you wrote was initiated. > > Thanks, > Sebastian > > > 2013/6/7 Maxim Solodovnik <[email protected]> > >> 1) video codec was switched in 2.1.0, this is bug fix release I'm sure we >> shouldn't add major changes. Additionally we will need to use h264 in >> 3.0/3.1 since it is html5 compatible. >> >> 2) this behavior was not changed since 2.0. Please file an issue (better >> with patch) if you feel this need to be changed. (I would add configuration >> value default_gain with default value == 50) >> >> 3) I'll doublecheck >> >> 4) according to VOTE rules every user voting for release must perform these >> tests >> >> 5) I would vote for adding NOTE regarding flash version to release notes of >> 2.1.1 and alert to 3.0 (on room enter). Actually auto reconnect works as >> before .... it is just don't reconnect into room :) >> You can file JIRA issue (better with patch), but to be fair I don't really >> like such "last minute" changes. The release was announced long time ago >> and I feel we should close this branch and focus on 3.0 >> On Jun 6, 2013 7:23 PM, "Alexei Fedotov" <[email protected]> >> wrote: >> >> > Maxim, I've looked through the changes: >> > https://gist.github.com/leshikus/5720534 >> > most of them relate to log improvements, thus harmless. >> > >> > Here are few points on 2.1.1 release: >> > >> > 1. We haven't carefully tested H.264. Our customers use H.263 because >> > H.264 didn't work well for them. H,254 is a default codec for a new >> > release. Won't this be a problem? >> > >> > 2. There are hardcoded microphone settings for silence level, gain, >> > etc. From the other side I noticed that for the new release microphone >> > sensitivity is low, and one needs to set the sensitivity bar to the >> > maximum every time to get it to normal. Or this can be also related to >> > codec. >> > >> > 3. There are several places which incorrectly refer to "2.1.0" and >> > "2.0". The only place where the release is correct is a build file. >> > >> > 4. Who will volunteer to pass all tests for the new release, >> > xdocs/ManualTesting.xml? >> > >> > 5. Our users believe that reliability is most important part of >> > conferences. The last flash version does not work well, and auto >> > reconnect feature stopped to work long ago. How this can be addressed? >> > Should a flash version check be added (against flash 11.6 and 11.7), >> > and auto reconnect fixed? >> > >> > -- >> > With best regards / с наилучшими пожеланиями, >> > Alexei Fedotov / Алексей Федотов, >> > http://dataved.ru/ >> > +7 916 562 8095 >> > >> > >> > On Thu, Jun 6, 2013 at 1:42 PM, Alexei Fedotov <[email protected] >> > >> > wrote: >> > > Maxim, thanks for the tags! >> > > -- >> > > With best regards / с наилучшими пожеланиями, >> > > Alexei Fedotov / Алексей Федотов, >> > > http://dataved.ru/ >> > > +7 916 562 8095 >> > > >> > > >> > > On Mon, Jun 3, 2013 at 2:28 PM, Maxim Solodovnik <[email protected] >> > >> > wrote: >> > >> I believe ""Release early, release often" assumes we use one branch" >> > can be >> > >> followed with any number of branches :) >> > >> >> > >>>> "It is uncool when users come to support engineers" >> > >> I currently feel I'm the only support engineer :) >> > >> >> > >>>> Don't expect it become easier with html5 >> > >> I DO expect!, and things are MUCH easier with HTML5 :) >> > >> >> > >>>> How to get 2.1 code? >> > >> http://svn.apache.org/repos/asf/openmeetings/tags/2.1RC3/ >> > >> >> > >>>> How to get 2.1.1 code? >> > >> https://svn.apache.org/repos/asf/openmeetings/branches/2.1/ >> > >> >> > >>>> Were all these mentioned fixes ported to 3.0 (taking into account >> how >> > many >> > >> times I get no for particular fix, it would be surprising and great to >> > get >> > >> yes)? >> > >> I'm not sure what do you mean by "all these mentioned fixes", all >> fixes >> > >> made by me are in both branches :) >> > >> >> > >> >> > >> >> > >> >> > >> >> > >> On Mon, Jun 3, 2013 at 5:06 PM, Alexei Fedotov < >> > [email protected]>wrote: >> > >> >> > >>> Just one comment. "Release early, release often" assumes we use one >> > branch, >> > >>> where we fix errors quickly. Having two branches gives us the >> expected >> > >>> result that developers care less about users when develop in their >> > >>> development branch. When something is released, developers switch to >> a >> > new >> > >>> develioment branch and continue being cool and creative. >> > >>> >> > >>> It is uncool when users come to support engineers and they feed them >> > with >> > >>> tweaks and workarounds. You won't imagine the quirkiness of >> workarounds >> > >>> used to demonstrate a SIP video integration. >> > >>> >> > >>> Everyone wants to fix bugs when they are easy to reproduce. Most of >> > them >> > >>> are not due to the fact that openmeetings has a lot of unreliable >> > >>> technologies at its base. >> > >>> >> > >>> Don't expect it become easier with html5. Contrary, we will get own >> > >>> problems for each browser. That's why testing infrastructure is >> > important >> > >>> before we start releasing things. >> > >>> >> > >>> Maxim, regardless of the discussions. How to get 2.1 code? How to get >> > 2.1.1 >> > >>> code? Were all these mentioned fixes ported to 3.0 (taking into >> > account how >> > >>> many times I get no for particular fix, it would be surprising and >> > great to >> > >>> get yes)? >> > >>> 02.06.2013 15:06 пользователь "George Kirkham" < >> [email protected] >> > > >> > >>> написал: >> > >>> >> > >>> > Hi, >> > >>> > >> > >>> > My apologies, I had thought that 2.1.1 had been released. >> > >>> > >> > >>> > It would be good to release a stable version, whether that be 2.1.1 >> > or >> > >>> > 3.0.0 I don't care too much (as I do not know the difference >> between >> > >>> > versions). >> > >>> > >> > >>> > The main point is to have; >> > >>> > a) a stable released version for production use. The version would >> > be >> > >>> > used for production while new features are prepared and fixed in >> the >> > >>> > current development version. Fixes for any bugs that are >> discovered >> > >>> either >> > >>> > in production or in development should be made to both this >> > production >> > >>> > version and to the development version (where still applicable). >> > >>> > >> > >>> > b) a development version to which new features are added, bug fixes >> > for >> > >>> > issues discovered in production. Full testing should be performed >> > often, >> > >>> > and at least any time a new feature is added. >> > >>> > >> > >>> > At some point, additional features are put on hold, a full testing >> > >>> > completed, and then the development version can be released to >> > update the >> > >>> > production version, after which the development version is >> > incremented >> > >>> and >> > >>> > the cycle continues, the development version is updated with new >> > >>> features. >> > >>> > >> > >>> > One question I would have, do we have a test procedure (document) >> > that >> > >>> > lists all features to be tested? Such a test procedure can be used >> > to >> > >>> test >> > >>> > a version as it is prepared for production release, and after any >> > time >> > >>> that >> > >>> > bug fixes are applied to the production version. If we have such >> a >> > >>> > document, I would suggest it is not complete, as at times I have >> seen >> > >>> > issues that are obvious and have not been removed. >> > >>> > >> > >>> > Thanks, >> > >>> > >> > >>> > George Kirkham >> > >>> > >> > >>> > >> > >>> > -----Original Message----- >> > >>> > From: Maxim Solodovnik [mailto:[email protected]] >> > >>> > Sent: Sunday, 2 June 2013 12:30 PM >> > >>> > To: [email protected] >> > >>> > Subject: Re: [ANNOUNCE] upcoming 2.1.1 release >> > >>> > >> > >>> > I would vote for 2.1.1 release because: >> > >>> > 1) 17 issues were fixed: 1 blocker, 1 critical, 9 major. >> > >>> > 2) logging was improved: there is currently trivial way to get the >> > exact >> > >>> > OM version >> > >>> > 3) 3.0 is not ready yet >> > >>> > 4) "release early, release often" strategy >> > >>> > >> > >>> > PS your grep containing network related issues "Connection >> refused", >> > "not >> > >>> > Authenticated", "Broken pipe" is not very helpful >> > >>> > >> > >>> > >> > >>> > On Sun, Jun 2, 2013 at 1:01 AM, Alexei Fedotov < >> > [email protected] >> > >>> > >wrote: >> > >>> > >> > >>> > > There are two possible options. >> > >>> > > >> > >>> > > 1. Test heavily and ship 2.1.1 release. >> > >>> > > 2. Test heavily and ship 3.0 release instead. >> > >>> > > 3. Ship 3.0 pre-release without heavy testing. >> > >>> > > >> > >>> > > Why should we bother about testing in 2.1.1? Haven't we just >> > improved >> > >>> > > a few things? We don't actually know if we have improved things. >> > Any >> > >>> > > change have a risk of breaking something. >> > >>> > > >> > >>> > > My suggestion is to have testing infrastructure implemented, this >> > >>> > > would help in testing. This will provide feedback. I prefer >> option >> > >>> > > (2), or (3) because (1) is a waste of heavy testing effort. We >> > don't >> > >>> > > have enough resources to test two branches thoroughly. >> > >>> > > >> > >>> > > I constantly get suggestions (the last one from you one Friday), >> > like >> > >>> > > let's have a low quality development branch, where we will >> > progress, >> > >>> > > and be creative, and let's have someone to produce reliable >> > branches >> > >>> > > from these development branches. For people in a role of support >> > >>> > > engineers it's hard. That's another reason why I prefer (2) or >> (3) >> > to >> > >>> > > (1). >> > >>> > > >> > >>> > > BTW, here is start of the error log in our 2.1.1 installation. >> One >> > can >> > >>> > > start fixing existing problems, if we want a release. >> > >>> > > >> > >>> > > 45 at >> > >>> > > >> > org.apache.catalina.valves.ErrorReportValve.invoke(ErrorReportValve.ja >> > >>> > > va:99) [tomcat-embed-core-jar-7.0.34.jar:7.0.34] >> > >>> > > 45 at >> > >>> > > >> > org.apache.catalina.core.StandardEngineValve.invoke(StandardEngineValv >> > >>> > > e.java:118) [tomcat-embed-core-jar-7.0.34.jar:7.0.34] >> > >>> > > 42 Caused by: java.net.SocketException: Broken pipe >> > >>> > > 42 at >> > >>> > > >> > org.apache.openjpa.jdbc.sql.SQLExceptions.getStore(SQLExceptions.java: >> > >>> > > 62) >> > >>> > > ~[openjpa-bundle-2.2.1.jar:2.2.1] >> > >>> > > 42 at >> > >>> > > >> > org.apache.openjpa.jdbc.sql.SQLExceptions.getStore(SQLExceptions.java: >> > >>> > > 136) >> > >>> > > ~[openjpa-bundle-2.2.1.jar:2.2.1] >> > >>> > > 42 at >> > >>> > > >> > org.apache.openjpa.jdbc.sql.SQLExceptions.getStore(SQLExceptions.java: >> > >>> > > 110) >> > >>> > > ~[openjpa-bundle-2.2.1.jar:2.2.1] >> > >>> > > 42 at >> > >>> > > >> > org.apache.openjpa.jdbc.sql.DBDictionary.newStoreException(DBDictionar >> > >>> > > y.java:4918) >> > >>> > > ~[openjpa-bundle-2.2.1.jar:2.2.1] >> > >>> > > 42 at >> > >>> > > >> > org.apache.openjpa.jdbc.sql.DBDictionary.narrow(DBDictionary.java:4958 >> > >>> > > ) >> > >>> > > ~[openjpa-bundle-2.2.1.jar:2.2.1] >> > >>> > > 42 at >> > >>> > > >> > org.apache.openjpa.jdbc.kernel.JDBCStoreManager.connect(JDBCStoreManag >> > >>> > > er.java:971) >> > >>> > > ~[openjpa-bundle-2.2.1.jar:2.2.1] >> > >>> > > 42 at java.net.SocketOutputStream.socketWrite0(Native >> Method) >> > >>> > > ~[na:1.6.0_27] >> > >>> > > 42 at >> > >>> > > >> > com.mysql.jdbc.SQLError.createCommunicationsException(SQLError.java:11 >> > >>> > > 17) ~[mysql-connector-java-jar-5.1.22.jar:na] >> > >>> > > 42 at com.mysql.jdbc.MysqlIO.send(MysqlIO.java:3851) >> > >>> > > ~[mysql-connector-java-jar-5.1.22.jar:na] >> > >>> > > 32 at >> > >>> > > sun.reflect.NativeConstructorAccessorImpl.newInstance0(Native >> > >>> > > Method) ~[na:1.6.0_27] >> > >>> > > 21 at >> > >>> > > >> > org.springframework.orm.jpa.JpaTransactionManager.doCommit(JpaTransact >> > >>> > > ionManager.java:516) >> > >>> > > ~[org.springframework.orm-jar-3.1.1.RELEASE.jar:3.1.1.RELEASE] >> > >>> > > 21 at >> > >>> > > >> > org.springframework.orm.jpa.EntityManagerFactoryUtils.convertJpaAccess >> > >>> > > ExceptionIfPossible(EntityManagerFactoryUtils.java:326) >> > >>> > > ~[org.springframework.orm-jar-3.1.1.RELEASE.jar:3.1.1.RELEASE] >> > >>> > > 21 at >> > >>> > > >> > org.springframework.orm.jpa.DefaultJpaDialect.translateExceptionIfPoss >> > >>> > > ible(DefaultJpaDialect.java:120) >> > >>> > > ~[org.springframework.orm-jar-3.1.1.RELEASE.jar:3.1.1.RELEASE] >> > >>> > > 19 FailedObject: select c from Sessiondata as c where >> > >>> > > c.session_id LIKE :session_id [java.lang.String] >> > >>> > > 16 java.lang.NullPointerException: null >> > >>> > > 12 at >> > >>> > > >> > org.apache.openmeetings.remote.ConferenceService.getRoomsByOrganisatio >> > >>> > > nWithoutType(ConferenceService.java:147) >> > >>> > > ~[openmeetings-2.1.1-SNAPSHOT.jar:na] >> > >>> > > 10 at >> > >>> > > sun.reflect.GeneratedConstructorAccessor283.newInstance(Unknown >> > >>> > > Source) ~[na:na] >> > >>> > > 4 at >> > >>> > > >> > org.apache.openmeetings.remote.red5.ScopeApplicationAdapter.roomLeaveB >> > >>> > > yScope(ScopeApplicationAdapter.java:564) >> > >>> > > [openmeetings-2.1.1-SNAPSHOT.jar:na] >> > >>> > > 2 java.lang.Exception: not Authenticated >> > >>> > > 2 ERROR 05-28 20:44:56.425 o.a.o.r.WhiteBoardService:408 >> > >>> > > [NioProcessor-6] - WhiteboardSyncLockObject not found for this >> > Client >> > >>> > > {} >> > >>> > > 2 ERROR 05-28 20:44:56.423 o.a.o.r.WhiteBoardService:408 >> > >>> > > [NioProcessor-6] - WhiteboardSyncLockObject not found for this >> > Client >> > >>> > > {} >> > >>> > > 2 Caused by: >> > >>> > > com.mysql.jdbc.exceptions.jdbc4.CommunicationsException: The last >> > >>> > > packet successfully received from the server was 84 612 291 >> > >>> > > milliseconds ago. The last packet sent successfully to the >> server >> > was >> > >>> > > 84 612 291 milliseconds ago. is longer than the server configured >> > >>> > > value of 'wait_timeout'. You should consider either expiring >> and/or >> > >>> > > testing connection validity before use in your application, >> > increasing >> > >>> > > the server configured values for client timeouts, or using the >> > >>> > > Connector/J connection property 'autoReconnect=true' to avoid >> this >> > >>> > > problem. >> > >>> > > 2 at >> > >>> > > >> > org.apache.openmeetings.axis.services.FileWebService.getFileExplorerBy >> > >>> > > RoomSelf(FileWebService.java:728) >> > >>> > > ~[openmeetings-WebService-2.1.1-SNAPSHOT.jar:na] >> > >>> > > 1 org.springframework.orm.jpa.JpaSystemException: The last >> > >>> > > packet successfully received from the server was 84 612 291 >> > >>> > > milliseconds ago. The last packet sent successfully to the >> server >> > was >> > >>> > > 84 612 291 milliseconds ago. is longer than the server configured >> > >>> > > value of 'wait_timeout'. You should consider either expiring >> and/or >> > >>> > > testing connection validity before use in your application, >> > increasing >> > >>> > > the server configured values for client timeouts, or using the >> > >>> > > Connector/J connection property 'autoReconnect=true' to avoid >> this >> > >>> > > problem.; nested exception is <openjpa-2.2.1-r422266:1396819 >> fatal >> > >>> > > general error> >> org.apache.openjpa.persistence.PersistenceException: >> > >>> > > The last packet successfully received from the server was 84 612 >> > 291 >> > >>> > > milliseconds ago. The last packet sent successfully to the >> server >> > was >> > >>> > > 84 612 291 milliseconds ago. is longer than the server configured >> > >>> > > value of 'wait_timeout'. You should consider either expiring >> and/or >> > >>> > > testing connection validity before use in your application, >> > increasing >> > >>> > > the server configured values for client timeouts, or using the >> > >>> > > Connector/J connection property 'autoReconnect=true' to avoid >> this >> > >>> > > problem. >> > >>> > > 1 org.apache.openjpa.persistence.PersistenceException: The >> > last >> > >>> > > packet successfully received from the server was 39 315 457 >> > >>> > > milliseconds ago. The last packet sent successfully to the >> server >> > was >> > >>> > > 39 315 457 milliseconds ago. is longer than the server configured >> > >>> > > value of 'wait_timeout'. You should consider either expiring >> and/or >> > >>> > > testing connection validity before use in your application, >> > increasing >> > >>> > > the server configured values for client timeouts, or using the >> > >>> > > Connector/J connection property 'autoReconnect=true' to avoid >> this >> > >>> > > problem. >> > >>> > > 1 java.net.ConnectException: Connection refused >> > >>> > > 1 INFO 05-29 14:53:34.184 o.a.o.d.u.d.UsersDao:163 >> > >>> > > [NioProcessor-4] - [getUser] Info: No USER_ID given >> > >>> > > 1 INFO 05-28 15:46:15.535 o.a.o.d.u.d.UsersDao:163 >> > >>> > > [NioProcessor-14] - [getUser] Info: No USER_ID given >> > >>> > > 1 FailedObject: select c from User as c where c.user_id = >> > >>> > > :user_id AND c.deleted <> true [java.lang.String] >> > >>> > > >> > >>> > >> > >>> > >> > >>> > >> > >>> > -- >> > >>> > WBR >> > >>> > Maxim aka solomax >> > >>> > >> > >>> >> > >> >> > >> >> > >> >> > >> -- >> > >> WBR >> > >> Maxim aka solomax >> > >> > > > > -- > Sebastian Wagner > https://twitter.com/#!/dead_lock > http://www.webbase-design.de > http://www.wagner-sebastian.com > [email protected] <[email protected]>
