About the plugin executer: The plugin executer was created (as the name suggest) as a bridge for base1 plugins and not as a generic way of executing none java plugins. As for now we don't have the recourses to evolve it in that direction, the first priority is to get it to work with existing plugins.
So if you want to develop a none java plugin for base, whats your options? - Using a java wrapper: The simplest way is to create a java application that wraps your plugin and handles the communication with base. The Base1PluginExecuter and the RMAExpress <http://lev.thep.lu.se/trac/baseplugins/wiki/thep.lu.se.RMAExpress> are two examples of this. Of course, this requires a little java knowledge but I think that you can take most of the code from the examples. - Webservice A web service branch has been created for Base and is making progress. It will be included in a future release (2.4 earliest). With this you communicate with base using any language. Note however that this interface won't be able to do all the things a plugin using the java api can. The webservices will be focused around BioAssaySets and the analytic data. There is a trac and a subversion repository created for external plugins, http://baseplugins.thep.lu.se/. Some answers for your questions in the reply; > I guess execName is sensibly ignored... The execName is used to execute the plugin. This is the code for that: String execPath = (String) configuration.getValue(execPathParameter.getName()); String execName = (String) configuration.getValue(execNameParameter.getName()); Process p = Runtime.getRuntime().exec(execPath + java.io.File.separator + execName, null, getExecDirectory()); This means that I grab the execpath and execname from the configuration. Then I create a new Process using Runtime.exec(...). > 'Error: Can't find the columns intensity1/intensity2 or > l2ratio1_2/l10intgmean1_2. No data could be inmported.' Ups. Guess I got clouded by all the two channel data that i test on. You should of course be able to import one channel if thats what your data is. > Maybe I am expecting too much until allIntensities is working.... I will load > in some 2 channel data and do some further testing. How do I reopen a ticket > if I need to ? > This has nothing to do with the allIntensities parameter, I just have to make the import part a little bit more sensible. You can reopen a ticket on the edit page but don't if you don't intend to follow up the discussion. As I see it, the bug in ticket 511 is fixed. /Johan Steve Taylor wrote: > Johan Enell wrote: > >> I think it's best to discuss this in a new thread. >> >> The easiest way to go around this bug is to define a directory for the >> Experiment. Edit Experiment and add a directory. >> > > Ok, thanks. That got me a bit further... > > >> You will also need to change the field '_allIntensities' to >> intesity1\tintensity2. The suport for allIntensities till be added in a >> future release. >> >> > > So I changed this to intensity1 (only), since I was trying to analyse some > single channel data (I haven't loaded any 2 channel data in BASE2 yet). I > then had to change the base 1.2 file export > > uniqueName molbio.mgh.harvard.edu/dpark/limitint > > to 'basefile-general.pl' > > and the exec path to > /usr/local/base2/www/plugins/molbio.mgh.harvard.edu/dpark/. > > I guess execName is sensibly ignored...I created the directory to run the > plugin within the experiment. I then run it and get: > > 'Error: Can't find the columns intensity1/intensity2 or > l2ratio1_2/l10intgmean1_2. No data could be inmported.' > Maybe I am expecting too much until allIntensities is working.... I will load > in some 2 channel data and do some further testing. How do I reopen a ticket > if I need to ? > > I have a few follow up questions and comments about plugins...I am trying to > get MGH plugins from BASE2 working. Unfortunately I am not having much luck > (see ticket http://base.thep.lu.se/ticket/511). > > > >> It is fixed >> >> >>> So my questions are: >>> >>> 1) Which plugins have been successfully tested in Base1PluginExecuter (just >>> so I can rule out it's not me!)? >>> >>> >>> >> I've successfully with MergeBioassays, >> http://lev.thep.lu.se/trac/baseplugins/wiki/onk.lu.se.MergBioAssay >> >> > > >>> 2) Do you require some help testing Base1PluginExecuter? I am happy to help >>> here to get this working for the common good and also I have a several >>> custom plugins I have written in R and perl (and other plugins used in BASE >>> 1.2) that I would really like to use on the data in BASE. Also I can't >>> really move all my users of BASE 1.2 across to BASE2 until I get these >>> working, since they use these as part of their daily analysis. >>> >>> >>> >> Yes, we appreciate all help that we can get. Especially testing since it >> feels like we never have time to do it as thorough as we would like. As >> far as I know all base 1 plugins should work except for those using the >> option 'geneAverages'. >> >> >>> 3) What are the plans to support non-Java plugins (via development of >>> Base1PluginExecuter or otherwise) given that R and perl are very popular >>> languages in bioinformatics and for arrays most serious >>> analysis is done using R? >>> >>> >>> >> I leave this question open for now. Let me come back on this. >> >> > > Just my 2c. If all the BASE1.2 plugins can be got to work robustly then > perhaps you can evolve the BASE 1.2 plugin executer code? With an extra > configuration screen for passing in variables this would > be useful...Also long term it could be worth looking at what the EBI are > doing with the developer API for the Expression Profiler > (http://www.ebi.ac.uk/expressionprofiler/). The bonus if this approach > was implemented would mean any plugins for BASE could in theory work in > Expression Profiler too. I realise the latter would take a fair amount of > work and coordination...:-) > > BTW where is the central list of BASE 1.2 plugins listed these days? If it is > verified they work in BASE 2 it would be good to check them into the > subversion repository, with the appropriate config files. > > Regards and thanks, > > Steve > ------------------------------------------------------------------ > Medical Sciences Division > Weatherall Institute of Molecular Medicine/Sir William Dunn School > Oxford University > > ------------------------------------------------------------------------- > Take Surveys. Earn Cash. Influence the Future of IT > Join SourceForge.net's Techsay panel and you'll get the chance to share your > opinions on IT & business topics through brief surveys-and earn cash > http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV > _______________________________________________ > The BASE general discussion mailing list > basedb-users@lists.sourceforge.net > unsubscribe: send a mail with subject "unsubscribe" to > [EMAIL PROTECTED] > ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ The BASE general discussion mailing list basedb-users@lists.sourceforge.net unsubscribe: send a mail with subject "unsubscribe" to [EMAIL PROTECTED]
