You can subscribe to this list here.
2003 |
Jan
|
Feb
|
Mar
|
Apr
(3) |
May
(7) |
Jun
|
Jul
|
Aug
(3) |
Sep
|
Oct
|
Nov
|
Dec
|
---|---|---|---|---|---|---|---|---|---|---|---|---|
2004 |
Jan
|
Feb
(46) |
Mar
(7) |
Apr
|
May
(1) |
Jun
|
Jul
|
Aug
|
Sep
(23) |
Oct
(9) |
Nov
|
Dec
|
2017 |
Jan
|
Feb
|
Mar
|
Apr
(1) |
May
|
Jun
(1) |
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
From: Nascif Abousalh-N. <na...@al...> - 2017-06-18 04:46:38
|
Hello JDiSeq http://www.myhaberci.net/vwd_justso.php?farmer=2vb6na8a0xku6mh |
From: nascif2 <na...@in...> - 2017-04-08 04:19:41
|
good morning JDiSeq http://www.churchofthelordjesus.com/wp-signup.php?quiet=2gw5mqscy6b0ny2 |
From: Orjan N. A. <aus...@co...> - 2004-10-05 06:44:26
|
prefuse was asesome. Really responsive. I have had a look on JAVAVIS some time ago, but (as of then) I found=20 that it didn't suit our need. Anyhow its interressting that JAVAVIS=20 tried to render sequence diagrams the same way as jDiSeq (JDI). I could no find any information on the work by Eichelberger... On the links page on prefuse.sf.net I found Piccolo, a replacement for=20 what was know as Jazz. Download the source and try out examples.jar. =D8rjan Nascif Abousalh-Neto wrote: > Found some interesting material on graph visualization today, check > this out: >=20 > http://prefuse.sourceforge.net/ (awesome, check the demos) >=20 >=20 > [20] H. Eichelberger, J. Wolf von Gudenberg, On the Visualization of > Java Programs, in: S. Diehl (Ed.), Software Visualization, Dagstuhl > Seminar, LNCS 2269, Springer, Berlin, 2002, pp. 295-306.=09 > [21] K. Oechsle, T. Schmitt, JAVAVIS: Automatic Program Visualization > with Object and Sequence Diagrams using the Java Debug Interface, in: > S. Diehl (Ed.), Software Visualization, Dagstuhl Seminar, LNCS 2269, > Springer, Berlin, 2002, pp. 176-175. >=20 >=20 >=20 > =09 > _______________________________ > Do you Yahoo!? > Declare Yourself - Register online to vote today! > http://vote.yahoo.com >=20 >=20 > ------------------------------------------------------- > This SF.net email is sponsored by: IT Product Guide on ITManagersJourna= l > Use IT products in your business? Tell us what you think of them. Give = us > Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out = more > http://productguide.itmanagersjournal.com/guidepromo.tmpl > _______________________________________________ > Jdiseq-developer mailing list > Jdi...@li... > https://lists.sourceforge.net/lists/listinfo/jdiseq-developer |
From: Nascif Abousalh-N. <na...@ya...> - 2004-10-04 21:06:33
|
Found some interesting material on graph visualization today, check this out: http://prefuse.sourceforge.net/ (awesome, check the demos) [20] H. Eichelberger, J. Wolf von Gudenberg, On the Visualization of Java Programs, in: S. Diehl (Ed.), Software Visualization, Dagstuhl Seminar, LNCS 2269, Springer, Berlin, 2002, pp. 295-306. [21] K. Oechsle, T. Schmitt, JAVAVIS: Automatic Program Visualization with Object and Sequence Diagrams using the Java Debug Interface, in: S. Diehl (Ed.), Software Visualization, Dagstuhl Seminar, LNCS 2269, Springer, Berlin, 2002, pp. 176-175. _______________________________ Do you Yahoo!? Declare Yourself - Register online to vote today! http://vote.yahoo.com |
From: Orjan N. A. <aus...@co...> - 2004-10-04 19:15:09
|
:-) Trond Andersen wrote: >>The activity percentile on the project page seems to be active again :-) > > Finally - it really has annoyed you :-D > > ---- Trond > > > ------------------------------------------------------- > This SF.net email is sponsored by: IT Product Guide on ITManagersJournal > Use IT products in your business? Tell us what you think of them. Give us > Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out more > http://productguide.itmanagersjournal.com/guidepromo.tmpl > _______________________________________________ > Jdiseq-developer mailing list > Jdi...@li... > https://lists.sourceforge.net/lists/listinfo/jdiseq-developer |
From: Trond A. <tro...@gm...> - 2004-10-04 06:27:48
|
> The activity percentile on the project page seems to be active again :-) Finally - it really has annoyed you :-D ---- Trond |
From: Orjan N. A. <aus...@co...> - 2004-10-03 22:17:00
|
Added support for preferences handling, cool splash screen (please=20 comment on jDiSeq logo!), and initial ui code. =D8rjan |
From: Orjan N. A. <aus...@co...> - 2004-10-03 21:40:58
|
The activity percentile on the project page seems to be active again :-) =D8rjan |
From: Orjan N. A. <aus...@co...> - 2004-10-01 05:36:32
|
Hardfilters: Hardfiltering are filters handed to the JDI environment. Such filters=20 can only filter classes for inclusion/exclusion (threads will maybe be=20 added later). Once filtered, the event is lost forever. In addition,=20 hardfiltering have a direct impact on observation performance. Just=20 think about not filtering out method events on the Java core classes.=20 Hardfilters only support 3 variations for specification: full classname: org.jdiseq.Foo glob at start: *iseq.Foo glob at end: org.jdis* The runtime configuration file supports a notion of package/class=20 filtering for hardfilters: <set name=3D"cglib" packages=3D"net.sf.cglib,org.objectweb"/> <set name=3D"foo classes" classes=3D"org.jdiseq.Foo,org.jdiseq.FooBar"/> Softfilters: Softfilters are applied on sequences post record time, and gives the use=20 more fine grained control over filtering. In an interactive jDiSeq=20 session, softfilters can be changed to provide different views of a=20 sequence. Softfilters (should) provide filtering for classes, packages,=20 objects, threads methods, and possibly a combination of the above. It is=20 expected that softfilters will be used to "tune" a diagram once recorded. Regards, =D8rjan |
From: Orjan N. A. <aus...@co...> - 2004-10-01 05:12:04
|
Nascif Abousalh-Neto wrote: > When are you planning to tag 0.2 and start 0.3? Should we have a > roadmap for the desired features for each release? > Good question! I am open for discussion about when and what should be in the relase. Trond is currently writing on an initial roadmap document. I guess that the sooner we can deliver a minimum relase, the better. Relasing jDiSeq and hopefully gaining some attention will help us drive development in the right direction. I have had a discussion with Trond off-list about releasing a command line version only as 0.2. For this to happen, we do miss one important feature -- softfiltering. In addition, some minimal documentation as "Quick start guide" and configuration file documentation must be written. What do you think? > I have to do some research on how to handle the CVS commit - some > technical problems in my environment I have to address. > Maybe I could help. I struggled some until I figured out how to handle cvs on sf. Just send me a mail and I'll se what I can do. > Hey, one question: what is the difference between hard and soft > filters? > Answered in a separate posting for easer access when I have to write documentation on this :-) |
From: Nascif Abousalh-N. <na...@ya...> - 2004-10-01 04:13:18
|
> Does the maven plugin require a special format of the tag name? From what I saw in the code the only check they make is for HEAD; they explicitly skip a release tagged as HEAD when creating a binary distribution. When are you planning to tag 0.2 and start 0.3? Should we have a roadmap for the desired features for each release? I have to do some research on how to handle the CVS commit - some technical problems in my environment I have to address. Hey, one question: what is the difference between hard and soft filters? Regards, Nascif --- Orjan Nygaard Austvold <aus...@co...> wrote: > Good work! > > (comments inline below) > > Nascif Abousalh-Neto wrote: > > > > There is a problem though: we need to have a "closed" version to > offer > > as a download. The Maven xdoc plugin will not generate download > links > > for open versions, meaning versions with the CVS HEAD tag, as the > > current 0.2. I had to create a fake 0.1 to get the link creation to > > work. > > My though was that we will tag version 0.2 before building a release. > > Does the maven plugin require a special format of the tag name? > > > > > Also, how should I submit the changes? I am accessing CVS using > > anonymous login; not sure if I have commit privileges. In the mean > > time, I sent a patch so you guys can see the changes I made. > > > > I checked the jdiseq sf site, and you do have commit privilege. > > Side note: To make cvs mangling on sf as smooth as possible, I've > gone > through these few steps: > > Generated a keyset with ssh-keygen (dsa type of keys) > Uploaded the public key to sf on my user-admin page > > Whenever I want to work on jdiseq I simply do ssh-add (to add my key > to > my local agent), and start working on the module checked out with > CVSROOT as :ext:aus...@sf...:/cvsroot/jdiseq > > > > I added JCoverage in the reports, a really cool companion for > JUnit, > > but since there were zero JUnit testcases to run, there was no code > > coverage report... have you guys forgot to commit the unit > testcases? I > > can see a JUnit report on the official site. > > > > This is kind of my fault (more correctly -- it _is_ my fault :-) The > only tests that existed was those to test the previous filtering > mechanism made by Trond. Since that has been replaces by > hard/softfiltering, the tests weren't needed anymore. Now.... on all > the > code that I have checked in there is an average test coverage of > about > ......... 0%. Shame on me! I'll do better -- promise :-) > > > > Ørjan > > > > ------------------------------------------------------- > This SF.net email is sponsored by: IT Product Guide on > ITManagersJournal > Use IT products in your business? Tell us what you think of them. > Give us > Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find > out more > http://productguide.itmanagersjournal.com/guidepromo.tmpl > _______________________________________________ > Jdiseq-developer mailing list > Jdi...@li... > https://lists.sourceforge.net/lists/listinfo/jdiseq-developer > _______________________________ Do you Yahoo!? Declare Yourself - Register online to vote today! http://vote.yahoo.com |
From: Orjan N. A. <aus...@co...> - 2004-09-30 18:52:03
|
Good work! (comments inline below) Nascif Abousalh-Neto wrote: >=20 > There is a problem though: we need to have a "closed" version to offer > as a download. The Maven xdoc plugin will not generate download links > for open versions, meaning versions with the CVS HEAD tag, as the > current 0.2. I had to create a fake 0.1 to get the link creation to > work. My though was that we will tag version 0.2 before building a release.=20 Does the maven plugin require a special format of the tag name? >=20 > Also, how should I submit the changes? I am accessing CVS using > anonymous login; not sure if I have commit privileges. In the mean > time, I sent a patch so you guys can see the changes I made.=20 >=20 I checked the jdiseq sf site, and you do have commit privilege. Side note: To make cvs mangling on sf as smooth as possible, I've gone=20 through these few steps: Generated a keyset with ssh-keygen (dsa type of keys) Uploaded the public key to sf on my user-admin page Whenever I want to work on jdiseq I simply do ssh-add (to add my key to=20 my local agent), and start working on the module checked out with=20 CVSROOT as :ext:aus...@sf...:/cvsroot/jdiseq > I added JCoverage in the reports, a really cool companion for JUnit, > but since there were zero JUnit testcases to run, there was no code > coverage report... have you guys forgot to commit the unit testcases? I > can see a JUnit report on the official site. >=20 This is kind of my fault (more correctly -- it _is_ my fault :-) The=20 only tests that existed was those to test the previous filtering=20 mechanism made by Trond. Since that has been replaces by=20 hard/softfiltering, the tests weren't needed anymore. Now.... on all the=20 code that I have checked in there is an average test coverage of about=20 ......... 0%. Shame on me! I'll do better -- promise :-) =D8rjan |
From: Orjan N. A. <aus...@co...> - 2004-09-30 18:36:47
|
Nascif Abousalh-Neto wrote: > --- Orjan Nygaard Austvold <aus...@co...> wrote: >=20 >=20 >>Looking for files matching a given pattern could generate to many=20 >>targets to run, I think. Aslak Helles=F8y mentioned a year ago that an=20 >>xdoclet plugin for jDiSeq could be worth while creating. Imagine >>simply=20 >>documenting scenarios with filters from within the JavaDoc of the=20 >>methods. Generating live diagrams for these scenarios would be just=20 >>plain simple with a Maven plugin for jDiSeq. >=20 >=20 > Hmm, I don't agree. How would doclet tags solve the problem of creating > a harness to execute the code, manage the dependencies, etc.? Unless we > are talking about annotating the JUnit testcases that we want to use as > scenarios. >=20 Yes, you are right. Xdoclet will not solve the problem. But I do think=20 that it could be put to use, especially for generating=20 scenario-descriptions. > Instead of using patterns we could just create yet another source > tree. Maven currently assumes three: the main one, one for JUnit, one > for Aspects. This one would be specific for the jdiseq scenarios. We > could even use JUnit to drive the execution, and JUnit testcase method > names as triggers to start and stop the jdiSeq monitoring and > generation of diagrams. >=20 I don't understand this part. Why would you create unit tests to start=20 jDiSeq monitoring? Could this be done as launching a new vm from within=20 maven? > In that approach xdoclet could be used to annotate jdiseq specific > configuration info as part of the scenarios (now JUnit testcases); or > we could just add the jdiseq config files to the testcase directories. >=20 Ah. I didn't see this before now. I see that we see xDoclet usage the=20 same way. >=20 >=20 >>>Cool, I think a new task is a good idea. Regarding the release >> >>build, I >> >>>have two questions/concerns: >>>1) there is an overlap between the SourceForge support for file >>>download and the Maven support for release download (the dist plugin >>>creates a "download" link in the site). How are we going to handle >>>this? Use Maven to generate the release and then >>>manually/automatically(how?) push the file to SourceForge? >>>2)The dist plugin generates quite large archives because it adds all >>>the Maven generated documentation. Not a major problem, just so you >>>know. >>>=20 >>> >> >>Task for maven-plugin has been created and assigned to you. >> >>1) A download link on the site seems fine to me. Trond led me to a >>maven=20 >>sourceforge plugin which have goals for deploying files into the SF >>File=20 >>Release System. Maybe this can help? >> >> >=20 > http://maven-plugins.sourceforge.net/maven-sourceforge-plugin/index.htm= l >=20 >>2) This shouldn't be a problem I guess. Later on we can try to >>minimize=20 >>download size and the alike if this becomes an issue. >> >> >>>Actually, I was considering the UML class diagram support of this >> >>tool. >> >>>Just like jdiSeq, it extracts the info from the code (static instead >> >>of >> >>>dynamic analysis) and builds a diagram. So with UMLGraph and dotuml, >> >>a >> >>>project would have both class and sequence diagrams automatically >>>generated, as part of the maven output - not bad at all... :-) >>> >>>=20 >>> >> >>Ahh.. Now I understand what you are looking for. Execelent tool=20 >>combination :-) >> >>Btw: I still think that creating UMLGraph (text) file output by >>jDiSeq=20 >>could be interesting. I relly liked the layout on diagrams rendered >>by=20 >>pic2plot. Running everything with pipes could be fun: >> >>jdiseq | UMLGraph | pic2plot > my_cool_diagram.svg >> >> >>>I am working with a guy called Dominik on the umlgraph maven plugin. >> >>I >> >>>already have it working in my environment, we are trying to make it >>>"official" soon. I can send it to you guys if you want to take a >> >>look >> >>>on the overal plugin architecture, and how the report pages are >>>generated. >>> >>>=20 >>> >> >>Yes, thank you. Just send it by mail off this list so that it doesn't >> >>become official by accident. >> >> >> >>------------------------------------------------------- >>This SF.net email is sponsored by: IT Product Guide on >>ITManagersJournal >>Use IT products in your business? Tell us what you think of them. >>Give us >>Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find >>out more >>http://productguide.itmanagersjournal.com/guidepromo.tmpl >>_______________________________________________ >>Jdiseq-developer mailing list >>Jdi...@li... >>https://lists.sourceforge.net/lists/listinfo/jdiseq-developer >> >=20 >=20 >=20 >=20 > =09 > _______________________________ > Do you Yahoo!? > Declare Yourself - Register online to vote today! > http://vote.yahoo.com >=20 >=20 > ------------------------------------------------------- > This SF.net email is sponsored by: IT Product Guide on ITManagersJourna= l > Use IT products in your business? Tell us what you think of them. Give = us > Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out = more > http://productguide.itmanagersjournal.com/guidepromo.tmpl > _______________________________________________ > Jdiseq-developer mailing list > Jdi...@li... > https://lists.sourceforge.net/lists/listinfo/jdiseq-developer |
From: Orjan N. A. <aus...@co...> - 2004-09-29 23:55:33
|
http://callflow.sourceforge.net/ |
From: Orjan N. A. <aus...@co...> - 2004-09-29 23:32:33
|
/home/austvold/projects/jdiseq cvs commit -f -R -m "Removed JiBX configuration scheme. Introduced jakarta commons configuration. Refactored application launch (dependency injection pattern). Added pale splash screen." src/java/org/jdiseq/configuration/Configuration.java src/java/org/jdiseq/configuration/ConfigurationFactory.java src/java/org/jdiseq/configuration/TerminationCriteria.java src/java/org/jdiseq/filter/HardFilter.java src/java/org/jdiseq/recorder/Recording.java src/java/org/jdiseq/ui/SplashScreen.java src/java/org/jdiseq/ConfigurationException.java src/java/org/jdiseq/HeadLessLauncher.java src/java/org/jdiseq/JDiSeq.java src/java/org/jdiseq/Main.java src/resources/default-hardfilter.xml src/resources/default_hardfilter.xml example-configuration.xml maven.xml project.xml run.sh Checking in src/java/org/jdiseq/configuration/Configuration.java; /cvsroot/jdiseq/jdiseq/src/java/org/jdiseq/configuration/Configuration.java,v <-- Configuration.java new revision: 1.2; previous revision: 1.1 done Removing src/java/org/jdiseq/configuration/ConfigurationFactory.java; /cvsroot/jdiseq/jdiseq/src/java/org/jdiseq/configuration/ConfigurationFactory.java,v <-- ConfigurationFactory.java new revision: delete; previous revision: 1.1 done Checking in src/java/org/jdiseq/configuration/TerminationCriteria.java; /cvsroot/jdiseq/jdiseq/src/java/org/jdiseq/configuration/TerminationCriteria.java,v <-- TerminationCriteria.java new revision: 1.3; previous revision: 1.2 done Checking in src/java/org/jdiseq/filter/HardFilter.java; /cvsroot/jdiseq/jdiseq/src/java/org/jdiseq/filter/HardFilter.java,v <-- HardFilter.java new revision: 1.3; previous revision: 1.2 done Checking in src/java/org/jdiseq/recorder/Recording.java; /cvsroot/jdiseq/jdiseq/src/java/org/jdiseq/recorder/Recording.java,v <-- Recording.java new revision: 1.6; previous revision: 1.5 done RCS file: /cvsroot/jdiseq/jdiseq/src/java/org/jdiseq/ui/SplashScreen.java,v done Checking in src/java/org/jdiseq/ui/SplashScreen.java; /cvsroot/jdiseq/jdiseq/src/java/org/jdiseq/ui/SplashScreen.java,v <-- SplashScreen.java initial revision: 1.1 done Removing src/java/org/jdiseq/ConfigurationException.java; /cvsroot/jdiseq/jdiseq/src/java/org/jdiseq/ConfigurationException.java,v <-- ConfigurationException.java new revision: delete; previous revision: 1.1 done Checking in src/java/org/jdiseq/HeadLessLauncher.java; /cvsroot/jdiseq/jdiseq/src/java/org/jdiseq/HeadLessLauncher.java,v <-- HeadLessLauncher.java new revision: 1.5; previous revision: 1.4 done RCS file: /cvsroot/jdiseq/jdiseq/src/java/org/jdiseq/JDiSeq.java,v done Checking in src/java/org/jdiseq/JDiSeq.java; /cvsroot/jdiseq/jdiseq/src/java/org/jdiseq/JDiSeq.java,v <-- JDiSeq.java initial revision: 1.1 done Removing src/java/org/jdiseq/Main.java; /cvsroot/jdiseq/jdiseq/src/java/org/jdiseq/Main.java,v <-- Main.java new revision: delete; previous revision: 1.12 done RCS file: /cvsroot/jdiseq/jdiseq/src/resources/default-hardfilter.xml,v done Checking in src/resources/default-hardfilter.xml; /cvsroot/jdiseq/jdiseq/src/resources/default-hardfilter.xml,v <-- default-hardfilter.xml initial revision: 1.1 done Removing src/resources/default_hardfilter.xml; /cvsroot/jdiseq/jdiseq/src/resources/default_hardfilter.xml,v <-- default_hardfilter.xml new revision: delete; previous revision: 1.1 done Checking in example-configuration.xml; /cvsroot/jdiseq/jdiseq/example-configuration.xml,v <-- example-configuration.xml new revision: 1.5; previous revision: 1.4 done Checking in maven.xml; /cvsroot/jdiseq/jdiseq/maven.xml,v <-- maven.xml new revision: 1.4; previous revision: 1.3 done Checking in project.xml; /cvsroot/jdiseq/jdiseq/project.xml,v <-- project.xml new revision: 1.24; previous revision: 1.23 done Checking in run.sh; /cvsroot/jdiseq/jdiseq/run.sh,v <-- run.sh new revision: 1.4; previous revision: 1.3 done |
From: Nascif Abousalh-N. <na...@ya...> - 2004-09-29 19:59:49
|
--- Orjan Nygaard Austvold <aus...@co...> wrote: > Looking for files matching a given pattern could generate to many > targets to run, I think. Aslak Hellesøy mentioned a year ago that an > xdoclet plugin for jDiSeq could be worth while creating. Imagine > simply > documenting scenarios with filters from within the JavaDoc of the > methods. Generating live diagrams for these scenarios would be just > plain simple with a Maven plugin for jDiSeq. Hmm, I don't agree. How would doclet tags solve the problem of creating a harness to execute the code, manage the dependencies, etc.? Unless we are talking about annotating the JUnit testcases that we want to use as scenarios. Instead of using patterns we could just create yet another source tree. Maven currently assumes three: the main one, one for JUnit, one for Aspects. This one would be specific for the jdiseq scenarios. We could even use JUnit to drive the execution, and JUnit testcase method names as triggers to start and stop the jdiSeq monitoring and generation of diagrams. In that approach xdoclet could be used to annotate jdiseq specific configuration info as part of the scenarios (now JUnit testcases); or we could just add the jdiseq config files to the testcase directories. > > >Cool, I think a new task is a good idea. Regarding the release > build, I > >have two questions/concerns: > >1) there is an overlap between the SourceForge support for file > >download and the Maven support for release download (the dist plugin > >creates a "download" link in the site). How are we going to handle > >this? Use Maven to generate the release and then > >manually/automatically(how?) push the file to SourceForge? > >2)The dist plugin generates quite large archives because it adds all > >the Maven generated documentation. Not a major problem, just so you > >know. > > > > > Task for maven-plugin has been created and assigned to you. > > 1) A download link on the site seems fine to me. Trond led me to a > maven > sourceforge plugin which have goals for deploying files into the SF > File > Release System. Maybe this can help? > > http://maven-plugins.sourceforge.net/maven-sourceforge-plugin/index.html > > 2) This shouldn't be a problem I guess. Later on we can try to > minimize > download size and the alike if this becomes an issue. > > >Actually, I was considering the UML class diagram support of this > tool. > >Just like jdiSeq, it extracts the info from the code (static instead > of > >dynamic analysis) and builds a diagram. So with UMLGraph and dotuml, > a > >project would have both class and sequence diagrams automatically > >generated, as part of the maven output - not bad at all... :-) > > > > > > > Ahh.. Now I understand what you are looking for. Execelent tool > combination :-) > > Btw: I still think that creating UMLGraph (text) file output by > jDiSeq > could be interesting. I relly liked the layout on diagrams rendered > by > pic2plot. Running everything with pipes could be fun: > > jdiseq | UMLGraph | pic2plot > my_cool_diagram.svg > > >I am working with a guy called Dominik on the umlgraph maven plugin. > I > >already have it working in my environment, we are trying to make it > >"official" soon. I can send it to you guys if you want to take a > look > >on the overal plugin architecture, and how the report pages are > >generated. > > > > > > > Yes, thank you. Just send it by mail off this list so that it doesn't > > become official by accident. > > > > ------------------------------------------------------- > This SF.net email is sponsored by: IT Product Guide on > ITManagersJournal > Use IT products in your business? Tell us what you think of them. > Give us > Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find > out more > http://productguide.itmanagersjournal.com/guidepromo.tmpl > _______________________________________________ > Jdiseq-developer mailing list > Jdi...@li... > https://lists.sourceforge.net/lists/listinfo/jdiseq-developer > _______________________________ Do you Yahoo!? Declare Yourself - Register online to vote today! http://vote.yahoo.com |
From: Nascif Abousalh-N. <na...@ya...> - 2004-09-29 19:43:46
|
I got the creation of binary distribution working, though they turned out to be to big to send as an attachment (around 4M): -rwxrwxrwx 1 naabou 4045518 Sep 29 15:27 jdiseq-0.2.tar.gz* -rwxrwxrwx 1 naabou 4383226 Sep 29 15:27 jdiseq-0.2.zip* The attached screenshot show the Download page at the Maven site. There is a problem though: we need to have a "closed" version to offer as a download. The Maven xdoc plugin will not generate download links for open versions, meaning versions with the CVS HEAD tag, as the current 0.2. I had to create a fake 0.1 to get the link creation to work. Also, how should I submit the changes? I am accessing CVS using anonymous login; not sure if I have commit privileges. In the mean time, I sent a patch so you guys can see the changes I made. I added JCoverage in the reports, a really cool companion for JUnit, but since there were zero JUnit testcases to run, there was no code coverage report... have you guys forgot to commit the unit testcases? I can see a JUnit report on the official site. I also added a changes.xml under xdocs to track the changes between versions. Regards, Nascif --- Orjan Nygaard Austvold <aus...@co...> wrote: > Nascif Abousalh-Neto wrote: > > >jdiSeq as a maven plugin would be basically a documentation tool, so > it > >should generate reports as the other plugins. I think something like > a > >JUnit report; jdiSeq would scan the source tree for all file > matching a > >given pattern (maybe in the class name, like JUnit does), execute > those > >classes, capture the sequence diagrams in a web-compatible format > (png, > >svg) and generate a web page linking to those diagrams. > > > > > > > Trond had an idea on using a container launcher (he mentioned a > codehaus > project) for launching the observed vm. Maybe this could be used in > conjuntion with unit testing? > > Looking for files matching a given pattern could generate to many > targets to run, I think. Aslak Hellesøy mentioned a year ago that an > xdoclet plugin for jDiSeq could be worth while creating. Imagine > simply > documenting scenarios with filters from within the JavaDoc of the > methods. Generating live diagrams for these scenarios would be just > plain simple with a Maven plugin for jDiSeq. > > >Cool, I think a new task is a good idea. Regarding the release > build, I > >have two questions/concerns: > >1) there is an overlap between the SourceForge support for file > >download and the Maven support for release download (the dist plugin > >creates a "download" link in the site). How are we going to handle > >this? Use Maven to generate the release and then > >manually/automatically(how?) push the file to SourceForge? > >2)The dist plugin generates quite large archives because it adds all > >the Maven generated documentation. Not a major problem, just so you > >know. > > > > > Task for maven-plugin has been created and assigned to you. > > 1) A download link on the site seems fine to me. Trond led me to a > maven > sourceforge plugin which have goals for deploying files into the SF > File > Release System. Maybe this can help? > > http://maven-plugins.sourceforge.net/maven-sourceforge-plugin/index.html > > 2) This shouldn't be a problem I guess. Later on we can try to > minimize > download size and the alike if this becomes an issue. > > >Actually, I was considering the UML class diagram support of this > tool. > >Just like jdiSeq, it extracts the info from the code (static instead > of > >dynamic analysis) and builds a diagram. So with UMLGraph and dotuml, > a > >project would have both class and sequence diagrams automatically > >generated, as part of the maven output - not bad at all... :-) > > > > > > > Ahh.. Now I understand what you are looking for. Execelent tool > combination :-) > > Btw: I still think that creating UMLGraph (text) file output by > jDiSeq > could be interesting. I relly liked the layout on diagrams rendered > by > pic2plot. Running everything with pipes could be fun: > > jdiseq | UMLGraph | pic2plot > my_cool_diagram.svg > > >I am working with a guy called Dominik on the umlgraph maven plugin. > I > >already have it working in my environment, we are trying to make it > >"official" soon. I can send it to you guys if you want to take a > look > >on the overal plugin architecture, and how the report pages are > >generated. > > > > > > > Yes, thank you. Just send it by mail off this list so that it doesn't > > become official by accident. > > > > ------------------------------------------------------- > This SF.net email is sponsored by: IT Product Guide on > ITManagersJournal > Use IT products in your business? Tell us what you think of them. > Give us > Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find > out more > http://productguide.itmanagersjournal.com/guidepromo.tmpl > _______________________________________________ > Jdiseq-developer mailing list > Jdi...@li... > https://lists.sourceforge.net/lists/listinfo/jdiseq-developer > _______________________________ Do you Yahoo!? Declare Yourself - Register online to vote today! http://vote.yahoo.com |
From: Orjan N. A. <aus...@co...> - 2004-09-29 10:38:12
|
Just wanted to let you know that we (yes _we_) are using jDiSeq at work=20 as we speak. We have a couple of scenarios involving user session handling with=20 callback which we are trying to understand. We're using batik-squiggle to view diagrams generated by jDiSeq. Even=20 though squiggle is kind of sluggish. It works! Cool! Right now I'm really missing two features: * Softfiltering. The hardcoded stuff in the zanthan render is not enough. * "diagram nesting". Some way of specifying that a method called event=20 within a sequence should be extracted as a new diagram, possibly with=20 some kind of link from the called diagram. =D8rjan |
From: Trond A. <tro...@gm...> - 2004-09-29 08:59:25
|
> SVG is cool, svg is fast... Batik made it possible :-) >=20 > Took me less than 20 minutes to write a new renderer to render the > sequence to SVG. The rendering itself was sleek. Attached is the first > document generated (forgot to set some colors). Excellent =D8rjan. That was extremely fast. I'm downloading all new source today and try playing around with it tonight. I've been thinking on the roadmap stuff, but I think I will have to see what has been done before starting to use the feature request on the sf.net site as a basis for the roadmap. Me and =D8rjan concluded that we should try using the feature request part as the roadmap tool - similar as being done in Jira. For information: I'm struggeling on getting my wireless card working on Linux where I have my development environment. ---- Trond |
From: Orjan N. A. <aus...@co...> - 2004-09-28 17:25:03
|
Nascif Abousalh-Neto wrote: >jdiSeq as a maven plugin would be basically a documentation tool, so it >should generate reports as the other plugins. I think something like a >JUnit report; jdiSeq would scan the source tree for all file matching a >given pattern (maybe in the class name, like JUnit does), execute those >classes, capture the sequence diagrams in a web-compatible format (png, >svg) and generate a web page linking to those diagrams. > > =20 > Trond had an idea on using a container launcher (he mentioned a codehaus=20 project) for launching the observed vm. Maybe this could be used in=20 conjuntion with unit testing? Looking for files matching a given pattern could generate to many=20 targets to run, I think. Aslak Helles=F8y mentioned a year ago that an=20 xdoclet plugin for jDiSeq could be worth while creating. Imagine simply=20 documenting scenarios with filters from within the JavaDoc of the=20 methods. Generating live diagrams for these scenarios would be just=20 plain simple with a Maven plugin for jDiSeq. >Cool, I think a new task is a good idea. Regarding the release build, I >have two questions/concerns: >1) there is an overlap between the SourceForge support for file >download and the Maven support for release download (the dist plugin >creates a "download" link in the site). How are we going to handle >this? Use Maven to generate the release and then >manually/automatically(how?) push the file to SourceForge? >2)The dist plugin generates quite large archives because it adds all >the Maven generated documentation. Not a major problem, just so you >know. > =20 > Task for maven-plugin has been created and assigned to you. 1) A download link on the site seems fine to me. Trond led me to a maven=20 sourceforge plugin which have goals for deploying files into the SF File=20 Release System. Maybe this can help? http://maven-plugins.sourceforge.net/maven-sourceforge-plugin/index.html 2) This shouldn't be a problem I guess. Later on we can try to minimize=20 download size and the alike if this becomes an issue. >Actually, I was considering the UML class diagram support of this tool. >Just like jdiSeq, it extracts the info from the code (static instead of >dynamic analysis) and builds a diagram. So with UMLGraph and dotuml, a >project would have both class and sequence diagrams automatically >generated, as part of the maven output - not bad at all... :-) > > =20 > Ahh.. Now I understand what you are looking for. Execelent tool=20 combination :-) Btw: I still think that creating UMLGraph (text) file output by jDiSeq=20 could be interesting. I relly liked the layout on diagrams rendered by=20 pic2plot. Running everything with pipes could be fun: jdiseq | UMLGraph | pic2plot > my_cool_diagram.svg >I am working with a guy called Dominik on the umlgraph maven plugin. I >already have it working in my environment, we are trying to make it >"official" soon. I can send it to you guys if you want to take a look >on the overal plugin architecture, and how the report pages are >generated. > > =20 > Yes, thank you. Just send it by mail off this list so that it doesn't=20 become official by accident. |
From: Orjan N. A. <aus...@co...> - 2004-09-28 15:39:12
|
SVG is cool, svg is fast... Batik made it possible :-) Took me less than 20 minutes to write a new renderer to render the=20 sequence to SVG. The rendering itself was sleek. Attached is the first=20 document generated (forgot to set some colors). =D8rjan Nascif Abousalh-Neto wrote: >This matches my previous experience with "real-life" projects; it is >very easy to reach those numbers without some sort of >filtering/aggregation going on. You just need to be using some kind of >medium-large data processing, or be using a framework. Both conditions >are quite common in enterprise software. > >There is another option for the visualization, though, that may scale >better[1]: generate SVG instead of PNG. I believe that the PngCreator >is based on Java 2D; if that is the case, there is this article talking >about creating SVG output from Java2D API calls. Not sure how well a >SVG viewer would handle a diagram this big, but since SVG supports >zooming, it should be more robust. This is the link to the google >cache, the original is not available anymore: >http://216.239.41.104/search?q=3Dcache:9jTwROPJAIoJ:www.sun.com/software= /xml/developers/svg/java2d-api/+java2d+svg&hl=3Den > >Come on, Orjan! Another sleepless night and jdiSeq will have its >SvgCreator.java! ;-) > >[1] At least from a pure memory/size point of view; the diagram might >still be too big to be used as an effective documentation/analysis >tool. > >--- Orjan Nygaard Austvold <aus...@co...> wrote: > > =20 > >>Couldn't sleep so I found the bug in SEQUENCE PngCreator.java He >>simply=20 >>forgot to upgrade the parser implementation to the newer version :-) >> >>I've attached the first working example of jDiSeq generation png's. I >> >>actually documented the startup of SEQUENCE just for fun. >> >>Btw: I have sucessfully created long (500-600 method calls) sequence=20 >>records without any hazzle, but as soon as I tried to do rendering to >> >>png, the vm used over 300MB memory and took 10 minutes to put >>everything=20 >>to file. The resulting file were about 12000x18000x pixels and 3MB >>big.=20 >>We need to look into a smarter way to render graphics objects. >> >>Here's the command lines I used for rendering of the attached >>diagram: >> >>jDiSeq: >>[austvold@localhost jdiseq]$ ./run.sh --headless >>example-configuration.xml >> >>SEQUENCE: >>java -Xdebug=20 >>-Xrunjdwp:transport=3Ddt_socket,address=3D8000,server=3Dy,suspend=3Dy >>-classpath=20 >> >> =20 >> >~/.maven/repository/log4j/jars/log4j-1.2.8.jar:lib/sequence/jars/sequenc= e-20040120.jar > =20 > >>com.zanthan.sequence.Main >> >> >>=D8rjan >> >> >> =20 >> > > =20 > >>ATTACHMENT part 2 image/png name=3DMain.main.20040928021058.png >> =20 >> > > > > > =09 >_______________________________ >Do you Yahoo!? >Declare Yourself - Register online to vote today! >http://vote.yahoo.com > =20 > |
From: Nascif Abousalh-N. <na...@ya...> - 2004-09-28 02:32:16
|
Hi, --- Orjan Nygaard Austvold <aus...@co...> wrote: > Creating a plugin wrapper would really be cool. Should we do anything > on the "jdiseq-side" to support this? > Will it be sufficient for a plugin > to launch the jdiseq with a configuration file given on the command > line? I need to think more about it. We would probably want to support as much as possible (or feasible) from the configuration from properties; this would allow for better integration with Maven (including inhertitance of configuration, which is a really nice feature). But just calling the CLI with a config file is a great way to start - but there is also the integration with the rest of the reports. jdiSeq as a maven plugin would be basically a documentation tool, so it should generate reports as the other plugins. I think something like a JUnit report; jdiSeq would scan the source tree for all file matching a given pattern (maybe in the class name, like JUnit does), execute those classes, capture the sequence diagrams in a web-compatible format (png, svg) and generate a web page linking to those diagrams. > Do you want me to create a task for this on the sf-site? I've updated > the maven build system task with info on adding release build > support. Cool, I think a new task is a good idea. Regarding the release build, I have two questions/concerns: 1) there is an overlap between the SourceForge support for file download and the Maven support for release download (the dist plugin creates a "download" link in the site). How are we going to handle this? Use Maven to generate the release and then manually/automatically(how?) push the file to SourceForge? 2)The dist plugin generates quite large archives because it adds all the Maven generated documentation. Not a major problem, just so you know. > Just looked at UMLGraph. If I understand it correctly, UMLGraph takes > a descriptive file as input and generates output suitable for > renderering by pic2plot. > It shouldn't be too hard to create a renderer (like the > Zanthan renderer I made) for the UMLGraph notiation. Actually, I was considering the UML class diagram support of this tool. Just like jdiSeq, it extracts the info from the code (static instead of dynamic analysis) and builds a diagram. So with UMLGraph and dotuml, a project would have both class and sequence diagrams automatically generated, as part of the maven output - not bad at all... :-) I am working with a guy called Dominik on the umlgraph maven plugin. I already have it working in my environment, we are trying to make it "official" soon. I can send it to you guys if you want to take a look on the overal plugin architecture, and how the report pages are generated. Regards, Nascif > > A simple pipe could do. If jDiSeq sent rendering to stdout, then > UMLGraph and pic2plot could do the rest. Cool! > > > Ørjan > > > > Nascif Abousalh-Neto wrote: > > >I know how to do a binary distribution, I recently did one for my > work > >project. I can look into that. > > > >Since the current release focus on a command-line interface, how > about > >a maven plugin wrapper to run jdiseq as part of a maven build? I > think > >it makes sense (generate up-to-date documentation), although a GUI, > >interative UI would also be useful. It would make a nice pair to > >dotuml, the maven plugin for UMLGraph, soon to be released... ;-) > > > >Regards, > > Nascif > > > >--- Orjan Nygaard Austvold <aus...@co...> wrote: > > > > > > > >>Hi Nascif. Good to hear from you to! > >> > >>It would be great to get some help on maven. There are some minor > >>things > >>I would like to see changed with the current maven build (ordered > by > >>priority): > >> > >>1. Build target for binary release distribution(s). I've noticted > >>that > >>there's a release plugin for maven. From the docs at > >>http://maven.apache.org/reference/plugins/release/ it seems to be > >>able > >>to pick out dependent jars for release distribiution by looking at > a > >>"special" dependecy property. Did not figure out how this works > >>though. > >>For now we could do by simply creating a release directory > >>containing: > >> > >>bin/jdiseq <- simple launch script > >>lib/jdiseq-0.2.jar > >>lib/jibx-run-1.0b3c.jar > >>lib/xpp3-1.1.3.3.jar > >>lib/log4j-1.2.8.jar > >>lib/commons-logging-api-1.0.4.jar > >>lib/commons-lang-2.0.jar > >>lib/sequence-20040120.jar > >>licenses/jibx.license.txt > >>licenses/sequence.license.txt > >>licenses/xpp3.license.txt > >>LICENSE.txt > >>NOTICE.txt > >>README.txt > >>examples/ <- some example configurations > >> > >>2. Application launch target during development (I'm using a run.sh > > >>script as of now) > >> > >>3. Extracting JiBX binding to a plug-in (The JiBX files in lib > should > >> > >>also be published to the maven repository) > >> > >>I will reorganize the task list on the sourceforge site and add 1) > as > >>a > >>task. Maybe 2) and 3) are RFE's? > >> > >> > >>Ørjan > >> > >> > >> > >>Nascif Abousalh-Neto wrote: > >> > >> > >> > >>>Hi Orjan, good to hear from you. > >>> > >>>In the mean time I started to use maven, and even played with some > >>>plugins, so I might be able to help on that front. > >>> > >>>I will try to check the new code and get back to you. > >>> > >>>Regards, > >>>Nascif > >>> > >>>--- Orjan Nygaard Austvold <aus...@co...> wrote: > >>> > >>> > >>> > >>> > >>> > >>>>Ok, so now the optional png-task mentioned below is more likely > to > >>>> > >>>> > >>be > >> > >> > >>>>realized in the first relase :-) I was browsing the new source > for > >>>>SEQUENCE and found > >>>> > >>>>src/com/zanthan/sequence/headless/PngCreator.java > >>>> > >>>>I'll look into how this works. By the way: I've talked with Trond > >>>>about > >>>>the possibility to use simpleUML (plugin for IntelliJ) to render > >>>>diagrams. I just found out (on the plugin site) that the plugin > >>>>actually > >>>>uses the sequence code for rendering sequence diagrams. Alex > >>>> > >>>> > >>Moffat > >> > >> > >>>>(author of sequence) have done a fine job! > >>>> > >>>> > >>>>Ørjan > >>>> > >>>> > >>>> > >>>>Orjan Nygaard Austvold wrote: > >>>> > >>>> > >>>> > >>>> > >>>> > >>>>>During the last couple of weeks I've commited new and some > changed > >>>>> > >>>>> > >>>>>functionality to jDiSeq. > >>>>> > >>>>>As of now the tool is actually usable in a headless kind of > mode. > >>>>> > >>>>>Users create a configuration file describing filters, actors and > > >>>>>possibly termination criterias, run the tool, and get cool > >>>>> > >>>>> > >>sequence > >> > >> > >>>>> > >>>>> > >>>>>diagram files in plain text :-) Such file can be opened manually > >>>>> > >>>>> > >>in > >> > >> > >>>>> > >>>>> > >>>>>the sequence tool for rendering. > >>>>> > >>>>>Could the tool be released in the current state? Nope :-) > >>>>> > >>>>>A few (small) tasks must be addressed first: > >>>>> > >>>>>1. Remove hard-coded soft-filtering for the ZanthanSequence > >>>>> > >>>>> > >>>>> > >>>>> > >>>>renderer. > >>>> > >>>> > >>>> > >>>> > >>>>>2. Write xdocs describing tool usage (setup and configuration) > >>>>>3. Configure maven build to generate a distro containing > dependent > >>>>> > >>>>> > >>>>> > >>>>> > >>>>jar > >>>> > >>>> > >>>> > >>>> > >>>>>files docs and more. > >>>>> > >>>>>Optionally, we could try to generate png's instead of sequence > >>>>> > >>>>> > >>>>> > >>>>> > >>>>files > >>>> > >>>> > >>>> > >>>> > >>>>>as output. > >>>>> > >>>>> > >>>>>Ørjan > >>>>> > >>>>> > >>>>> > >>>>>------------------------------------------------------- > >>>>>This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of > 170 > >>>>>Project Admins to receive an Apple iPod Mini FREE for your > >>>>> > >>>>> > >>>>> > >>>>> > >>>>judgement on > >>>> > >>>> > >>>> > >>>> > >>>>>who ports your project to Linux PPC the best. Sponsored by IBM. > >>>>>Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php > >>>>>_______________________________________________ > >>>>>Jdiseq-developer mailing list > >>>>>Jdi...@li... > >>>>>https://lists.sourceforge.net/lists/listinfo/jdiseq-develope > >>>>>r > >>>>> > >>>>> > >>>>> > >>>>> > >>>>------------------------------------------------------- > >>>>This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of > 170 > >>>>Project Admins to receive an Apple iPod Mini FREE for your > >>>> > >>>> > >>judgement > >> > >> > >>>>on > >>>>who ports your project to Linux PPC the best. Sponsored by IBM. > >>>>Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php > >>>>_______________________________________________ > >>>>Jdiseq-developer mailing list > >>>>Jdi...@li... > >>>>https://lists.sourceforge.net/lists/listinfo/jdiseq-developer > >>>> > >>>> > >>>> > >>>> > >>>> > >>> > >>> > >>>_______________________________ > >>>Do you Yahoo!? > >>>Declare Yourself - Register online to vote today! > >>>http://vote.yahoo.com > >>> > >>> > >>> > >>> > >> > >>------------------------------------------------------- > >>This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170 > >>Project Admins to receive an Apple iPod Mini FREE for your > judgement > >>on > >>who ports your project to Linux PPC the best. Sponsored by IBM. > >>Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php > >>_______________________________________________ > >>Jdiseq-developer mailing list > >>Jdi...@li... > >>https://lists.sourceforge.net/lists/listinfo/jdiseq-developer > >> > >> > >> > > > > > > > > > >__________________________________ > >Do you Yahoo!? > >Yahoo! Mail Address AutoComplete - You start. We finish. > >http://promotions.yahoo.com/new_mail > > > > > >------------------------------------------------------- > >This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170 > >Project Admins to receive an Apple iPod Mini FREE for your judgement > on > >who ports your project to Linux PPC the best. Sponsored by IBM. > >Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php > >_______________________________________________ > >Jdiseq-developer mailing list > >Jdi...@li... > >https://lists.sourceforge.net/lists/listinfo/jdiseq-developer > > > > > > > __________________________________ Do you Yahoo!? Yahoo! Mail - 50x more storage than other providers! http://promotions.yahoo.com/new_mail |
From: Nascif Abousalh-N. <na...@ya...> - 2004-09-28 02:20:01
|
This matches my previous experience with "real-life" projects; it is very easy to reach those numbers without some sort of filtering/aggregation going on. You just need to be using some kind of medium-large data processing, or be using a framework. Both conditions are quite common in enterprise software. There is another option for the visualization, though, that may scale better[1]: generate SVG instead of PNG. I believe that the PngCreator is based on Java 2D; if that is the case, there is this article talking about creating SVG output from Java2D API calls. Not sure how well a SVG viewer would handle a diagram this big, but since SVG supports zooming, it should be more robust. This is the link to the google cache, the original is not available anymore: http://216.239.41.104/search?q=cache:9jTwROPJAIoJ:www.sun.com/software/xml/developers/svg/java2d-api/+java2d+svg&hl=en Come on, Orjan! Another sleepless night and jdiSeq will have its SvgCreator.java! ;-) [1] At least from a pure memory/size point of view; the diagram might still be too big to be used as an effective documentation/analysis tool. --- Orjan Nygaard Austvold <aus...@co...> wrote: > Couldn't sleep so I found the bug in SEQUENCE PngCreator.java He > simply > forgot to upgrade the parser implementation to the newer version :-) > > I've attached the first working example of jDiSeq generation png's. I > > actually documented the startup of SEQUENCE just for fun. > > Btw: I have sucessfully created long (500-600 method calls) sequence > records without any hazzle, but as soon as I tried to do rendering to > > png, the vm used over 300MB memory and took 10 minutes to put > everything > to file. The resulting file were about 12000x18000x pixels and 3MB > big. > We need to look into a smarter way to render graphics objects. > > Here's the command lines I used for rendering of the attached > diagram: > > jDiSeq: > [austvold@localhost jdiseq]$ ./run.sh --headless > example-configuration.xml > > SEQUENCE: > java -Xdebug > -Xrunjdwp:transport=dt_socket,address=8000,server=y,suspend=y > -classpath > ~/.maven/repository/log4j/jars/log4j-1.2.8.jar:lib/sequence/jars/sequence-20040120.jar > > com.zanthan.sequence.Main > > > Ørjan > > > ATTACHMENT part 2 image/png name=Main.main.20040928021058.png _______________________________ Do you Yahoo!? Declare Yourself - Register online to vote today! http://vote.yahoo.com |
From: Orjan N. A. <aus...@co...> - 2004-09-28 00:38:16
|
Couldn't sleep so I found the bug in SEQUENCE PngCreator.java He simply=20 forgot to upgrade the parser implementation to the newer version :-) I've attached the first working example of jDiSeq generation png's. I=20 actually documented the startup of SEQUENCE just for fun. Btw: I have sucessfully created long (500-600 method calls) sequence=20 records without any hazzle, but as soon as I tried to do rendering to=20 png, the vm used over 300MB memory and took 10 minutes to put everything=20 to file. The resulting file were about 12000x18000x pixels and 3MB big.=20 We need to look into a smarter way to render graphics objects. Here's the command lines I used for rendering of the attached diagram: jDiSeq: [austvold@localhost jdiseq]$ ./run.sh --headless example-configuration.x= ml SEQUENCE: java -Xdebug=20 -Xrunjdwp:transport=3Ddt_socket,address=3D8000,server=3Dy,suspend=3Dy -cl= asspath=20 ~/.maven/repository/log4j/jars/log4j-1.2.8.jar:lib/sequence/jars/sequence= -20040120.jar=20 com.zanthan.sequence.Main =D8rjan |
From: Orjan N. A. <aus...@co...> - 2004-09-27 22:04:16
|
I tried to set up commons configuration but failed. Partly due to either=20 me not understanding how to go about configuring the thing, or lack of=20 support in the API. Commons configuration seems to require a configuration file (!) on which=20 configuration files to be included must be described. I can't see how=20 this would fit the requirement of letting the user specify a file as a=20 command line argument. We could copy til file given on the command line to a well-known name on=20 tmp, but this seems like a hack to me. Well, all in all - I attempted to add two features to jDiSeq tonight,=20 and failed on both ;-) =D8rjan |