The Goldfish Chronicles

Notable

33965 [lang] Can’t XMLDecode an Enum

An interesting one - it’s referring to the Lang version of an Enum rather than the version in Java in later JDKs. Appears to be only on JDK 1.5 - by the sound of it it’s related to another Lang bug to do with a change in how classes are getting initialised in 1.5, issue #36666.

36157 [VFS] post 1.0, add get rid of deprecation warnings in FT…

The beginning of a VFS-fest tonight. This one involves a change to not use the deprecated API of Commons Net.

37991 [vfs] Local filesystem provider returns 0 from URLConnect…

Missing implementation.

38122 [vfs] getChildren() faild on missing folder access rights

Missing implementation.

38491 [vfs] / logs warning instead of throwing …

Mario added a failonerror attribute.

38659 [VFS] Maven repository SNAPSHOT upgrade

Comment discussion on how to get the jar into the snapshot repository.

38688 [net][PATCH] TelnetInputStream.read sometimes hangs if re…

Rob Hasselbaum is doing a great job dealing with a bug on a mature product - [net] doesn’t have much in the way of development so it’s hard to get an active conversation going on the matter, but Rob has pushed on and submitted a patch to his original bug report.

38846 [VFS] New provider: iso9660

Nice catch - a patch offered which depends on loopy, an LGPL library on sourceforge. Not something we can do currently.

38847 [VFS] New provider: crypt

Good questions from Mario - why does the patch have copyright statements in and does anyone else hold a copyright on the code. I’ve bumped it up a level by asking that we go further and get the reporter to provide a software grant. It’s painful - Robert’s currently undergoing pain discovering how to handle this with Commons Pool - but it’s something we need to do if we’re going to try and exist in a way that works with the law (rather than ignoring it).

38929 [configuration] Support event listeners for configurations

More discussion on this enhancement - always a bit sucky to have discusions like this in the issue tracker, doesn’t seem to work as well as the list; but frankly I think it’s just good that there’s activity.

38935 [VFS] remove commons-collections dependency from ram prov…

Always good to have less dependencies on a component. Some comments on @author tags too - which I stumbled into with a half-awake response.

38959 [configuration] Drop 1st class dependency on commons-logging

More discussion here too.

New

38987 [dbcp] commons dbcp does not supports Firebird DB.

Most peculiar, didn’t know [dbcp] had specific support for databases, I assumed it just sat on top of JDBC. The user has found the issue via DB Torque.

38992 [daemon] JSVC does not support all the standard java 5.0 launcher …

Interesting to see what response there is on this one - not sure how active [daemon] is - which is no indication of inactivity, I thought [JEXL] was inactive and it turned out they’d been busily being active for the last couple of months. Issue is asking for a few more parameters to be passable to the JVM.

Summary

The story of the day is [vfs] activity. It will be interesting to see if the [dbcp], [daemon] and [net] issues surface again - are those components active or slumbering.

Notable

38659 [VFS] Maven repository SNAPSHOT upgrade

Request for a new snapshot on the ibiblio maven repo. This is something which we need to get a bit more organized about - shouldn’t have snapshots going to ibiblio.

New

38959 [configuration] Drop 1st class dependency on commons-logging

Suggestion from Jorg Schaible (who’s been very active in Commons recently) to drop the logging dependency from configuration (and beanutils I think). Not sure I understand yet, seems to be a lot of logging usage.

38963 [dbutils] BeanListHandler and BeanHandler fail to support…

A dbutils fix for java.sql.Date support. Seems worth applying.

Summary

Quiet day, wonder why?

Notable

38894 [digester] ParserFeatureSetterFactory uses a wrong method…

Simon Kitching has resolved a bug in Digester - well half resolved, the less useful, more painful half has been documented in the javadoc. It was one of those lovely issues when XML implementations clash.

38907 [dbutils] ScalarHandler subclasses that return Integers a…

Greg Hawkes suggests an enhancement to DbUtils (Int and Long ScalarHandlers). There’s not a lot of maintenance going on in DbUtils at the moment - not much since February 2005, so I’m not sure when this would get looked at. It’s one I’d like to find time for, though I’d also like to investigate Brian McCallister’s JDBI API - get him to bring it to db.apache.org (he’s the chair) and fold Commons DbUtils into it.

38911 [lang] StringUtils#getLevenshteinDistance() performance i…

This was cool. Cedrik Lime pointed out an unapplied patch from October 2003 - and now it’s applied.

38927 [io] Tests Fail on Windows 2000

Few more comments on Niall’s Windows IO problems. Nothing to report though.

38929 [configuration] Support event listeners for configurations

This was a new enhancement suggestion yesterday - and today there were some comments on the idea. All quite positive.

38944 [email] Please add Email.setReplyTo( java.util.Collection…

dIon went ahead and added the missing method.

New

38953 [configuration] clone() implementation for all configurat…

A request from the mailing list for configuration to have a systematic approach to cloning the various configuration classes. Sounds like it’ll happen.

Summary

Positive - except for the unlikelyness of any activity on the [dbutils] issue (and other ones in Bugzilla).

Notable

33591 [dbcp][PATCH] Connection leak in PoolableConnection.close…

Sandy and Phil knocking off another issue on the way to a [dbcp] release.

36799 [dbcp] Docs inaccurately describe default value of testOn…

Javadoc improvement.

37090 [dbcp] Old Xerces version in project.xml

Some specific XML dependencies that have been yanked from the Maven build.

37359 [PATCH] SharedPoolDataSource - close() method closes poo…

And lastly, a fourth patch applied by the [dbcp] juggernaut.

38927 [io] Tests Fail on Windows 2000

Niall’s ended his invetigations into build problems on Windows 2000.

New

38936 [fileupload] FileUploadBase::parseRequest has to be more …

Ken Yee suggests an enhancement to [fileupload] to make it do as much as possible in the face of errors.

38939 [email] Errors when sending MultiPartEmail with another e…

Dave Cherkassky reports a bug in [email].

38942 [io] LockableFileWriter issues

Stephen reports and fixes a bug in [io]. This is a pattern we’ve come to refer to as ‘internal affairs’ at work

38944 [email] Please add Email.setReplyTo( java.util.Collection…

Dave Cherkassky also suggests an enhancement to the [email] API.

Summary

[dbcp] charging on is the message of the day.

Notable

29716 [configuration] Disabling string splitting

Housekeeping for a bug that was fixed months ago via another issue.

33591 [dbcp][PATCH] Connection leak in PoolableConnection.close…

Sandy McArthur, a new Commons committer has gone ahead and applied James Ring’s patches to fix this one.

37359 [PATCH] SharedPoolDataSource - close() method closes poo…

Phil applied a fix to this one. Another [dbcp] issue resolved.

38073 [dbcp] JNDI problems with SharedPoolDataSource

Again with the [dbcp], Phil applied a patch from Sandy, which was a modification of an original patch from Thomas Fischer. Listening to the mailing list, it sounds like a [dbcp] release will have to wait on a [pool] release, but that’s in the works.

38920 [cli][PATCH] cli2.commandline.Parser.parseAndHelp need no…

A patch from James Ring for [cli] being applied by yours truly. This is cleanup for a previous patch that James had applied to simplify the [cli] v2 API.

New

38927 [io] Tests Fail on Windows 2000

Niall’s running his own show here with some Windows based test failures he’s found in the upcoming release of [io]. Some are getting resolved, but I think from reading this that there are still some left to deal with.

38929 [configuration] Support event listeners for configurations

A new API enhancement from Oliver Heger.

38935 [VFS] remove commons-collections dependency from ram prov…

A patch to remove a dependency on [collections] from VFS. A common problem at Commons - eating our own dogfood causes pain to the users, so often you see inlined code from another component, as in this example (I presume).

Summary

And that’s it for today. Lots of activity.

So how do people manage their OpenPGP (specifically GPG) keys? For the last year or more I’ve avoided releasing things at Apache due to a general confusion at just how to manage my keys. There’s nothing available at Apache to do so, so you have to come up with something locally. Putting it on another server seems insecure, but client-wise I’m never that stable. Currently I’m on a work laptop, so probably not something to get too attached to in the long-term.

My current solution is to use a usb key and symlink my .gnupg directory over to that as I’ve not seen anything in the gpg command help to let me spcify where the keys live. This worked fine for signing, but I just realised that it means that I have to have the usb key in to verify keys - ie: it needs to find my list of public keys and those are in the .gnupg directory too. This sucks. Any ideas on how to better manage this out there?

One downside of the usb key approach is that it gets automounted in a way that gpg doesn’t like. The files are world readable, and lock() isn’t supported. So lots of errors. This is on a mac, so in the long-term I’d have to look into understanding how to configure its automounting.

I spend my days checking up on the issue reports in various projects - the major focus being on Struts/Axis/Spring/Hibernate. So why not do a Commons one on my blog?

Today in Commons there were 5 modified issues. 2 new, 3 notable.

New

38912 - An enhancement discussion on a version of java.util.Service for Commons Lang.

38917 - A bug with much in the way of patches in Commons Daemon - in which the Windows prunsrv prohibits the use of non-alphanumeric names.

Notable

38073 - Continued discussion on a issue in Commons DBCP to do with JNDI problems with the SharedPoolDataSource. It’s a few months old and now seeing some action with DBCP looking release-wards.

38401 - A bug in DurationFormatUtils in Commons Lang. Yu Peng has submitted a patch for a gnarly bug; least I thought it was gnarly when I was looking at how to fix it surgically a couple of nights ago.

38885 - An enhancement for Commons JCI; sounds like it’ll go in once a dependency is updated.

Summary

All in all, a nice haul. Positive outcomes, quick responses.

I’ve got a little tool called ReportRunner that is currently rather tricky to use - as you have to grab the jar, create the configuration files and get the exploded webapp; ie) lots of jsp and javascript etc.

I’m not too concerned with the config; but does anyone have any thoughts on how better to do the small web wizard? I’d like it to be in the jar if possible - currently I’m slowly migrating the jsp to velocity - but other ideas would be welcomed.

Ignoring discussions of adjusting things at Jakarta, there’s an interesting surge in life at Jakarta Commons at the moment. The following seem to be in the works:

• Collections 3.1.
• Validator 1.2.1.
• Lang 2.2
• IO 1.2
• Compress 1.0 (needs interested contributors to get promoted out of the sandbox)
• SCXML 1.0 (needs interested contributors to get promoted out of the sandbox)
• Logging 1.1/2.0
• a Pool release
• a DBCP release

And of course I’m trying to get [CLI] back on target, help appreciated - though help is appreciated on any Commons component.

I’d love to hear opinions on Commons CLI and the two APIs it has within the current SVN HEAD (it’s been that way for a good couple of years). Am slowly working towards getting a new version released.

• org.apache.commons.cli (the 1.0 API) is better, gimme my damn bugfixes
• org.apache.commons.cli2 (the 2.0 API) is better, release already
• Commons CLI is not as good as this equally open licenced alternative (please, no copyleft)

I know there are people out there using the CLI-2 API; and so I’m staring at a bunch of CLI-1 bugs that will require more than a surgical bugfix and are pointless to fix if the CLI-2 API is seen as the superior.