Tina Coleman

Ranch Hand
+ Follow
since Dec 12, 2001
Merit badge: grant badges
For More
Cows and Likes
Total received
In last 30 days
Total given
Total received
Received in last 30 days
Total given
Given in last 30 days
Forums and Threads
Scavenger Hunt
expand Ranch Hand Scavenger Hunt
expand Greenhorn Scavenger Hunt

Recent posts by Tina Coleman

Does the new book go into utilizing p2 from within applications? I've seen documentation that talks about how to add p2 meta-information to my plugin, but haven't yet found a good source as to how to use p2 to let me implement an auto-update / auto-install feature for my own products.

Great fan of the earlier versions of your book, btw, and excited to see it updated and extended for Ganymede!
Our team has implemented a core model using JAXB, which we'd now like to allow other teams to extend.

Thus, we'd like to distribute a jar file containing our generated compiled code, with the XSDs as resources. Optimally, other teams would then import our XSDs (so that the new XSD could use its own namespace) and generate JUST their own code. Having difficulty making that happen, though. Two difficulties that I'm having so far:
- pointing to the schema definition within a jar file: I've seen mention of classpath schemaLocations, but those don't seem to work for Xerces.
- If I extract the XSDs, and give explicit schemaLocations (or presumably, post them in an accessible location), the JAXB code generation generates all of the classes, rather than just the set of interest.

I should mention we're using the JAXB reference implementation from Sun, which uses Xerces.

Code for my test extension:

Code for my jaxb binding:
Found my own answer, after seeing a thread on comp.java.cruise-control.user: the svn plugin only deals with checking to see if there's been a modification. It doesn't do the actual update.

Current intent is to add an additional step to my schedule set to retrieve the files from subversion.
16 years ago
I have CruiseControl 2.7 running, and have set up a configuration file that has an svn plugin configured as so:

<plugin name="svn" localworkingcopy="${svnLocalCopy}" property="svnUpdate"

I then utilize that plugin within a specific project like so:
<svn />

CruiseControl is appropriately triggering our builds when a file changes in the repository. However, if we add a new file to the repository, the SVN plugin is not picking it up.

I can tell via the SVN command-line that the file is in the remote repository, using svn status -u. How do I need to configure my build process to recognize those new files?
16 years ago
My first check would be to see if your sequence exists in the database, and verify that it is indeed incrementing appropriately.
If you're using log4j, you can add the following properties to your log4j.properties file to get some additional information:
Verify: did you assign the alertId for the child? I see that it's an assigned id. Given that your debug message claims that it's an 'Update on child', that suggests that you're reusing your previous child based on reusing an assigned id, rather than creating a new one.

Can you generate any of the hibernate-specific logs?
We have a object hierarchy that includes both inheritance and and interface implementation. In certain cases, it makes sense for our object model to return an interface, rather than an implementation. However, when we attempt to retrieve, we don't get back any rows (empty resultset). What we see in the generated SQL is that hibernate appends a "1=2" to what would be an otherwise valid SQL query.

Has anyone else run into this before, or can you point me to examples of retrieval via interface? (E.g., getInterfaceA, rather than getClassA.)
Looking for some guidance on using StrutsTestCase with Tiles.

Within struts-config, I have an action entry like so:
<action path="/servicesAndRates" forward=".servicesAndRates" />

.servicesAndRates refers to a tiles definition within an XML file mapped via its plugin definition to Struts.

If I bring up my application, the application resolves the tiles definition appropriately. In my StrutsTestCase,though, I've been unable to verify the tiles forward.

What would be the appropriate mechanism to determine that the action appropriately resolves its configuration?

Notes on what I've tried thusfar, and my attempt at explaining why they don't work:
verifyTilesForward: seems to require a forward as a result of an action, rather than directly via the mapping
verifyForwardPath: gives me the base layout from which my main tiles all derive, and thus not a very good test check
verifyInputTilesForward: since there's no input to return to, this doesn't seem to work.
[ April 05, 2006: Message edited by: Tina Coleman ]
18 years ago
I was disappointed to not get to see a table of contents and index for the book out on Amazon. I know that's not likely the authors' doing, but passing along that feedback. Speaking as someone who's done a good bit of .NET programming of late, I wonder if the authors could speak to how much of the text is J2EE-specific, and how much would be more widely applicable. I expect that since this is a patterns book, it should be more widely applicable. Definitely interested in some of the various topics listed in the book blurbs.
18 years ago
It would seem to me that if I roll my own, even though I come up with new code, I'm likely to fall into the same patterns as someone else coming up with code to solve the same problem. So, effectively, my choice becomes whether I'm implementing a pattern (likely bad) that I don't recognize, or implementing patterns that have been examined with an eye for security. Given that patterns are usually not created, but are generally documented after being recognized from best practices, I figure I want to copy the best.
18 years ago
Our company's interested in getting several folks certified. Our expectation is that there a core group of folks who could reasonably pass the exam with a bit of brushup. Does anyone have a recommended format or syllabus for a group study for SCJP 1.4?
Ethics in computer science: I'd almost posit that ethics in computer science would be ethics in whatever problem domain it was you were applying technology to. So, if I were building an accounting system, then the ethical domain would be that of accounting. In a medical system, I'd be worry about respecting the appropriate ethical rules in medicine. In a publishing system, I'd be worried about respecting copyright laws. Etc, etc.
I guess I'm interested in seeing the ethics list Valentin offered to post. I'm not getting the connection. Perhaps directly demonstrating the need for instruction in ethics...
Run JavaDoc on your test classes, if you have any sort of method description there. If you're slick, then you can apply a transform on the documentation to get a document in exactly the format your boss wants. If you're like me, and not slick, at least you've got an easy set to copy/paste from.
20 years ago
Object-relational mapping frameworks seem to be cropping up all over the place. I'm aware of Torque (actually seen it in use in the Scarab bug tracking database), Cayenne ( http://www.objectstyle.org/cayenne/index.html), OJB ( http://db.apache.org/ojb/), and Hibernate ( http://www.hibernate.org/). Then there's JDO, of course, which tackles the same basic problem (object-relational mismatch) from a different perspective. What are some of the differentiators folks use to make a decision amongst the toolsets out there?