Finding the root cause of a Maven build issue made me question my sanity, as seemingly everything I knew about Maven seemed to be wrong. Luckily, in the end logic was victorious, and quite a nasty bug was found.
Caches are bugs waiting to happen. Rob Pike
The last two days I was hunting an issue in the Maven/Tycho based build of VIATRA: some artifacts from Maven Central were not found anymore, seemingly without any related change. A more detailed analysis has shown that the issue is specific to the build server at the Eclipse Foundation: in all other cases the builds run successfully.
When looking at the debug output of the build (the -X switch for mvn is truly a killer feature…) it was interesting that the build did not try to download the dependency from Maven Central nor from the mirror set up for builds at the Hudson instance at eclipse. Given that the first twenty-two modules compiled (with dependencies to Maven Central) this made no sense.
After managing the sanity loss and calling for help (thanks Ábel and Balázs) we have managed to identify that the settings.xml used on Hudson caused the problem. After calling for more help (thanks for the quick help from webmaster) I have finally managed to find out the root cause of the issue: repository identifier clash caused by the maven central mirror and a dependency issue.
For performance and network utilization reasons a mirror for maven central was set up about a month ago using the following fragment:
<name>Eclipse Central Proxy</name>
Our problematic bundle had a repository declaration as follows:
What happened here is that we declared some repositories with the same identifier as the proxy repository declared by webmaster, and Maven got confused, and thought that everything from Central is reachable from our declared dependency repository. Changing the identifier to avoid the name clash solved this issue nicely.
Now only one question remains: why was this issue not found earlier. The answer is trivial: the plugins were available in the local repository, effectively hiding the unresolvable dependencies. A completely unrelated cleanup of the local repository however has brought forth this issue, causing much headache to find.
The main lesson we learned here: DO NOT reuse the same repository identifier for multiple repositories. It can cause very subtle, hard to debug issues in the long run. However, there are a few other aspects you have to keep in mind:
- All repositories and plugin repositories defined in your parent project are also inherited, so their name should also be unique.
- Identifier clash might not be a problem in case of deployment repositories, but have not tested this aspect yet.
- A more specific lesson for Eclipse projects building on Foundation Servers:
DO NOT use the identifier “repo.eclipse.org” for your repositories. Maybe it is worth checking whether the identifier is used (e.g. use following the github search link to check projects developed or mirrored to the Eclipse Github organization), and update accordingly. Update: the mirror repository id was updated to “eclipse.maven.central.mirror” that should not clash with manual repository names.
Parsing textual notation has a long history in computer science. This long time has produced many great and efficient parsing algorithms, which were optimized to the extreme by compilers for different programming languages. However in the recent years the development of integrated development environments are accelerated, adding more and more services to previously simple features like code editors and automatic background compilation. These with the rise of domain specific language frameworks which allows the user to define an arbitrary language declaratively causes that present-day computers still struggle under the resource requirements of these technologies.
While waiting in front of a textual editor, I’ve started thinking about why does the editor need to re-run the parsing algorithm every time I hit button (of course I’m aware of the fact that the main cause of the slow responsibility is probably not the parsing algorithm itself). A fully incremental approach could make these tools more responsive which would remove a lot of pain from its users. As I’m already familiar with an incremental query engine (IncQuery) the thought about using it to parse text couldn’t leave my mind.
The following article presents the experiment I’ve done as a proof-of-concept. It’s far from being useful and reveals more problems than it solves it does not produce an AST just accepts or rejects the input string, however it may be interesting as an approach and maybe in the future it will be more than just a stupid idea. Continue reading “Parsing textual input with IncQuery”
Most of the time I try to avoid reinventing the wheel. And most of the time I fail and forced to do so. That’s what happened to me when I decided to write my own textual modeling framework. I’ve used (and currently using) Xtext in several projects and tried EMFText once. Both of them are great projects with great possibilities, but I’ve come across a problem which required a different approach. I needed a more dynamic toolkit, where the language is extensible and easier to maintain. Continue reading “Yet another textual modeling framework”
Since I’ve always been interested in computer-aided music notation and a strong proponent of open-source software, I’ve been advocating MuseScore among my fellow musician colleagues since multiple releases. When learning the usage of a new piece of software, questions always pop up about the most common sheet music engraving tasks. When Packt Publishing asked me to review Instant MuseScore, I was eager to see whether this e-book fulfills its promise of being a pragmatic hands-on quickstart guide to the first steps of using MuseScore. Here are some of my thoughts.
Its workflow-oriented structure is very practical (although I would definitely not put the “Barlines and repeats” section into the “Formatting” chapter, especially repeats are clearly not a matter of formatting). The tip about setting the velocity of special notes (e.g. slashes) was new and useful for me. The notes about potential caveats are also helpful and can save a lot of frustration.
For an “Instant” book which aims to be quickly skimmable, more formatting emphases would be good, as well as indicating the keyboard shortcuts for every command mentioned. The used version of MuseScore could be more prominent because there are major UI changes between new versions.
I would note that ties can be inserted with the numeric plus sign (on most keyboards this character is bound to Shift+3, which is reserved for inserting a third below), and also I think it would be important to emphasize that unfortunately, extracted parts are separate copies, not views (as opposed to MuseScore’s commercial counterparts).
When describing expression anchors, I miss mentioning that they also affect playback obviously, not just the layout of the sheet music. In the lyrics chapter, it would be also worth showing the keyboard shortcut of inserting a space into a syllable. (Though this is not the most common case, it occurs quite frequently in some languages, e.g. in Italian.)
One tiny but apparent technical issue with the PDF version of the e-book: the embedded font does not support the Cmd and the spacer icon.
Although I doubt that a book (especially a printed one) is the most suitable medium for teaching the usage of such a complex application (screencasts are better, the best would be tutorials integrated into the interface of the application itself), if you can afford its not too low price, Instant MuseScore will give you a gentle introduction to this feature-packed but yet maturing program.
Today a new Eclipse version codenamed Kepler was released, with a lot of nice new features, including conflict handling during update, an update user interface for the Marketplace client or various EGit updates (my favorite is the Commit and Push button in the Commit dialog).
About the less visible stuff, in the modeling projects we use, e.g. EMF, Graphiti or Xtext also got some internal updates, resulting in the need to update our projects. As of now, I updated the Xtext Reflective Editor, as a change in EMF 2.9 made it unusable in Kepler.
I often use this tool to debug the Xtext-based parsers that use Xbase and model inferrers, as it displays the generated model using the EMF reflective editors. Version 0.5.5 is a recommended minor update – it works on both newer EMF/Xtext versions, but maintains compatibility with older EMF versions. It is still downloadable from our update site http://eclipse.cubussapiens.hu.
Thanks for all contributors the new, nice features of Kepler (and of course fixes as well). This can become a new platform for new projects for me – the reflective editor update is only the first of them.