Wikipedia:Articles for deletion/JamVM


 * The following discussion is an archived debate of the proposed deletion of the article below. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's talk page or in a deletion review).  No further edits should be made to this page.

The result was   keep. However, please add more third-party references, or a lot of people might want to nominate it again. Max Semenik (talk) 19:43, 6 January 2012 (UTC)

JamVM

 * – ( View AfD View log )


 * See this blog post by the software's author. Chris Cunningham (user:thumperward) (talk) 17:21, 6 January 2012 (UTC)

Procedural nom mostly. Last prod is fundamentally sound, removed by an editor who added a current work paragraph. I don't see enough indications of notability here. Shadowjams (talk) 09:48, 23 December 2011 (UTC)


 * The primary claim to fame here appears to be the assertion that the Dalvik Virtual Machine is based on JamVM. However, not only can I not find a reliable source for this (a blog post and a forum comment both assert it individually, but with no evidence other than "seems to be" or "read the code", and none of the reliable sources given in the Dalvik article mention it at all), but notability isn't inherited and if that's the only noteworthy aspect of the subject then little more is needed than a footnote to that effect in the Dalvik article. Personally I wouldn't even both with that unless a better source can be found. Chris Cunningham (user:thumperward) (talk) 12:36, 23 December 2011 (UTC)
 * Note: This debate has been included in the list of Software-related deletion discussions.  —Tom Morris (talk) 13:51, 23 December 2011 (UTC)
 * Comment I think we need to be very careful here as we are potentially making a serious allegation.  Even if the article is kept, unless there is a more reliable source, I think it needs to be removed. Pit-yacker (talk) 19:52, 29 December 2011 (UTC)
 * Comment Early android/dalvik technical pages mentioned JamVM explicitly as the runtime used. See for example http://web.archive.org/web/20091027081958/http://pdk.android.com/online-pdk/guide/debugging_native.html which explains how to debug native (JNI) code on Android and includes instructions on how to run GDB and gives stack traces that matches the JamVM source code. Dalvik seems to be a newer runtime that Android uses now, which might or might not be based on or inspired by JamVM. — Preceding unsigned comment added by 80.101.103.228 (talk) 21:57, 30 December 2011 (UTC)

Firstly some information regarding its current status. JamVM is very much actively developed. It has had several new releases, and it is widely used, both in its GNU Classpath version, and its OpenJDK version. The last release date of January 1, 2010 is misleading. In the past, JamVM only worked with GNU Classpath. GNU Classpath consisted of the Java class library, it did not contain a Java Virtual Machine. As such, JamVM was packaged and released separately, but both are needed for a working Java system. Regular releases of JamVM were needed to keep in sync with changes in GNU Classpath. Since the release of OpenJDK, GNU Classpath development has mostly ceased, its last release being GNU Classpath 0.98, on February 5, 2009. The JamVM release date of January 1, 2010 relates to the last GNU Classpath version of JamVM. I rewrote JamVM to support OpenJDK during 2010, with an initial announcement on my blog on February 16, 2011, and an announcement at FOSDEM. It was then integrated into IcedTea by Xerxes Ranby and committed. Since then JamVM has been released as part of IcedTea, thus removing the need for separate releases of JamVM. It is supported by both IcedTea 1.x (OpenJDK6) and IcedTea 2.0 (OpenJDK7) (the latest release). See the release announcement of IcedTea 2.0 for proof of JamVM support. As part of IcedTea, JamVM is packaged for most Linux distributions, including Debian and Ubuntu, across a wide range of architectures. After discussions at the Ubuntu Developer Summit in Budapest (of which I was a guest), and after extensive testing, JamVM was selected as the default Java Virtual Machine for Ubuntu 11.10 on ARM. Without putting too fine a point on it, this means that when you run IcedTea (java) on the current version of Ubuntu/ARM, you are using JamVM. The Ubuntu blueprint can be found here:[ https://wiki.ubuntu.com/FoundationsTeam/Specs/JvmOnARM] and the changelog making JamVM the default VM can be found here: The full changelog for OpenJDK6 is here, with references to JamVM as the default VM for oneiric/ARM (Ubuntu 11.10) and armhf. Further points for JamVM notability is its inclusion into all the popular embedded build systems (the GNU Classpath version of JamVM), including openembedded and busybox buildroot. JamVM has also been extensively used within academia as the basis of research. This is indicated by over 200 hits from a Google scholar search: and JamVM is also featured in books on embedded Linux, e.g. Pro Linux Embedded Systems (Gene Sally) Linaro (http://en.wikipedia.org/wiki/Linaro) also recently discussed Java support on ARM. As with Ubuntu, their decision was to support JamVM. — Preceding unsigned comment added by 80.177.127.218 (talk) 03:30, 24 December 2011 (UTC)
 * I'm the author of JamVM (Robert Lougher). For the record, I didn't create the JamVM page, nor have I edited it (beyond updating the release information on one occasion).  I prefer others to make the case for JamVM, but I will provide some information in case it can help the case for non-deletion.

*Delete: Only claim to notability is based on an allegation from a blog that Google copied it. If we can substantiate the claim, then per User:Thumperward, that fact can be included in Dalvik Virtual Machine as a) the specific allegation would suggest it is a fork from an earlier version and b) It is Dalvik that is notable not a product that it was based on. Pit-yacker (talk) 19:52, 29 December 2011 (UTC)


 * Keep: I never used JamVM yet. However, since JamVM is being used as Java Virtual Machine for ubuntu as the default VM (in ARMs Ubuntu), I think it is an important VM and a notable software. –ebraminiotalk 09:40, 30 December 2011 (UTC)


 * Keep: JamVM is not only interesting because other virtual machines like Dalvik are inspired by it, but also because it is ported to so many architectures, it is the only JVM ported to ARMEL, that it is used in university courses to teach about garbage collection (see the Jam-o-lizer garbage collection visualization system), optimizations for (java) runtime systems (see YETI: a graduallY Extensible Trace Interpreter based on JamVM) and various university courses that include teaching common virtual machine techniques like stack caching, direct-threading, super-instructions, etc, because it is so neatly written. Should these facts and references be added to the article at this point? Or is it more appropriate to wait to add new stuff till the article is no longer marked up for deletion? — Preceding unsigned comment added by 80.101.103.228 (talk) 09:57, 30 December 2011 (UTC)

Rémi — Preceding unsigned comment added by 88.167.50.47 (talk) 15:01, 30 December 2011 (UTC)
 * Keep: JamVM is an important VM because of it's a simple and elegant designed JVM. As already stated, it's the default JVM for ARM Ubuntu and it is used/patched in several Universities (included mine (University of Paris East)) because the code is not hard to read compared to other VM. There is no direct relation between JamVM and Dalvik, but when I read the code of Dalvik I've found that some parts use tricks pretty similar to the ones use (or used to be exact) by JamVM but does it something that it worth mentioning, I don't think so.


 * Keep: Having a readable and understandable VM is important, especially as already mentioned when its being used for academic study.  In and of itself that would suffice for notoriety to my mind.  Not only that but debian provide opt-in popularity statistics for packages that they distribute, which shows that out of the 118626 debian packages (at the time of writing) 'icedtea-6-jre-jamvm' is in 2406th place - ie within the top 2.5% most frequently installed packages.  Bear in mind that that places it above the 'flac', 'wireshark' and 'mencoder' packages, none of which have wikipedia pages flagged for deletion on notoriety grounds.  I'm not claiming that popularity is a necessary criteria for notoriety, but surely its a sufficient condition? You can check http://popcon.debian.org/by_inst for full listings.  — Preceding unsigned comment added by RichardWarburton (talk • contribs) 18:43, 30 December 2011 (UTC)


 * Keep: I have used JamVM as a basis for a modified research VM used in my academic research. Results have been published in IEEE proceedings as a peer reviewed paper http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?arnumber=1541156 and have been followed up by further work building on the first work. For its source code's ease of reading and understanding JamVM is a frequent choice for virtual machine researchers. Aszegedi (talk) 20:30, 30 December 2011 (UTC)


 * keep: Many open-source VMs have pages on wikipedia. Unlike JamVM, many of these are no longer maintained and are only of historical interest.  JamVM, however, is active, up to date, packaged by leading distributions and has many current users.  It seems odd that JamVM gets fingered for deletion ahead of these.  Also, the paragraph that caused so many problems has been removed. Ploppy (talk) 22:23, 30 December 2011 (UTC)


 * Comment: I'm striking the delete for now solely due to the apparent coverage in peer reviewed literature. However, I will note that I disagree with most of the other justifications for keeping made so far. For example, WP:OTHERCRAP is not a reason to keep - if you have problems with other articles nominate them too! Furthermore, readability of code does not necssarily make the subject notable - and it certainly isn't the job of Wikipedia to make the subject notable because it has readable code. Equally, JamVM could shipped with every electronic device on the planet, however it still wouldn't be notable if there weren't reliable independent sources top atest to this Pit-yacker (talk) 23:39, 30 December 2011 (UTC)


 * Keep - mostly verifiable content, many mentions in literature, several articles with special focus on the JamVM, subject of a USENIX talk; for links, start with http://scholar.google.de/scholar?q=jamvm, whose first five results include two substantial citations . Shouldn't there be some requirement for spending 90 seconds on basic checks like the above Google Scholar search before AfD-listing articles receiving ongoing constructive input?   Having said that, kudos to the editors who pointed out the issues with sourcing in the article, which should be and can be fixed.  &mdash; Charles Stewart (talk) 13:50, 31 December 2011 (UTC)
 * The top five results on that Scholar search are all to papers by the project's author. I wouldn't exactly describe them as secondary sources even if they're peer-reviewed. Chris Cunningham (user:thumperward) (talk) 17:19, 6 January 2012 (UTC)
 * The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made on the appropriate discussion page (such as the article's talk page or in a deletion review). No further edits should be made to this page.