Eclipse Community Forums
Forum Search:

Search      Help    Register    Login    Home
Home » Eclipse Projects » Virgo » Virgo Project wil be archived?
icon4.gif  Virgo Project wil be archived? [message #1111602] Wed, 18 September 2013 08:49 Go to next message
Eduardo Frazão is currently offline Eduardo Frazão
Messages: 118
Registered: January 2012
Senior Member
Hi folks. Talking on IRC, a friend shows me this THREAD:

http://dev.eclipse.org/mhonarc/lists/virgo-dev/msg01488.html

The Virgo project can really be archived? Its the only suitable solution for pure OSGi applications, with the power of Spring. We will be lost ... =/

Re: Virgo Project wil be archived? [message #1111736 is a reply to message #1111602] Wed, 18 September 2013 13:05 Go to previous messageGo to next message
Miles Parker is currently offline Miles Parker
Messages: 1338
Registered: July 2009
Senior Member
Hi Eduardo,

I think, sadly, the message speaks for itself. If this technology is truly important for the community then someone in the community needs to step up to see it continue to be supported. So now may be the time for the user community to rise to that occasion..

Note that the level of support does not need to be that great. The fact is that Virgo works quite well for its intended purposes now. The chief need would be to coordinate and maintain builds and ensure that basic bug triage is accomplished -- ideally that would be someone familiar with Eclipse Development Process along with Virgo internals. And then there should be thought given to where the technology evolves to, and what future releases would look like, but I'm sure that there would be a lot of flexibility with respect to that.

cheers,

Miles

[Updated on: Wed, 18 September 2013 13:06]

Report message to a moderator

Re: Virgo Project wil be archived? [message #1111795 is a reply to message #1111736] Wed, 18 September 2013 14:57 Go to previous messageGo to next message
Eduardo Frazão is currently offline Eduardo Frazão
Messages: 118
Registered: January 2012
Senior Member
Hi Miles Parker! Thanks by your answer!

Only by the moviment on the forum, we can conclude that this Technology is import for many. I have 4 products heavy based on Virgo, and the bigger one is comming.

I dont see some Virgo features in any other projects.

- Spring support out of the box, and with Spring OSGi Integration
- Easy AOP support with Equinox Weaving hooks
- Gemini Projects.
- GEMINI WEB! (Goddbye Servlet http service bridges, and other things)
- Blueprint
- DB Access
- JPA Made easy with EclipseLink and Gemini
- And others as naming, etc
- Simplified Deploy Model
- Easy Logging.
- Various flavors for minimal dependencies

- More can be read here: http://www.eclipse.org/virgo/benefits/ (The ideal OSGi Server Runtime?)

Virgo was the light on the end of tunel for Enterprise OSGi Applications. (At least for me).

I think that ARCHIVE a project like that will be a big mistake, after so many efforts to it be what it is today, and right now, when it reaches the desired stability of tooling, etc. We, normal users that belive on this solution, and builds our projects based on, are very very concerned, even more when Spring simple says that they will not release Spring with OSGi metadata anymore. What is going on? Smile. Modularity isn't anymore a concern of enterprise development?


[Updated on: Wed, 18 September 2013 15:07]

Report message to a moderator

Re: Virgo Project wil be archived? [message #1112110 is a reply to message #1111795] Thu, 19 September 2013 02:20 Go to previous messageGo to next message
Tin N/A is currently offline Tin N/A
Messages: 46
Registered: December 2010
Member
Eduardo Frazão wrote on Wed, 18 September 2013 20:57
Modularity isn't anymore a concern of enterprise development?


The problem is that the threshold to adopting OSGi is still much too high. Java has zillions of useful libraries that people use in their enterprise applications, and although the majority of those are OSGi-compatible, when you hit a classloader-related problem you're toast - your project relying on that particular functionality could be halted for an unspecified amount of time or required to be re-engineered from scratch.

So you need adoption, but to have adoption you need developers to build a useful "battle-set" of libraries and technologies that cover a much wider range of functionalities than what's in Virgo core. And they won't show up for any of that if they're not sure that they won't need to repeat a very painstaking process of testing and adapting each new version of the library to OSGi. IMHO, any OSGi based project such as this will have heaps of troubles to go mainstream and acquire a critical mass of developers until the whole Java community turns to modularity which is quite an extensive mindset change. I have no doubts that this will happen, but right now Virgo is too far ahead of its time and it's faith is to languish at the margins for a few more years or even decades (the source is open so even if Eclipse archives it, I'm sure there will be some life left in it from its current small community through alternative means like github or similar).

Re: Virgo Project wil be archived? [message #1112324 is a reply to message #1112110] Thu, 19 September 2013 08:39 Go to previous messageGo to next message
Eduardo Frazão is currently offline Eduardo Frazão
Messages: 118
Registered: January 2012
Senior Member
Well. For this point of view, seems like the Virgo, and others of this type is doomed. So... its time to revert our projects to a flat class path model? The problem for me, right now, isn't the flat classpath, but the high modular design of our solutions, totaly built around of plugins, etc.

Maybe it was written to early.


Future proof
Virgo is an open source Eclipse project with a liberal license and active participation from multiple vendors, which positions it ideally for the future.

From: http://www.eclipse.org/virgo/benefits/

Sad

[Updated on: Thu, 19 September 2013 08:43]

Report message to a moderator

Re: Virgo Project wil be archived? [message #1112381 is a reply to message #1112324] Thu, 19 September 2013 11:48 Go to previous messageGo to next message
Tin N/A is currently offline Tin N/A
Messages: 46
Registered: December 2010
Member
Eduardo Frazão wrote on Thu, 19 September 2013 14:39
Well. For this point of view, seems like the Virgo, and others of this type is doomed. So... its time to revert our projects to a flat class path model? The problem for me, right now, isn't the flat classpath, but the high modular design of our solutions, totaly built around of plugins, etc.


I don't think you need to migrate back if you have the whole system working in modules right now (and expertise to build on it). In the company I work for, we also have Virgo in production (in fact, still using the 3.0 version), and it'll stay in production for quite a few more years with upgrades to the functionalities as needed. Right now, I don't see any new features that we would need from Virgo to keep our system running, so in that sense Virgo really is futureproof - it already provides everything we need for the next few years. And hopefully this will be enough for the rest of the Java world to bridge the gap and the whole modularity concept becomes mainstream.
Re: Virgo Project wil be archived? [message #1112417 is a reply to message #1112381] Thu, 19 September 2013 13:06 Go to previous messageGo to next message
Eduardo Frazão is currently offline Eduardo Frazão
Messages: 118
Registered: January 2012
Senior Member
Yes. I agree with you. Right now, i'm working on a project that is 60% done. Its is uses Spring (DI, Transaction Management, and other stuff), AOP via LTW, Gemini JPA, Snaps. I can't see this needs, supplied by another OSGi Runtime Server right now, and I cant break development, and start a transition in this moment. My concern is about possible bug fixes, IDE Tooling, etc. Spring will continue its evolution, and Virgo versions will be outdated. Even if we can upgrade minor versions of Spring, how will the major updates be applied?
Will be other Gemini Projects continued? Gemini JPA have some bugs, and Eclipselink is releasing new versions.
Re: Virgo Project wil be archived? [message #1112738 is a reply to message #1112417] Fri, 20 September 2013 01:14 Go to previous messageGo to next message
Tin N/A is currently offline Tin N/A
Messages: 46
Registered: December 2010
Member
Eduardo Frazão wrote on Thu, 19 September 2013 19:06
My concern is about possible bug fixes, IDE Tooling, etc. Spring will continue its evolution, and Virgo versions will be outdated. Even if we can upgrade minor versions of Spring, how will the major updates be applied?
Will be other Gemini Projects continued? Gemini JPA have some bugs, and Eclipselink is releasing new versions.


Well this is where all the advantages of modularity should really come into view. Because everything that you mentioned is provided as bundles, and you should be able to just plug it into your Virgo runtime with minimal or no effort required. AFAIK, other Gemini projects (except Management I think) are not affected with this situation.
Re: Virgo Project wil be archived? [message #1123365 is a reply to message #1112738] Wed, 02 October 2013 09:45 Go to previous messageGo to next message
Glyn Normington is currently offline Glyn Normington
Messages: 1222
Registered: July 2009
Senior Member
If the project does get archived, then maintenance will sooner or later become very difficult, even with modularity in place. Dependency upgrades for critical bug fixes would need to be handled very carefully since it won't be possible to do a proper Virgo release (which runs all the tests right up the stack of git repositories) for an archived project. (In particular, I'd be careful always to perturb dependencies by the smallest version change possible, especially if "monkey patching". For example, rather than trying to pick up a bug fix by upgrading the major or minor version of a dependency, try instead to backport just the necessary fix so as to upgrade only the micro or qualifier version.)
Re: Virgo Project wil be archived? [message #1123393 is a reply to message #1123365] Wed, 02 October 2013 10:15 Go to previous messageGo to next message
Eduardo Frazão is currently offline Eduardo Frazão
Messages: 118
Registered: January 2012
Senior Member
Glyn, until now, no news about a new Project Leader? I'm sure that the Virgo community is very concerned about that. I think that for most projects, replace Virgo is not a trivial task.
Re: Virgo Project wil be archived? [message #1123405 is a reply to message #1123393] Wed, 02 October 2013 10:27 Go to previous messageGo to next message
Glyn Normington is currently offline Glyn Normington
Messages: 1222
Registered: July 2009
Senior Member
Correct. Several people have expressed concerns, but no-one is quite concerned enough to volunteer their services. Wink
Re: Virgo Project wil be archived? [message #1123410 is a reply to message #1123405] Wed, 02 October 2013 10:35 Go to previous message
Eduardo Frazão is currently offline Eduardo Frazão
Messages: 118
Registered: January 2012
Senior Member
What a pity. Unfortunatelly, my knowledge is very far from any Virgo commiter. I never have study their internals. Otherwise, I will be happy to contribute, as I do with some other small projects Sad.
Previous Topic:Is it possible to independently upgrade the version of Tomcat used by Virgo Tomcat?
Next Topic:using autbind on linux for virgo tomcat server to bind on ports < 1024
Goto Forum:
  


Current Time: Thu Jul 31 23:59:49 EDT 2014

Powered by FUDForum. Page generated in 0.01858 seconds