[edit] This site was an experiment and as of April 7th, 2009 is no longer accepting changes. It is kept up for archival and reference purposes. New work is ongoing in the direction of having a Pointrel Social Semantic Desktop instead of a Halo Semantic MediaWiki.

Project Status

From Oscomak

Jump to: navigation, search

Contents

Project Status

Projects

Google Group

http://groups.google.com/group/openvirgle

This is the primary discussion location for OpenVirgle projects.

Project Leader: Doram

Status: Update In Progress

Status Date: 5/7/2008

Details:

I will be writing the Google Group Page duplicates of these Wiki files, as soon as the Wiki files are completed. See Full Weekly Report for more details.


Semantic MediaWiki Plus

http://www.oscomak.net/wiki/OpenVirgle

This is the main repository for information on OpenVirgle Projects. Also supports limited focused discussion on individual topics.

Project Leader: Paul and Mike

Status: Needs Organization

Status Date: 5/4/2008

Details:

Discussions are underway for determining the best way to proceed with semantic tagging and importing information into the Wiki, as well as utilizing the other Wiki add-ons to their highest potential (Halo, gardening, etc.).

See thread: Examples of of a procedures for the OSCOMAK Wiki and related issues


SKDB

-

This will be the eventual back end to the OpenVirgle project.

Project Leader: Bryan

Status: Needs Help - see Calls to Action and Calls for Help

Status Date: 5/5/2008

Details:

Python programming work continues, but is currently slowed down by a lack of support from external source providers:

"the prototype client interface (in python) is going smoothly, it's actually to the point where I need some help deciding where to drop files on the user's computer -- where will the packages go? What about the metadata files? What's the usual directory structure in debian for this? And what about the possibility of using makefile? That has to be investigated. Come to think of it, a debian programmer onboard would really, really help at this point."

... "I'd be interested in organizing more clearly with the debian deity revamp project team members before I do anything stupid with the codebase."

..."Second, after that is solved for agx-get, I need to do some sample files in the 'metadata format', some sample scripts, and then have Ben do his autospec implementation. The 'ghost aggregation layer' aspects will come in at this point, as we write up some tutorials on how to join your git repos into 'metarepo', how to crawl them, and so on. But before this happens, I need some confirmation from the debian deity team re: the new architecture for debtorrents. " (The debian project is currently overwhelmed by their traffic load)

Also see Bryan's blog: http://heybryan.org/~bbishop/cgi-bin/blosxom.cgi/


Google Code Project

http://code.google.com/p/openvirgle/

This is the center for programming-based simulation work for the OpenVirgle Project.

Project Leader: Paul

Status: On hold while figuring out Wiki

Status Date: 5/5/2008

Details:

Not mentioned in last update, but has had no updates in a while.


OSCOMAK/OpenVirgle.net Server Maintenance

http://www.openvirgle.net/

This is the Front Page/Gateway to the OpenVirgle Project. Use this for all official links.

Project Leader: Paul and Doram

Status: Stable, Upgrade Complete

Status Date: 5/7/2008 4:21 pm EST

Details:

The server has been updated to Pair Networks dedicated server QS-1

Core 2 Duo E4400 (2GHz) CPU

120GB Hard Drive

Daily back-ups

See Mid-week important updates for more details.


Other Organizational Stuff

Project Leader: Doram

Status: Update In-Progress

Status Date: 5/7/2008

Details:

Building wiki pages as fast as possible... need to learn this semantic stuff better...


Full Weekly Report from Project Leads

Last update: May 5, 2008

Doram

Current general consensus:

  • phpbb/mail2forum is too much overhead in security right now.
  • Wiki pages have individual, essentially tasked, "discussion" areas, which can serve as generally organized discussion space.
  • Further organizational efforts would best serve in the Wiki so as to be properly connected to Wiki entry based project discussion/flow
  • Semantic part of Wiki is solid, even if further layers are not quite stable (or capable of stabilizing) just yet (HALO, gardening, etc.)
  • SKDB/OSCOMAK/Whatever is going to continue slow parallel growth, but is not ready for integration yet, Needs Debian and agx-get support both within our system, and within the respective source communities.
  • We are just at the point where we really need more contributions to get things rolling properly.

What this means:

  • New wiki pages, mostly maintained by me - Project/Subprojects Status, News, and Calls to Action/Calls for Help
  • New Google Group Page - duplicate of Wiki news page (by me) - both for access to people not ready/aware of the Wiki, and for RSS feed support for those interested in continuing to lurk.
  • After, a few support pages are figured out within the Wiki structure (I will analyze), notice to Google Group "forum" to start heading over to the Wiki and start contributing.
  • Soon - Update to HTML page at OpenVirgle.net to reflect new internal structure/flow, plus more links to relevant info. Plus rework of existing Google Group Pages to interlink with Wiki resources.
  • SKDB/OSCOMAK/Whatever both needs help and a name. First call for help is Debian and agx-get people. Plus, possible "Let's name this thing" contest for group members... or we could just call it the OpenVirgle Framework?? Comments/flaming arrows?

Mike

Currently I'm just capriciously creating and supplementing wiki articles and researching.

I added some thoughts on how we could keep track of future projects here: http://www.oscomak.net/wiki/WikiProject_OpenVirgle

One of the great things about the smw is that you can have multiple categories, so if you decided to do a project called, say, influenza transmission simulation, you could create a project page for coordinating it and its resources and put it in the appropriate category, like health risks, and also the "project" category. Then in the description property you could put a general status like "started, incomplete" and make a membership property if you'd like. That way we have, in effect, a self-updating project log in the ontology browser. Of course, some projects, like the ones we are engaged in now, are best coordinated through a post like this.

Paul

Thanks for bringing this up.

The server upgrade won't affect halo performance while editing as that is limited by the speed of JavaScript on the client (browser's) computer. It may (but likely not) fix the gardening bot issue (if a local MySQL database has more permissions like with the "func" table than a remote one at Pair Networks -- somehow I doubt that, although they may be willing to grant more permissions in the special case of a dedicated server).

I'm working towards adding articles and understanding related ontological and tagging issues (following Mike's example. :-) And I continue to appreciate feedback on that.

Bryan is right when he says there are ontological (tagging) issues that remain to be solved for better organization of manufacturing information. But I hope, to the extent Semantic MediaWiki can support them, that we can work these through as we go -- in part because adding real content forces us to think about these issues (as opposed to focusing on programming more infrastructure.) But there is still a lot of thinking to be done on that.

I have no time estimates -- other than, the rest of my life? (Not to imply it would be done by then, either. :-) Or, more realistically, until I might need to go back to work full-time :-( (but that is unlikely for the next year or two).

I also think Mike is right that the Wiki can easily be used to coordinate projects. And any help on writing articles can mostly happen there in an organic way.

From when I ran my own web server (with root access) I know that there are a vast number of daily attacks against any popular software, like phpbb, so I am not eager to install or maintain anything on the server which is non-critical and needs care and feeding, sorry. (I feel that way about our chickens sometimes too. :-) Worrying about the MediaWiki and related security patches etc. seems like chores enough for now. :-) However, that does not mean I am absolutely opposed to new things on the server. I just don't see the cost benefit right now. For example, does Wikipedia also use phpbb? I don't think so. They do have extra mailing lists though. And down the road I am not opposed in any way to more lists, or even moving off of Google groups entirely due to the lack of downloadable full archives. Also, on phpbb, now that we have semantic tagging with the Wiki, I think we would find forums less useful because content could not be semantically tagged after the fact, unlike Wiki articles. I'd suggest we give the Wiki a try for a couple months, and if we still feel like a forum would be helpful, we could revisit the issue.

I'm going to also eventually talk to Pair about whether the server can have a second account on it for secure experimentation (like learning PHP) which would not affect the primary account. If worst comes to worst, and Pair won't or can't do that (actually, it may fall under their "reseller" concept?) another shared account just for openvirgle.net is not out of the question for me to setup (since the OpenVirgle and OSCOMAK may overlap, but they are not identical, and I do not want to stand in the way of OpenVirgle having separate forums etc. if that is what most people want.) Again though, I feel focusing on adding content in the MediaWiki, including project planning coordination, may be able to meet the needs which forums have supported in the past. And I expect, as Mike suggests, even this Google mailing list will then eventually be used either for higher level project coordination discussions or perhaps more social or introductory aspects. Or, as a sort of "scaffolding" we might even in a sense take down when something better is running (virtually, since this list will likely always be here at Google Groups, but we could regularly post, see the Wiki, or see the forum, or see this other mailing list).

Bryan

On Sun, May 4, 2008 at 7:51 PM, Doram <DoramBaram...@gmail.com> wrote: > What are you working on right now (for OpenVirgle)? (Short > description, full list)

You'll have to excuse my apparent lack of progress over the past few days; I came across some technical difficulties involving 150 volts and 2.5 amps (yes, not milliamps). But otherwise, the prototype client interface (in python) is going smoothly, it's actually to the point where I need some help deciding where to drop files on the user's computer -- where will the packages go? What about the metadata files? What's the usual directory structure in debian for this? And what about the possibility of using makefile? That has to be investigated. Come to think of it, a debian programmer onboard would really, really help at this point. Second, after that is solved for agx-get, I need to do some sample files in the 'metadata format', some sample scripts, and then have Ben do his autospec implementation. The 'ghost aggregation layer' aspects will come in at this point, as we write up some tutorials on how to join your git repos into 'metarepo', how to crawl them, and so on. But before this happens, I need some confirmation from the debian deity team re: the new architecture for debtorrents. You see, they are looking for a new software infrastructure since they have this ridiculously awesome load on their servers because of the many tens of thousands of people using apt-get daily. So somebody was proposing a meta-project of sorts that solves that problem as well as a few other similar problems at the same time, and it's basically the 'metarepo' idea.

http://heybryan.org/exp.html (click the 'review' link to see the debtorrents side of the story)

> How far along are you with each? (Wild gross estimates are fine)

Good question. Far enough along that I know that there's only documents and people that I need to read and talk with at this point. I'd be interested in organizing more clearly with the debian deity revamp project team members before I do anything stupid with the codebase.

> What do you need help with? (if anything)

See above. Hunting down a debian developer, contacting the revamp project team members, that'd be great.

> As far as I can tell, future SKDB/OSCOMAK/Whatever systems seem like > they will take a bit more time to finish writing the back end for > (Bryan, confirm, please), but should be able to fairly easily digest > the Semantic Wiki data.

Heh, the only reason they may take some more time is because this month represents the month of testing for me, and without my typical computer setup my workflow is significantly disrupted, and at the end of the month I immediately start a lab position, but at which point I will probably have enough time to read the documents I've been hoarding.

> With both of those parts of the project reaching a stabilization point > of sorts, it may be time to ramp up contributions of members of > Openvirgle that are not the four of us. I would like Bryan's and > Paul's opinions on whether or not we should try installing phpbb/ > mail2forum on OpenVirgle.net in the hopes of future easy absorption of > forum/email/discussion into SKDB/OSCOMAK/Whatever (plus, bringing it > officially under our control). Now would be the time to make that > switch from Google owned servers to ours, before we start getting the > rest of the group riled up and contributing.

Paul's comments re: phpbb are basically correct, the system does in fact get lots of spam, and if you're not careful it will be taken over. But you see, phpbb can be installed nearly anywhere, so go ahead and do it, install the mail2forum gateway, and then lots and lots of captchas and other silly systems [while also maintaining a link to access Google Groups, since Google has their own layer of spam protection].

> Please let me know what you guys think. Maybe we could agree to a > weekly update for the news thing. Not too often, not too distant. Of > course, you are free to tell me that I out of my mind, too. All > comments welcome...

Hm, weekly news. You may be interested in: http://heybryan.org/~bbishop/cgi-bin/blosxom.cgi/ (my blog)

- Bryan

Mid-week important updates

Server report from Paul

5/4 3:13PM

OK, I just can't stand the idea of OSCOMAK not standing up to a slashdotting, if we should be so lucky someday. Plus, while I don't think the Wiki is ready for promoting much (for example, the issues that Mike lists)

 http://www.oscomak.net/wiki/Reported_Errors

I want to be optimistic for a couple months from now.

So, based on my optimism of the server being driven into the ground someday,

-) I just put in an order at Pair Networks to upgrade the shared account to

their lowest cost standard dedicated server (QS-1).

 http://www.pair.com/services/dedicated/
 "Core 2 Duo E4400 (2GHz) CPU ... 120GB Hard Drive ... Daily back-ups"

I don't know yet when the changeover will occur, presumably sometime this week. From what I understand, Pair does these things with minimal downtime, so hopefully no one will notice the change, except things might get a little snappier with OSCOMAK.net and the OpenVirgle.net pages. But should the sites be unavailable for half-an-hour or so, that's probably why. Or, of course, longer if something unexpected goes wrong, as it almost always does. :-)

5/7 3:37PM

The new server is online and seems snappier, both for openvirgle.net and oscomak.net.

Please let me know if any issues related to the switch over show up (the database was also moved, though it seems to be OK).

There is also a memory upgrade likely to happen tonight to (maybe) 4GB, with additional minor downtime.

5/7 4:04PM

I just gave Pair Networks permission to do the memory upgrade now, so if the server is down later, that is the reason. What are the plans for 4GB of memory on the server, you ask? The answer is, I don't know for sure. :-)

5/7 4:19PM

OK, I barely had time to finish writing that second Lost in Space joke and Pair says the upgrade to 4GB and reboot is complete. :-)

%top

last pid: 1447; load averages: 0.00, 0.00, 0.00 up 0+00:14:47 15:58:46 104 processes: 1 running, 102 sleeping, 1 zombie CPU states: 0.0% user, 0.0% nice, 0.0% system, 0.2% interrupt, 99.8% idle Mem: 246M Active, 46M Inact, 74M Wired, 12K Cache, 68M Buf, 3270M Free Swap: 2048M Total, 2048M Free (Hmmm .. Mem doesn't quite add up? About 200MB missing for what? Oh well. Good enough. Probably something I don't understand. :-)

%df Filesystem 1K-blocks Used Avail Capacity Mounted on /dev/ad0... 138673860 7062474 120517478 6% /usr

Anyway. nothing like an idle CPU core or two, an empty 3GB+ of memory, and about 120GB of backed up free disk space to fire the imagination. :-)

Or in other words, how can I now retroactively justify all this without total embarrassment? :-)

--Paul Fernhout