Contact:cdent@burningchrome.com
I posted to Freshmeat last night and people are poking around, but noone is actually downloading the software. This suggests to me that I've not done a very good job explaining why people might want to set up their own Arts repository. So I'll explain how it got started at Kiva and maybe that will help. A couple of years ago Kiva was a very small company and nearly all the employees were on one mailing list and all things technical were discussed there. Anything that anyone might need to know was distributed by that mailing list and pretty much everyone was in the loop. As the company grew, volume on that list got out of hand, so it was split up into smaller lists. People then fell out of loops and information was being lost. Someone suggested that relevant messages be cc'd to shared mail folders that people could read for more info later. This was helpful for historical data but not very good for documentation purposes. Everyone knows that _great_ documentation shows up in email conversations when it is least expected. Someone asks a question and someone else answers it in email and there it sits. That stuff ought to be saved somewhere. One day, a co-worker, Matt Liggett, suggested that we should set up mail addresses that could be Bcc: or bounced to whenever something good showed up in the mail. I thought it was a good idea but sat on it for a few months. Then in a flurry of activity I squirted out the first version of blackarts and the systems department started building a repository of arcane knowledge that helped us get our job done. Soon people in other departments became jealous and wanted ways to store information themselves and things grew. Kiva now has blackarts, greyarts, yellowarts, headarts, projarts, softarts, custarts, bluearts with something over 500 documents, all index by webglimpse. All of the documents began life as a mail message that was posted to some people and either also Bcc'd to the arts or one of the recipients said, "Hey, this should go in blackarts." As things are currently configured, all the documents are managed by the webwatcher system (see: http://systhug.com/webwatcher/ so that any document that goes into the system is announced for revision in 15 days so that corrections and adjustments can be made. Then the document remains in webwatcher with whatever expiration settings are desired. This means that documents don't get lost forever: people are reminded every now and again that they should take a poke at stuff to make sure it is not stale. Our latest thought with the system is to create something that allows easy generation of simple FAQs that are crosslinked. Something like a faqarts where there are sections based on particular software tools, machine or services. Some tools are closely related to particular machines so those questions would be crosslinked between the tool section and the machine section. That idea is what prompted the latest revision of arts and the repackaging of the software so that it could be used elsewhere. Back to the Index