I like the idea of working on this, especially as a set of Cornerstone services used by a UUIDService to generate UUIDs. At this time however I am just using vmid to fill my immediate need. I have two apps that need to be finished before I do the UUID services. If you don't mind waiting I'll work with you on this.
Bill Barnhill
Tim Reilly wrote:
Hello Bill,
I started on something quite a few months ago: commons-lang welcomed adapting a UUID implementation. The problem I ran into was - the consensus at the time was to use the Tyrex UUID for various reasons (also found in Castor with slight changes.) And elsewhere in exolab projects. I started the discussion wanting to get the Axis uuid moved into commons. http://marc.theaimsgroup.com/?l=jakarta-commons-dev&w=2&r=1&s=UUID&q=b
But since I want this as a pure Apache license class for the sake of reuse in apache projects I did not know how to take from one license source Exolab and put under ASL. I tried contacting Assaf Arkin; rather [EMAIL PROTECTED] to see about a contribution put no luck. -- Would you leave the original license in the file? But then all ASF source must have the ASF license. Would you pre-pend both licenses? Add the Tyrex license to the distribution? Reduce a modified Tyrex jar to distribute with only the 3 classes and the license and uuid.properties? And call it something else? -- FYI just noticed Turbine has org.apache.turbine.util.db.UUIdGenerator that allows for a properties file with the MAC address. (That was one of the requests from commons-lang.) -- When I didn't get answers to these questions I toyed with creating an implementation modeled after the Tyrex version. I planned to allow the clock thread to die after some timeout between calls. No idea if that's a smart thing or not; I also didn't understand the need to re-synchronize with the System time? Can threads sleep longer than you say tell them - I mean I know load and thread priority come into play, but I thought not that long. Anyhow, not feeling confident in what I was doing I decided to abandon it.
I know attachments sometimes don't make it, but I'll try to attach a rough **draft** here. If you have interest in continuing this feel free or we can work on it together. I don't have a much time but I can make some somewhere and I'll learn some stuff. It would be nice to follow through on the original thought and get something to commons-lang identifiers; so all these UUID that are about might live in commons.
Let me know your thoughts. -TR
-----Original Message----- From: Barnhill William [mailto:[EMAIL PROTECTED] Sent: Friday, December 05, 2003 7:10 PM To: Jetspeed Developers List Subject: A copyright question
Here is a copyright question...
I came across some code for generating a guid unique over IEEE 802 address space and a time of 35 years. The URL is http://www.theserverside.com/patterns/thread.jsp?thread_id=4976 The author is listed as Adrian Maclean, but I could find no way of contacting them. The code implements a standard IEEE algorithm.
What is the viability of using including this code in a greater whole and donating the whole to the ASF?
I'm guessing it's a no-go, but could use some info from more license-aware people.
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
------------------------------------------------------------------------
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
