Yep! That was it...I just ran it this morning and all was fine :) I missed the "PM" for the 10 when I looked at the timestamp yesterday of the plugins within maven-1.0.2/plugins. Sorry about that, that would have answered the question yesterday when you asked me to compare the timestamps!
Chad -----Original Message----- From: Brett Porter [mailto:[EMAIL PROTECTED] Sent: Wednesday, December 08, 2004 2:10 AM To: Maven Users List Subject: Re: 1.0.2 startup slow How funny! Being in +11 at the moment, that certainly made it as bad as it could be. Thanks Leif! Ok, I'll either set my clock back or build with significant lead time in future... Chad - can you confim it was the same for you? Thanks, Brett On Wed, 08 Dec 2004 01:01:17 -0700, Leif Nelson <[EMAIL PROTECTED]> wrote: > I think I figured out what the problem is. When I did what you suggest > (checkout the tagged version of maven, add logging, build and test) it > works fine. 2 seconds on windows xp. So, on a hunch, I looked in the > .zip file for maven-1.0.2.zip. And guess what, the date/time stamp of > all the files in there is for 12/7/2004 at 10:13 PM. Well, us "early > adopters" were running 1.0.2 before 10:13 PM on 12/7/04! The code in > PluginManager.unpackPlugin() compares the date/time of the plugin.jar > file with the date/time of the plugin cache dir. And of course, every > time, it found that the jar was newer because it had a future date/time > stamp, causing a cache invalidation for every invocation. > > So, for anyone who had this problem on 12/7, it will magically fix > itself when their local time ticks past 10:13 PM. So, is the time wrong > on the box that created the maven-1.0.2.zip file? Or, are zip files > dumb enough to not contain a GMT offset? Who knows. > > Anyway, I'd bet the problem goes away now. I re-unzipped the > distribution zip file, and since it's 12/8/04 12:58AM my time now, it > takes 2 seconds to run maven on my windows box. > > Hope that helps, > --Leif > > > > Brett Porter wrote: > > >Ok, this has got me stumped. I don't know what's different. > > > >Let me outline where we are at: > >- first run is always slower > >- second run should not unpack any plugins or touch the cache > >- the reason that blowing away the cache is faster than running with > >it when the unpacking is happening is because it is actually manually > >erasing each, then recaching instead of just the last step, so that's > >fine. > > > >We need to sort out what is triggering the unpacking. > > > >Is anyone experiencing this problem comfortable with building from tag > >MAVEN-1_0_2, then adding debug statements to PluginManager.java and > >recompiling? > >If not, I can do this and provide a debugging maven.jar to try out... > > > >Can someone let me know (Chad?) when you are able to look at this, and > >if possible jump onto irc.codehaus.org (you can get there by HTTP if > >needed) so we can walk through it. > > > >Thanks, > >Brett > > > >--------------------------------------------------------------------- > >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] --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
