[00:41:44 EDT(-0400)] * jayshao (n=jayshao@ool-44c59034.dyn.optonline.net) has joined ##uportal
[09:16:47 EDT(-0400)] * esm (n=esm@clue.mse.jhu.edu) has joined ##uportal
[09:36:55 EDT(-0400)] * jayshao (n=jayshao@jshao.rutgers.edu) has joined ##uportal
[09:43:41 EDT(-0400)] * EricDalquist (n=dalquist@bohemia.doit.wisc.edu) has joined ##uportal
[09:58:02 EDT(-0400)] * EricDalquist (n=dalquist@bohemia.doit.wisc.edu) has joined ##uportal
[10:10:39 EDT(-0400)] <EricDalquist> uhg
[10:10:50 EDT(-0400)] <EricDalquist> so this svn pruning is not going to be very easy
[10:17:17 EDT(-0400)] <esm> 
[10:17:40 EDT(-0400)] <esm> Chris D. has more experience than I do with dumpfilter
[10:17:45 EDT(-0400)] <EricDalquist> the bad rev we ran into back in r2475 is causing lots of problems
[10:17:50 EDT(-0400)] <EricDalquist> oh dumpfilter is cake
[10:17:59 EDT(-0400)] <esm> oh ok good
[10:18:01 EDT(-0400)] <EricDalquist> problem is I can't do a dump/filter/load
[10:18:13 EDT(-0400)] <EricDalquist> without changing all the rev #s
[10:18:18 EDT(-0400)] <EricDalquist> which would be BAD
[10:18:53 EDT(-0400)] <esm> ok i'll be devils advocate
[10:18:55 EDT(-0400)] <esm> why is it bad
[10:19:13 EDT(-0400)] <EricDalquist> well
[10:19:17 EDT(-0400)] <EricDalquist> not really that bad
[10:19:23 EDT(-0400)] <esm> we have references to revisions in jira issues - in comments, not the fisheye plugin
[10:19:26 EDT(-0400)] <EricDalquist> but anyone who has anything checked out will have to re-checkout
[10:19:36 EDT(-0400)] <EricDalquist> in would invalidate every checked out version
[10:19:45 EDT(-0400)] <EricDalquist> and make any local changes very difficult to merge back in
[10:19:55 EDT(-0400)] <EricDalquist> it would*
[10:20:03 EDT(-0400)] <esm> ok so that can be avoided with some announcments/planning
[10:20:16 EDT(-0400)] <EricDalquist> yeah
[10:20:19 EDT(-0400)] <EricDalquist> so we'll see
[10:20:23 EDT(-0400)] <EricDalquist> I still may be able to get around it
[10:21:05 EDT(-0400)] <EricDalquist> I found a tool for fixing corrup revs in the fsfs system
[10:21:15 EDT(-0400)] <esm> yeah. I guess the good thing is that mostly only up3 people are affected, the portlets and other stuff just got migrated
[10:21:19 EDT(-0400)] <esm> nice
[10:21:25 EDT(-0400)] <EricDalquist> so I did a filesystem copy of the repo last night to my local box and will be trying it out there
[10:21:39 EDT(-0400)] <esm> nice
[10:21:48 EDT(-0400)] <esm> i found my problem with assembly, now to fix it
[10:21:48 EDT(-0400)] <EricDalquist> if I can get around this my dumpfitler stragegy looks like it will work nicely
[10:21:52 EDT(-0400)] <esm> it was a buffer balooning
[10:21:53 EDT(-0400)] <EricDalquist> cool
[10:21:58 EDT(-0400)] <esm> nice
[10:22:02 EDT(-0400)] <EricDalquist> where in the code?
[10:22:13 EDT(-0400)] <esm> not yours (of course) it was mine.
[10:22:29 EDT(-0400)] <esm> I provide a ByteArrayOutputStream to store the results of war assembly
[10:22:48 EDT(-0400)] <esm> and the buffer for BAOS grows a power of 2 each time the buffer would overflow
[10:23:07 EDT(-0400)] <EricDalquist> nice
[10:23:10 EDT(-0400)] <esm> but I have a fix in my head
[10:23:27 EDT(-0400)] <esm> so i'll try to do it this afternoon but JHU stuff is calling right now
[10:28:31 EDT(-0400)] <EricDalquist> sweet!!!
[10:28:45 EDT(-0400)] <EricDalquist> looks like this fsfsverify tool can fix that bad rev!
[10:28:56 EDT(-0400)] <esm> awesome
[11:15:36 EDT(-0400)] <EricDalquist> I REALLY don't like out of office replies
[11:15:46 EDT(-0400)] <EricDalquist> I've gotten 15 so far from that email to jasig-announce
[11:51:28 EDT(-0400)] <EricDalquist> yay ... looks like the dump/filter/load is going to work quite well
[11:51:31 EDT(-0400)] <EricDalquist> and preserve rev #s
[11:51:52 EDT(-0400)] <EricDalquist> we're going to go from a repo of ~ 900MB to ~ 300MB
[11:54:38 EDT(-0400)] * colinclark (n=atrcwrk2@142.150.154.101) has joined ##uportal
General
Content
Integrations