Erik Thanks. It turns out that I had experimentally changed IndexWriter.BUFFER_SIZE from 1024 to 4096 (just looking for a bit of performance), which upset RAMDirectory which I wasn't aware was used implicitly. Anyway close the bug as 'cannot reproduce'.
EJP ----- Original Message ----- From: "Erik Hatcher" <[EMAIL PROTECTED]> To: "Lucene Developers List" <[EMAIL PROTECTED]> Sent: Thursday, November 20, 2003 1:11 AM Subject: Re: Lucene bug 24237 - content-dependent indexing failure > On Wednesday, November 19, 2003, at 12:03 AM, Esmond Pitt wrote: > > I'm not sure whether anybody is still waiting for me to advise that I > > had > > uploaded a corrected test case for this, but I did this & reopened the > > bug a > > couple of weeks ago (4th Nov), and haven't seen any progress on the > > bug - > > it's not even assigned yet. Seems like a problem worth investigating > > to me, > > and it will become important for us around the end of the month. > > Esmond - I tried your sample code with the URL you specified and it > worked fine for me with the CVS version of Lucene. It is not that > different from the 1.3RC versions, but perhaps something has fixed it > (although I'm skeptical). I'd suspect some other issue with your > environment or code, but first try a build from HEAD of Lucene's CVS > and see if it is fixed. > > Erik > > > --------------------------------------------------------------------- > 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]
