trevor@trevan.uucp (Trevor J. Harris) (10/01/89)
I am having problems with the batcher getting out of sync with about 20 percent of my incoming batches. I just 'mv' these from the bad directory back into in.coming and they go through ok. I am runing cnews on Microport V/AT 2.4 with the small model and dbz with the following changes: 1) Fixed hdrdefs.c to compile with Microport by applying patches posted to net. 2) In rna/lib/makefile made RANLIB = echo 3) Used microport stdio. 4) Added own spacefor 5) I use a 1k buffer in expire.c I would like to hear from anyone with a similar setup. regards trevor -- ------------------------------------------------------------------------- Trevor J. Harris ukc!trevan!trevor
ruiu@dragos.uucp (dragos) (10/04/89)
In article <1989Oct1.094543.14743@trevan.uucp> trevor@trevan.uucp (Trevor J. Harris) writes: >I am having problems with the batcher getting out of sync with about >20 percent of my incoming batches. I just 'mv' these from the bad directory >back into in.coming and they go through ok. >I am runing cnews on Microport V/AT 2.4 with the small model and dbz with >the following changes: >Trevor J. Harris ukc!trevan!trevor I was having this problem. I eliminated it by using a large model relaynews. I never tracked down the problem fully once I found a solution. I just built a duplicate source tree where I put the large model stuff. This way I can still patch the sources and keep separate versions of the cnews libraries. NOTE: the three fast stdio routines do not pass the tests in the large model, so you should make sure to avoid linking them in. -- Dragos Ruiu | Programming Languages: alberta!dragos!ruiu | uunet!myrias!dr | find / -name \*\[fF\]77\* -exec rm -rf {} \; bbs: (403) 439-0229 |