[comp.mail.elm] HELP <- ELM ALIAS PROBLEM

mwitten@ut-emx.cc.utexas.edu (matthew witten) (02/05/91)

I could use some advice/help/fixes for the following problem.
I am using ELM on a SUN. For some reason, ELM has a present 
number of aliases that can be entered. In this case approximately
250. When you exceed that, you get an error message. However,
ELM will not rebuild the .hash and .data files for the aliases.
This, in spite of the fact that it will save aliases in the .TEXT
file. Thus, while alias 251 may exist in the .TEXT file it will
not exist in any of the hash or data tables. Further, ELM does
not rebuild the tables. As a consequence of this action,
all of the aliases become unusable. 

Does anyone have a fix for such a problem. I'd hate to have
to constantly GREP from another file to get my addresses.

I need to be able to store a large (~1,000) number of aliases.


Matthew Witten, Ph.D.
Associate Director,
UT System,CHPC

M.WITTEN@HERMES.CHPC.UTEXAS.EDU
M.WITTEN@UTCHPC.BITNET

syd@DSI.COM (Syd Weinstein) (02/05/91)

mwitten@ut-emx.cc.utexas.edu (matthew witten) writes:
>I could use some advice/help/fixes for the following problem.
>I am using ELM on a SUN. For some reason, ELM has a present 
>number of aliases that can be entered. In this case approximately
>250. When you exceed that, you get an error message. However,
>ELM will not rebuild the .hash and .data files for the aliases.
>This, in spite of the fact that it will save aliases in the .TEXT
>file. Thus, while alias 251 may exist in the .TEXT file it will
>not exist in any of the hash or data tables. Further, ELM does
>not rebuild the tables. As a consequence of this action,
>all of the aliases become unusable. 
Yup, the error is fatal...

>I need to be able to store a large (~1,000) number of aliases.
Contact the person who installed Elm on your system and
see if they agree you need 1000 aliases, if so, have them
recompile Elm with a larger alias table, its one of the
configure options in Elm.  Its a memory tradeoff in 2.3.
Hopefully in 2.4 we will have a different algorithm in place
that won't care about the hash table size.
(its been talked about, but I've yet to see the code :-) )

-- 
=====================================================================
Sydney S. Weinstein, CDP, CCP                   Elm Coordinator
Datacomp Systems, Inc.                          Voice: (215) 947-9900
syd@DSI.COM or dsinc!syd                        FAX:   (215) 938-0235