LSTOWN-L Archives

LISTSERV List Owners' Forum

LSTOWN-L

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Topic: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Eric Thomas <[log in to unmask]>
Thu, 29 Jun 1995 07:44:20 +0200
text/plain (55 lines)
On Thu,  29 Jun  1995 01:03:42  -0400 Vickie  Banks <[log in to unmask]>
said:
 
>Seems like this must be a common  problem, and there must be a number of
>versions of a UNIX-based UNIX conversion  program out there, but I'm not
>sure what to look for. Any suggestions would be appreciated.
 
It shouldn't  be difficult to do,  but, never having had  that problem, I
have no idea where you could find that.
 
>Eric --  Any plans to  have the UNIX Listservs  offer an option  of UNIX
>format to people ordering list archives?
 
Since the  files are  stored in  a different  format, this  would require
either a new command called "GET_ARCHIVE_WITH_UNIX_MAILBOX_CONVERSION" or
the addition of  a new option for  the F= keyword, which  would only work
for list archives and would perform  the necessary conversion on the fly.
The problem  is that the  F= file  formats weren't designed  as something
that can only work  on a very specific type of  file. They're supposed to
accept arbitrary input, with the  possible exception of very exotic cases
like special system  files that in principle LISTSERV  shouldn't even try
to send and encode.  The only example I can think  of is a non-sequential
VMS file.  This can't be represented  as, say, Netdata. You  get an error
traceback that normal users won't understand, but then normal users won't
know what a "non-sequential file" is anyway, so it's no big deal :-)
 
But anyway, this is a  mainframe-thinking type of solution. Unix problems
call  for unix  solutions. A  simple unix  solution is  to pipe  the mail
message through  a simple yet powerful  filter which converts on  the fly
and  creates a  mail  folder  with all  sentences  beginning with  "From"
properly mangled to ">From" and the  "From" lines inserted. You can write
this filter in a variety of simple yet powerful languages including C and
perl. Now if  you combine this unix thinking with  the mainframe approach
of using  specialized controllers  for specialized  tasks to  offload the
central CPU, you arrive at the following algorithm:
 
1. Locate local unix student, preferably bored specimen, or specimen busy
   reading usenet.
 
2. Open umbrella.
 
3. State  that big time mainframe  bigot XYZ asserted that  this can't be
   done on unix in  less than 3 days of work because  there is no support
   for an advanced file system  on unix and extensive application support
   code needs to be engineered into  the program before the first line of
   application code can ever get written.
 
4. Close umbrella when expostulations are over.
 
5. Come back 30 minutes later and collect program.
 
:-)
 
  Eric

ATOM RSS1 RSS2