[Info-vax] OpenVMS - DCL - Data entry filtering

[mcleanjoh at gmail.com]

On Monday, March 30, 2015 at 11:08:06 AM UTC+11, JF Mezei wrote:
> On 15-03-29 18:03, Stephen Hoffman wrote:
> 
> > MAIL is a (good?) (bad?) example of what happens when you don't have a 
> > modern database, when you do have to work within the limitations on how 
> > fast you can scan directories (and this also from within a 
> 
> 
> But MAIL was originally designed like NOTES, with the expectation that
> most messages would be stored in the main indexed MAIL.MAI file.
> 
> The size limitations that triggered external storage were fine
> originally, but once just about every internet email exceeded that size
> limit, the mail store morphed into something it had not been designed
> for with every email stored externally.
> 
> 
> ALL-IN-1 had an interesting design. Created a database of email documeht
> attributes with pointer to random filename stored in a "random"
> directory. It was designed that way to distribute the tousands of files
> amongst a number of directories to avoid the performance problems of
> large directories.
> 
> It also had the advantage of having 1 copy of an email sent to multiple
> recipients. Each had a pointed to the shared copy. It only got deleted
> when the shared count dropped to 0.
> 
> Note that Unix mail stores are not exactly very efficient either. But
> they seem to work and scale.
> 
> dovecit for instalce uses strange file names on OS-X for each message:
> "1404580067.M555292P14187.velo.vaxination.ca,W=2670,S=2621:2,RSa"
> 
> But it files message files in folders that correspond to IMAP folder
> names. and there is some binary index to each file.
> 
> 
> VMS designers have to decide whether to go with the flow and just make a
> file system very efficient for large number of files (turning the file
> system into a database) or  provide database tools that allow apps to
> store large number of "things" in a single file instead of one file per
> "thing".

Anyone who has all of their emails in one Mail folder is asking for trouble.  Not only are file operations slow but scrolling up and down the list trying to find something is a pain.  It's far better to use subfolders and divide the emails up by some criteria (eg. source, subject or even date).

I agree with JF that Allin1 was smarter than Mail because it used multiple subdirectories.  Another time that I've seen this done was with the Ada development environment on VMS.