[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Xv 3.10a-11 (i386) uploaded to master



In article <[🔎] xe1n2seth0j.fsf@maneki-neko.cygnus.com> you write:
>
>> <sigh> Why? This has to be one of the most annoying things about most of
>> the packages using mime-support.
>
>Indeed.  There's been some discussion of this in the context of some
>of the bug reports I've filed, namely that this way isn't right either
>because if the string used changes, the command lines end up
>accumulating in /etc/mailcap.  [Frankly, I find the behaviour annoying
>enough anyway that I'd put up with that :-) Disk space for one small
>file is a lot cheaper than my frustration from answering the same
>question again and again and again and again...]

NO NO NO. This is why lynx, Netscape et al are taking 30 secs to start 
up - there are so many testndisplays to do that the process stops in 
its tracks while it sorts it out. 

>There were some ideas for adding "memory" to the process; I just
>noticed that install-mime is a fairly clear perl script so I might
>just try implementing some of them.  (For example, having a
>/var/lib/mime where relative priorites are cached... or maybe even an
>init.d style directory which is used for the real work and converted
>into a mailcap :-)

A cache is a good idea. A per-login cache would be even better (so 
that e.g. lynx knows it's running on the console and doesn't have 
to test $DISPLAY 80 times before starting up.

We _must_ fix this, it makes the whole system so much slower than 
pre-mime-support systems. 

Jon.


Reply to: