Re[2]: [xmlblaster] Queue and starving

[Michele Ravani <Michele.Ravani at swissonline.ch>]

On Tue, 10 Dec 2002 19:58:43 +0100 Michele Laghi <laghi at swissinfo.org> wrote:

ML> Ciao Michele,
ML> in fact there is no mechanism to avoid such "starving" of low priority 
ML> messages. I believe it is a correct behaviour to priorize high
ML> priority 
ML> messages. If there are no free resources available to sent all
ML> messages, 
ML> then it is correct to hold back low priority messages.

ML> Of course you could discuss the fact that messages which are in the 
ML> queue since a long time become a higher priority but this is a
ML> specific 
ML> behaviour (you could have other such behaviours) which could be solved 
ML> with a specific plugin.

Yes, I thought about that.
Some time based or 'number of fetch' operations based mechanism could be
implemented which increases the priority of the messages still in the
queue.

Ciao

-- 
Michele Ravani                  Michele.Ravani at swissonline.ch
"Those who live hoping, die singing" My Gran