[openi18n-im:01197] user-running iiim server discussion
Leon Ho <[email protected]>
| Newsgroups | gmane.comp.internationalization.input-methods |
|---|---|
| Message-ID | <[email protected]> |
Hi Kazunori san, Great idea. Your implementation idea actually will help from not running iiim service in boot time. So it will help typical desktop user who has one desktop/user per system. And then next implementation will be in gimlet, which it can pull in private iiim server if the preference "Enable Input methods" is enabled and normal iiim service is not running at gimlet startup. I think my another intention is similar but just slightly different - a user iiim server to share with other users. It serves some systems with multiple desktops/users. Cheers, Leon On Fri, 2005-02-25 at 17:18 +0900, AShiZaWa KaZuNoRi wrote: > Date: Fri, 25 Feb 2005 17:38:11 +1000 > From: Leon Ho <[email protected]>: > > > I'm also interested in non-root iiim server in your list. > > > > > My intension is to provide a private iiim server, not shared > with other users, which is required by some secure system. The > private iiim server also reduce the risk of single point of > failure, which the current implementation suffers. > > > > > Let me tell you some people suggestion on it and see if it is possible > > or not: > > > > Different type of desktop users usually want to enable (start) iiim > > server when they needed. Currently we are starting this in boot time, > > and it costs around 40MB of memory (roughly). > > > > Typical desktop user only have 256-512MB RAM. So definitely we don't > > want to run the server when no user(s) is using input. > > > > So if running in any user is possible, then it should resolve this > > problem by delay to run before it is needed, specially for that > > configuration. > > > > Other server configuration etc can run server at startup at normal. > > All seem reasonable and implementable. > > If possible, please post how implementation will be. I will > write and post my proposal next week. > > Regards,