[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
a
From: |
Luke Kenneth Casson Leighton |
Subject: |
a |
Date: |
Sat, 14 Jan 2006 01:14:58 +0000 |
User-agent: |
Mutt/1.5.5.1+cvs20040105i |
> - popup window to answer/cancel the call
> - addressbook summary of the contact appearing
> - calendar/todo/notes related to this contact appearing
> - ...
> I don't think it is easily doable with RPC.
it's _very_ easy with DCE/RPC!!
you use UDP or any other broadcast-based transport as the transport!
> With D-BUS you don't need to care of all this endiannes shit.
endian-ness is _totally_ below the level at which dce/rpc is used.
the dce/rpc runtime _takes care_ of endian-ness - that's part of
its job.
when you use the IDL compiler, you aren't dealing with "endian-ness",
you're dealing with functions - almost c-code except with essential
specifiers like [in], [out] and [length_is()] which help save data
bandwidth.
d-bus is _very_ very pointless.
if it came with an IDL compiler, i would understand, rave
about it, complain about how deeply unsatisfying the IDL syntax, was
compared to freedce, but i would probably actually _use_ it.
but there's really _nothing_ in d-bus that a couple of thousand
lines of code won't replace, and there's _ab_solutely no advantage for
it.
i am _so_ angry with the dickheads who created d-bus.
the opportunity to put in an IDL compiler was lost the moment they
specified that it was okay to use those _stupid_ data parsing functions
that even _python_ does better with a) repr and eval b) with that funny
%-specified data packing stuff it has.
why lost? because now you have dozens of projects that use those
pathetic data parsing functions, you now can't go back: you can't
retrofit an IDL compiler - the _one_ thing that would make d-bus
bearable - without having those dozens of projects have to rewrite all
their code.
anyway. this is a bit off-topic and it's unfair to burden gnokii users
with it.
--
--
<a href="http://lkcl.net">http://lkcl.net</a>
--
[Prev in Thread] |
Current Thread |
[Next in Thread] |
- a,
Luke Kenneth Casson Leighton <=