Re: Calling handlers manually
"Conan C. Albrecht" <conan-RXmsyT/Szrj2fBVCVOL8/[email protected]>
| Newsgroups | gmane.comp.python.spyce.general |
|---|---|
| Message-ID | <[email protected]> |
I think I agree with you. I'll just implement this special case without handlers. ____________________________________ Conan C. Albrecht, Ph.D. Information Systems Department Brigham Young University Web: http://warp.byu.edu/ Email: conan-RXmsyT/Szrj2fBVCVOL8/[email protected] Skype: doconix (please IM in active working groups; send email otherwise) On Dec 22, 2005, at Thu, Dec 22, 2005 10:30 PM, Jonathan Ellis wrote: > On Thu, 22 Dec 2005 22:18:01 -0700, "Conan C. Albrecht" > <conan-RXmsyT/Szrj2fBVCVOL8/[email protected]> said: >> So, the question is now, is there a way to design these handlers so >> they *can* be called from a direct link? > > Anything may be possible, but IMO this is one of the more difficult > anythings. > > The MD5ing, for instance, is a way of generating unique tag ids > with two > important properties: > > - it's low-overhead (happens only at compile time; compiled spyces are > then cached until the source changes) > - it doesn't depend on things being processed in any particular > order -- > important for it to work in any non-single-process Spyce deployment > > (We need unique IDs in the first place because you could conceivably > have multiple instances of the same tags, with different handlers. > The > "chatbox" demo illustrates this in a somewhat contrived fashion.) > > -Jonathan > -- > C++ is history repeated as tragedy. Java is history repeated as > farce. --Scott McKay >
smime.p7s
(application/pkcs7-signature, 2.3 KB) - not displayed