[PD-dev] loaderHEX bug report

IOhannes m zmoelnig zmoelnig at iem.at
Wed Nov 23 13:22:51 CET 2005


Thomas Grill wrote:
>>
> using my loader patch the naming of the loader function would be 
> irrelevant since it's just including in a chain of loader functions, 
> starting with the classic (now called sys_loader_lib_internal) PD loader.

so this basically means, we could write an external that provides the 
character escaping loader (and would not have to patch pd-core)

> 
> Concerning the setup function naming - how about xyz_altsetup ?
> it's true that it should be distinguished from the old naming, but 
> should not have a prefix that excludes some object names.

again the problem with this is, that the current escaping does (for 
"readability" replace "invalid" characters by "0x??" where "??" is the 
hex-value of the character. starting a function-name with numbers is 
generally a bad idea, if you want a compiler to read it.
so that's why we thought of a prefix that guarantees a valid indentifier.
prefixing "setup_" (which i still think is nice ,-)) would solve both 
the problem with valid id's and establish a distinct naming scheme 
(which, as hans pointed out, could also be seen as a problem)

mfg.asd.r
IOhannes




More information about the Pd-dev mailing list