[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

My christmas whishlist for OO.o



My christmas wishlist for OOo:

	idlc:
		astexpression.cxx doesn't compile here and some 
		checks like (uint64)foo < 0 generate unneccessary 
		warnings. [1]
                Switches covering all the theoretical
                possible values or using "default:"s would 
                also make it compile on gcc with less warnings.
	store: 
		Please do not give internal and external classes 
		the same name. [2]
	sal:    
		Please do not read the command-line arguments
		from /proc[3]. Or at least do not return NULL in
		getCmdLine or do not expect non-NULL in 
		osl_psz_getExecutableFile, i.e. give some error
		in one of these rotines instead of just segfaulting.


[1] While I also think it good the be paranoid, my buggy gcc is just to
    much confused by checking unsigned values to be less than zero, or
    64-bit values to represent larger numbers than fit in 64 bit. Could
    this somehow be tested while compiling instead of runtime?
[2] Yes, I now, there are .map files to sperate them. But while compiling
    it manually, I did again something wrong and got segfaults on the
    stranges places. Yust renaming the internal OStoreDirectory with
    OStoreDirectoryInternal makes it bulletproof against missed .maps
    and should therefore be a good investment in the future. (Especially
    as it does not affect anything outside libstore)
[3] Yes, no /proc being mounted is not that unnormal, for examples debian
    autobuilders tend to build packages in chroots. 
    I know that this is quite an very large wish. Resolving it the
    right way by saving the arguments in main would need changes in any
    program calling these rotines, including all the helper-programms like
    idlc and co.

Hochachtungsvoll,
	Bernhard R. Link
-- 
The man who trades freedom for security does not deserve 
nor will he ever receive either. (Benjamin Franklin)

Attachment: pgpzni2Odk0G5.pgp
Description: PGP signature


Reply to: