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

Re: [MiNT] Cookies

2010/10/17 m0n0 <ole@monochrom.net>:
>
> Am Sonntag, den 17.10.2010, 13:27 +0200 schrieb Kåre Andersen
> <kareandersen@gmail.com>:
>
>>
>> But it was an error in mintlib's getcookie() function, and nothing to
>> do with old software unable to relocate the cookie jar.
>
> Yes... but the software is using getcookie() to check if the cookie is
> already present, the app adds the cookie if getcookie() ain't returns
> success...
> It could have been possible...

Sure, but that would take a recompilation of the software in question,
as we only do static linking. Unless the kernel itself is using that
function internally and Jupiter and the Moon were aligned in a very
funky way... ;-)

I'm pretty sure this is caused by an old TSR in the AUTO folder not
reallocating the cookie jar when its full. TOS gives this
responsibility to the application, and its potentially dangerous not
to handle it properly. Of course, this particular case has not been
proven yet, so maybe a test app is in order? Does anyone else have
this problem at all?

Anyway, if as Roger suggests Lattice C was unable to do this
correctly, it may be an idea for someone to look into binary-patching
the executables still in use that we don't have source code for. But
since I don't use Lattice I can't really help, and am honestly more
interested in up to date software.

-K