"No error" condition

[Andre Grueneberg]

Hi Nick

Nick Andre schrieb:

NA> Is there any workaround that can be done in BinkD?

Best we could do is to treat this M_GET as unexpected (as if there is no such
file). But I have no clue how IRex would react to it.

Alternatively one could treat offset > size like offset == 0, but here again
it's unclear how implementations would react ... this could end in an endless
loop of M_FILE(offset 0)+M_GET(offset invalid).

CU Andre E-Mail: ***@grueneberg.de

[mark lewis]

NA> Is there any workaround that can be done in BinkD?

AG> Best we could do is to treat this M_GET as unexpected (as if there
AG> is no such file). But I have no clue how IRex would react to it.

AG> Alternatively one could treat offset > size like offset == 0, but
AG> here again it's unclear how implementations would react ... this
AG> could end in an endless loop of M_FILE(offset 0)+M_GET(offset
AG> invalid).

fully agreed but i recall some tests in the past that laid out the supported
features of other binkd compatible software... some of that software was seen,
at that time, as unsupported, but there was still feature enhancements added
that the other software did not support... then there was also the problem that
some previous software /could not/ be supported because of various reasons...
the main one being that the source code was lost...

so now what? how to continue to support those other packages that cannot
support newer binkd capabilities and bug fixes???

)\/(ark

[Andre Grueneberg]

Hi mark

mark lewis schrieb:

ml> so now what? how to continue to support those other packages that
ml> cannot support newer binkd capabilities and bug fixes???

New binkd capabilities shall be specified and implemented in a backward
compatible way. Bug fixes and known issues should be documented in the FTSC
protocol documentation in order to implement a backward compatible correction.
[I've got one good example: older binkD and qico do not conform with CRAM
specification -> FTSC documentation shall reflect reality]

CU Andre E-Mail: ***@grueneberg.de