C
Charles Prince
We have found out recently that when our original code was written by the
original team of developers, they incorporated a sstream header that was
crafted by GNU for a situation where the system library did not provide
the beast.
Now we have found this issue, we have used the sstream header supplied by
the system.
Firstly we hit a problem where the original coders were coding around a
bug in the ctor.
Having solved that problem we then went on to analyse the differences
between the original sstream and the current FreeBSD supplied one. We are
now concerned over the inefficiency of the overflow method which
repeatedly deletes and reallocated the buffer each time the buffer is
extended by one character.
I have seen various postings on the net about this problem but no
particular fix by GNU though I accept that this may just be bad google'ing
on my part
So my question on this subject is, what is the thinking by GNU on this
issue. Does the current sstream use a buffer increment larger than one
byte?
TIA,
Pep.
original team of developers, they incorporated a sstream header that was
crafted by GNU for a situation where the system library did not provide
the beast.
Now we have found this issue, we have used the sstream header supplied by
the system.
Firstly we hit a problem where the original coders were coding around a
bug in the ctor.
Having solved that problem we then went on to analyse the differences
between the original sstream and the current FreeBSD supplied one. We are
now concerned over the inefficiency of the overflow method which
repeatedly deletes and reallocated the buffer each time the buffer is
extended by one character.
I have seen various postings on the net about this problem but no
particular fix by GNU though I accept that this may just be bad google'ing
on my part
So my question on this subject is, what is the thinking by GNU on this
issue. Does the current sstream use a buffer increment larger than one
byte?
TIA,
Pep.