[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index][Thread Index][Top&Search][Original]

Re: [ID 19991224.002] Truncate does not flush the output buffer before

Ilya Zakharevich <ilya@math.ohio-state.edu> writes:
>On Wed, Dec 29, 1999 at 10:22:41PM +0000, Nick Ing-Simmons wrote:
>> >> I thought p5p agreed it made sense, and that SIFO did (it to some extent)
>> >> already. We just have not done it yet.
>> >
>> >Glad to hear that there is some progress.  However, I do not
>> >*remember* any progress.  What exactly do you mean?  Are there flags
>> >on a filehandle "buffer may be active", "last action was write/read"?
>> >[I do not understand how something may be done without these flags.]
>> SFIO's equivalent of FILE * does indeed have those flags.
>> So iff PerlIO == SFIO so does perl - but I mean the 'and only if' :-(
>Since we have an IO level *above* SFIO, and the read/write flags
>should be active on sysread/syswrite operations too, so we need to
>able to freely check/set/reset these flags, SvIO (sp?) is where the
>flags should be stored.  (Of course, add "IMO" wherever appropriate.)

I think 'my opinion' aligns with yours in this. But noone has added
those flags yet. 

It is far from clear that the "level *above*" is 
necessary - i.e. we could roll all the SFIO/STDIO features we need 
into PerlIO or SvIO and then only ever call open/read/write/lseek/close
+ (recv,send,...)

Nick Ing-Simmons <nik@tiuk.ti.com>
Via, but not speaking for: Texas Instruments Ltd.

References to:
Ilya Zakharevich <ilya@math.ohio-state.edu>

[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index][Thread Index][Top&Search][Original]