This issue tracker has been migrated to GitHub, and is currently read-only.
For more information, see the GitHub FAQs in the Python's Developer Guide.

Author vinay.sajip
Recipients Arfrever, amaury.forgeotdarc, benjamin.peterson, georg.brandl, vinay.sajip
Date 2011-06-19.23:04:53
SpamBayes Score 5.3225914e-08
Marked as misclassified No
Message-id <27487.70654.qm@web25805.mail.ukl.yahoo.com>
In-reply-to <1308521417.57.0.56329397885.issue12291@psf.upfronthosting.co.za>
Content
> Benjamin Peterson <benjamin@python.org> added the  comment:
> 
> I think you're right about playing with the bare fd being too  fragile. I think 
>a simpler solution is to read say 1024 bytes at a time and  buffer it  
>internally.

Doesn't this suffer from a similar problem? Namely, external Python code 
expecting the stream pointer to always be just past the object just read. See my 
earlier comments pointing out that there's nowhere to store the buffer state 
between successive calls to marshal.load. And the synchronising can be a problem 
to achieve with non-seekable streams (including, but not limited to, sockets).

I'm not trying to be difficult. No, really! :-)
History
Date User Action Args
2011-06-19 23:04:53vinay.sajipsetrecipients: + vinay.sajip, georg.brandl, amaury.forgeotdarc, benjamin.peterson, Arfrever
2011-06-19 23:04:53vinay.sajiplinkissue12291 messages
2011-06-19 23:04:53vinay.sajipcreate