From: Harry Ohlsen <harryo@...>
Date: 2003-11-14T12:22:13+09:00
Subject: Re: can't flush I know why, but no fix

Hi tO,

T. Onoma wrote:

> I figured out why it won't work. i'm using 1.8's open_uri to retrieve the url with open(url). as soon as this executes everything stops until it downloads the whole file! it isn't a real IO. blah! that stinks. not i wonder if net/http and net/ftp do the same since that's what oprn_uri is based on. if so then means i might have to get down to the socket level just to do a progess reported download. frig'n wonderful!
> 
> If anyone knows of a better way, or a code snippet i can steal, Please let me know.

Sorry, I don't know of any code snippets you could use as a starting point.

However, it might be worthwhile modifying open_uri to allow the asynchronous download and progress notification you want, and then submit a patch, so this can end up in the core code.

Cheers,

Harry O.