I forgot to describe the 'work-around'.
When the 'side effect' kicks in.....
While still viewing the partially downloaded page.
Hit DEL to remove it from the cache.
Go back in history to the page containing the link.
Hold the Ctrl key this time while taking the link.
(background download)
Stay on this page until the download has completed.
(leaving this page will terminate the D/L)
After the D/L is complete... go forward in history to the complete page.
I just used this metod to get aratest.cgi
Is this the corrct filesize ?
(the BIG one) ;-)
Volume in drive L is MS-RAMDRIVE
Directory of L:\ARACHNE\CACHE
. <DIR> 6-22-03 8:04p
.. <DIR> 6-22-03 8:04p
HEADERS <DIR> 6-22-03 8:04p
56329968 HTM 6220 6-22-03 8:58p
$IDX$CNM IDX 0 6-22-03 9:04p
$SRT$CNM IDX 0 6-22-03 9:04p
56329980 HTM 3001 6-22-03 8:59p
56330279 HTM 6661 6-22-03 9:04p
56330322 HTM 415763 6-22-03 9:05p
56330474 HTM 3059 6-22-03 9:07p
56330575 HTM 6875 6-22-03 9:08p
11 File(s) 59305984 bytes free
And is this the correct last line of the file ? (split for emailing)
91.72.255.62 => 1047148488,Mozilla/4.0
(compatible; MSIE 6.0; Windows NT 5.0; ntlworld v2.0)
On Sun, 22 Jun 2003 20:35:54 -0400, Glenn McCorkle wrote:
> On Mon, 23 Jun 2003 01:24:38 +0000, [EMAIL PROTECTED] (Udo Kuhnt)
> wrote:
>> On Fri, 06 Jun 2003 21:52:11 -0400, Glenn McCorkle wrote:
>>> On Sat, 07 Jun 2003 00:40:53 +0000, [EMAIL PROTECTED] (Udo Kuhnt)
>>> wrote:
>>>> I just wanted to know if the address [EMAIL PROTECTED] is still valid for
>>>> sending Arachne-related bug reports. Having sent several mails to
>>>> arachne.cz in the last few years without getting an answer, I sometimes
>>>> wonder if the mail server just forwards them all to NUL. Well, at least
>>>> majordomus is still responding. ;-)
>>> Yes,
>>> That address is still valid.
>>> However,
>>> Michael has not had the time read very many of those
>>> messages in over 2 years.
>>> Therefore,
>>> This is a better address to use for reporting any bugs.
>>> [EMAIL PROTECTED]
>> Well, I had hoped that you would say this. So I take my chances and send
>> you a report of what I think is a bug in Arachne. :-)
>> Recently, I ran a CGI script on my web site to dump the contents of the
>> connections database and was surprised that Arachne did not show the
>> full list. I thought it was a bug in the script, but a try with Lynx did
>> result in the whole html document being downloaded, so the script was
>> sending the full content. Arachne seems to stop the download shortly
>> after 26000 bytes have been received from the server - without an
>> apparent reason. The version of Arachne I am currently using is
>> 1.71UE01.
>> If you want to have a look at this yourself, I have set up a test CGI
>> script at http://www.drdosprojects.de/cgi-bin/aratest.cgi .
> What you experienced is a 'side effect' of my "Quick-N-Dirty fix" of the
> 'Runaway download counter bug'.
> On very large http pages where thre server is not telling us the
> file size. The RDLC bug fix will terminate the download when
> 24000 unsuccessful attempts have been made to get new data from the
> connection.
> This is the bit of code I used to prevent the RDLC bug from causing
> the unstoppable loop we were having in versions 1.69 and 1.70
> --- in html.c of the src code ---
> //!!glennmcc: begin Feb 24, 2002 (Quick-n-dirty Fix for RDLC bug)
> bflenold=bflen;
> //!!glennmcc: end
> bflen=readHTML(cache,p->html_source);
> if(bflen<=0 || p->memory_overflow) //end of page, or out of memory
> goto exitloop;
> else
> i=0;
> }
> //!!glennmcc: begin Feb 24, 2002 (Quick-n-dirty Fix for RDLC bug)
> if (cache->knowsize)goto knowsize;
> if (bflenold==bflen)
> {
> retry++;
> }
> else
> {
> retry=0;
> }
> if(retry>24000)
> {
> retry=0;
> goto exitloop;
> }
> knowsize:
> //!!glennmcc: end
> _________________________________
--
Glenn
http://arachne.cz/
http://www.delorie.com/listserv/mime/
http://www.cisnet.com/glennmcc/
http://www.cisnet.com/glennmcc/aqc/
