hi guys,

thnx for ur input.. if i suppress this using @sign, @curl_exec($ch).. this
will not show that error, my concern was, looking at my sample code -- the
only complete download im getting is the first one (zip file)... the 2nd
(zip) and 3rd (exe) were not downloaded completely coz i can't open it. But
if i use cURL in downloading one link -- it gives me no problem. But for
these, i have 3 site links in my text file.. U guys r still great, tnx for
ur help.

****************************************************************************
**
This email (which includes any files transmitted with it) is intended for
the above named only. It may contain
privileged, confidential or price sensitive information. If you are not the
intended recipient, please notify the
sender immediately and destroy all copies of this email, as well as deleting
the email from your computer system.
This email is protected by copyright. If you are not the intended recipient,
you are not entitled to use,
disclose or copy the email or rely on it in any way whatsoever.
****************************************************************************
**
----- Original Message -----
From: "Wayne Dawson" <Wayne_Dawson_at_inventuresolutions.com>
To: "curl tool talk" <curl-users_at_cool.haxx.se>
Sent: Thursday, June 02, 2005 3:45 AM
Subject: RE: Getting error code 28 on one non-existant file but 22 on
another

> Hi Daniel, my replies to your questions are in [ ], below.
>
> On Wed, 1 Jun 2005, Wayne Dawson wrote:
>
> > I know what it means, but it doesn't make sense for that... Therer's
> > no reason for the timeout. There's no network condidtion for it to
> happen.
>
> And you are absolutely sure about that?
>
> [I'd swear to it. ]
>
> > The file is smaall. The thing is that it doesn't exist. A file with
> > that name *could* exist, but yesterday there ws no such file. I
> > should have got error 22.
>
> Can you show us a command line against a public URL that repeats this
> problem?
>
> [Well, I'd like to... but I can't without giving you our password to the
> site. This site is an ASP's SSL site, and authentication is required.
> The files downloaded have client's personal information in them, and
> well, you know they wouldn't appreciate it. ;)
>
> In any event, it's never happened to the daily files, only the monthend
> ones, and month end was yesterday. Today there would be only daily
> files. If the problem is related to the long filename it would need at
> least one existing monthend file.]
>
> Anyway, it would help if you'd use --trace-ascii in your problematic
> case as then we could analyze that log more when this error 28 occurs.
> And if you'd upgrade to 7.14.0 and --trace-time it would be even better.
>
> [Since, it won't happen now until next month -- at least I haven't seen
> it on the daily files -- this will have to wait until then. I can try
> it. Using 7.14.0 is probably a good idea, regardless, too. Actually,
> that's what I have on my local workstation (your build, as a mater of
> fact). However, I updated only yesterday, and didn't do it on the
> production machines.]
>
> > I did get error 22 when I tried to get the other style of filename,
> > that also didn't exist.
> >
> > This may sound weird, and I tried to explain it, but maybe it wasn't
> > too clear.
>
> I think you were clear, but it still sounds odd.
> [OK, then.... Since we get all the files that are there; I'll change the
> error message to be a warning message; if I can't find any reason for
> this behaviour.
>
> Thanks for the ideas - Wayne ]
>
Received on 2005-06-02