The Wiert Corner – irregular stream of stuff

Jeroen W. Pluimers on .NET, C#, Delphi, databases, and personal interests

  • My badges

  • Twitter Updates

  • My Flickr Stream

  • Pages

  • All categories

  • Enter your email address to subscribe to this blog and receive notifications of new posts by email.

    Join 2,004 other followers

Archive for the ‘wget’ Category

curl/wget: use content-disposition for the file names

Posted by jpluimers on 2020/04/27

For me, on Windows, curl works better than wget, but on Linux/Mac OS X, curl tends to work better. Some people find wget easier for downloading multiple URLs at the same time. So here the parameter switches for both so they download to the file specified by the Content-Disposition http header:

  • curl --remote-name --remote-header-name
  • wget --content-disposition

My experience is that wget is better at this, especially when redirects are involved (by adding a [WayBack] --location parameter to  thecurl command line).

So for instance the first fails, but the second succeeds determining the download to be VSCodeUserSetup-x64-1.27.2.exe (so curl keeps the name stable):

curl.exe --location --remote-name --remote-time --remote-header-name

wget.exe --content-disposition

This takes into account the name after all followed redirects.



Posted in *nix, cURL, Power User, wget | Leave a Comment »

Redirect Detective – A Free Tool To Trace Where Redirects End Up

Posted by jpluimers on 2020/02/27

Cool tool, but be aware it does aJavaScript:xmlhttpPost from , so your data can be logged [WayBackRedirect Detective – A Free Tool To Trace Where Redirects End Up.

For, it shows these redirects (where in bold I’ve listed the changes):

Oh and it runs this call: JavaScript:xmlhttpPost("/linkdetect.px")? from [

Note that this shows more redirects than the plain http ones, so wget from [WayBack] wget – How do I display all URLs in a redirect chain? – Unix & Linux Stack Exchange shows this:

$ wget 2>&1 | grep Location:
Location: [following]
Location: [following]


Posted in *nix, *nix-tools, Communications Development, Development, HTTP, Internet protocol suite, Power User, TCP, wget | Leave a Comment »

wget and curl: downloads that sometimes fail

Posted by jpluimers on 2018/10/19

For my archive somewhere between cURL 7.21.0 and 7.34.0 it does not like to be started from an RDP based tsclient share:

C:\Users\jeroen\Downloads>\\tsclient\bin\curl.7.21.0.exe --remote-name
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100 86465    0 86465    0     0  60805      0 --:--:--  0:00:01 --:--:-- 70012

C:\Users\jeroen\Downloads>\\tsclient\bin\curl.7.34.0.exe --remote-name
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
  0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0curl: (6) Could not resolve host:

C:\Users\jeroen\Downloads>\\tsclient\bin\curl.7.61.0.exe --remote-name
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
  0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0curl: (6) Could not resolve host:

C:\Users\jeroen\Downloads>copy \\tsclient\bin\curl.7.61.0.exe
        1 file(s) copied.

C:\Users\jeroen\Downloads>curl.7.61.0.exe --remote-name
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100    13    0    13    0     0     10      0 --:--:--  0:00:01 --:--:--    10

It fails the same way after net use B: \\tsclient\bin, so that does not matter.

The best link I could find until I got to the real problem was [WayBack] curl: (6) Could not resolve host: application – Stack Overflow which shows a different problem: properly quoting.

In addition to remote-name, you can also grab the file name from the headers using --remote-header-name, and --remote-time use the remote file time. The --location follows 302-redirects. You can see that in the example below which I build based on

[WayBack] unix – Curl to grab remote filename after following location – Stack Overflow: The remote side sends the filename using the Content-Disposition header.curl 7.21.2 or newer does this automatically if you specify –remote-header-name / -J.curl -O -J -L $url

C:\Users\jeroen\Downloads>b:\curl.7.21.0.exe --location --remote-name --remote-time --remote-header-name ""
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
100 86465    0 86465    0     0  45748      0 --:--:--  0:00:01 --:--:-- 50772
curl: Saved to filename ''

wget failed big time:

C:\Users\jeroen\Downloads>B:\wget.exe --no-check-certificate -v -v -v --content-disposition --restrict-file-names=windows ""
wget: Cannot read b:/.wgetrc (No such file or directory).
--2018-07-12 09:55:23--
Connecting to||:443... failed: Invalid argument.


--2018-07-12 09:55:23--  (try:20)
Connecting to||:443... failed: Invalid argument.
Giving up.

This is not caused by the filename (Windows does not like the ? question mark in output file names, so  – like & ampersand in file URLs – you have to quote the full URL, but also provide the --restrict-file-names=windows parameter; see [WayBack] wget – I can’t download files with “?” – Super User).


Posted in *nix, *nix-tools, cURL, Power User, wget | Leave a Comment » TLS certificate does not match domain name

Posted by jpluimers on 2018/09/07

One of the domains not yet monitored at, was the altd download server for ISOs and installers on http and https level. Ultimately you want https, as most of these are about installers, so you do not want any man-in-the-middle to fiddle with them.

TLS on altd fails

Upitmerobot is not yet smart enough to check validity of TLS certificates on https connections.

Chrome, Firefox, Safari, Internet Explorer, wget, curl and ssllabs however are.

altd hides as much from itself as possible

Uptimerobot did not like monitoring the plain and URLs, because the altd is not browsable, so it tries to hide most of its structure from access. This means they both return an odd response:

Those responses are actually 404 errors (note the - minus sign after curl --trace-ascii: it sends the trace to stdout):

$ wget
--2018-09-05 10:44:23--
Resolving (,
Connecting to (||:80... connected.
HTTP request sent, awaiting response... 404 Not Found
2018-09-05 10:44:23 ERROR 404: Not Found.

$ curl --verbose
*   Trying
* Connected to ( port 80 (#0)
> GET / HTTP/1.1
> Host:
> User-Agent: curl/7.54.0
> Accept: */*
< HTTP/1.1 404 Not Found
< Server: Apache
< Content-Type: text/html; charset=iso-8859-1
< Content-Length: 16
< Date: Wed, 05 Sep 2018 08:45:57 GMT
< Connection: keep-alive
* Connection #0 to host left intact
File not found."

$ curl --trace-ascii -
== Info:   Trying
== Info: TCP_NODELAY set
== Info: Connected to ( port 80 (#0)
=> Send header, 84 bytes (0x54)
0000: GET / HTTP/1.1
0010: Host:
002c: User-Agent: curl/7.54.0
0045: Accept: */*
<= Recv header, 24 bytes (0x18)
0000: HTTP/1.1 404 Not Found
<= Recv header, 16 bytes (0x10)
0000: Server: Apache
<= Recv header, 45 bytes (0x2d)
0000: Content-Type: text/html; charset=iso-8859-1
<= Recv header, 20 bytes (0x14)
0000: Content-Length: 16
<= Recv header, 37 bytes (0x25)
0000: Date: Wed, 05 Sep 2018 08:47:19 GMT
<= Recv header, 24 bytes (0x18)
0000: Connection: keep-alive
<= Recv header, 2 bytes (0x2)
<= Recv data, 16 bytes (0x10)
0000: File not found."
File not found."== Info: Connection #0 to host left intact

This is also the reason that WayBack does not want to archive that link, but it can be archived at [Archive.is

Luckily, a Google search for revealed there is a non-installer file short enough (~72 kibibytes) for Uptime robot to check, so it now verifies it can access these:


Read the rest of this entry »

Posted in *nix, *nix-tools, cURL, Encryption, HTTPS/TLS security, Monitoring, Power User, Security, Uptimerobot, wget | Leave a Comment »

Use TLS 1.2 or higher, as TLS 1.1 is phased out on many sites, after TLS 1.0/SLL has been disabled by most for a while now

Posted by jpluimers on 2018/07/23

If you get an error like this in one of your tools

OpenSSL: error:1407742E:SSL routines:SSL23_GET_SERVER_HELLO:tlsv1 alert protocol version

it means you are using a tool not yet properly supporting TLS 1.2 or higher.

Or in other words: update your tool set.

The reason is that – after turning off TLS 1.0 a while ago – more and more sites do the same for TLS 1.1.

A prime example of a site that warned on this in a clear way very early on is github:

Others have done this too, for instance:

TLS 1.0 is vulnerable to many attacks, and certain configurations of TLS 1.1 as well (see for instance [WayBack] What are the main vulnerabilities of TLS v1.1? – Information Security Stack Exchange), which means that properly configuring the non-vulnerable TLS 1.1 over times gets more and more complex. An important reason to say goodbye to that as well, as TLS 1.2 (from 2008) is readily available for a long time. The much more recent TLS 1.3 (from 2018) will take a while to proliferate.

I ran in the above error because on one of my systems, an old version of wget was luring around, so I dug up the easiest place to download recent Windows binaries for both win32 (x86) and win64 (x86_64):

[WayBack] GNU Wget for Windows having a table indicating the OpenSSL version for each wget build.


Reference: Transport Layer Security – Wikipedia: History and development

Posted in *nix, https, HTTPS/TLS security, OpenSSL, Power User, Security, wget | Leave a Comment »

%d bloggers like this: