[ANNOUNCEMENT] Updated: VirtualGL-3.0-1

2021-11-20 Thread DRC via Cygwin-announce via Cygwin
The VirtualGL package contains the VirtualGL Client, an X11 application 
that receives, decompresses, and displays OpenGL-rendered frames from a 
remote 3D application that were compressed and sent in real time by the 
VirtualGL Faker via the VGL Transport protocol.  For more information, 
visit https://VirtualGL.org.


Change log is here:
https://github.com/VirtualGL/virtualgl/releases/tag/3.0

Note that Cygwin packages for VirtualGL 2.6.5 and 3.0 beta1 were not 
created, because those releases contained no changes to the VirtualGL 
Client relative to VirtualGL 2.6.4.


--
Problem reports:  https://cygwin.com/problems.html
FAQ:  https://cygwin.com/faq/
Documentation:https://cygwin.com/docs.html
Unsubscribe info: https://cygwin.com/ml/#unsubscribe-simple


Re: ssmtp-config fails on step 6

2021-11-20 Thread Kevin Andrew Lipscomb via Cygwin



On 2021-11-19 18:54, Doug Henderson wrote:

On Fri, 19 Nov 2021 at 11:00, Kevin Andrew Lipscomb via Cygwin 
 wrote:


ssmtp-config fails silently on step 6 at the first call to 
/usr/sbin/alternatives.

I isolated the alternatives command, added the --verbose switch to it, and ran 
the result from a bash shell that was launched as Administrator.  The command 
fails immediately after emitting:

reading /var/lib/alternatives/mta

I am also reporting this as a bug in alternatives at bugzilla.redhat.com.
--
-- Kevin


Please reply quoting the complete failing command line and output from
"uname -a".


From a shell launched (from Windows) as Administrator (which is a user that has 
fully elevated rights)...

Administrator@WIN-O968FBUMJDS ~
$ /usr/sbin/alternatives --verbose --install /usr/sbin/sendmail mta 
/usr/sbin/ssmtp.exe 0 \
--slave /usr/lib/sendmail mta-sendmail /usr/sbin/ssmtp.exe \
--slave /usr/bin/mailq mta-mailq /usr/sbin/ssmtp.exe \
--slave /usr/bin/newaliases mta-newaliases /usr/sbin/ssmtp.exe
reading /var/lib/alternatives/mta

Administrator@WIN-O968FBUMJDS ~
$ uname -a
CYGWIN_NT-6.1 WIN-O968FBUMJDS 3.3.2(0.341/5/3) 2021-11-08 16:55 x86_64 Cygwin


Also please post the bug number or a link to the bug on bugzilla.redhat.com.


I haven't managed to accomplish that yet.  Their system requires me to select a 
"classification" or "product" that the bug applies to, and it's not clear to me 
what I should select.


On 2021-11-19 19:07, Andrey Repin wrote:
:

alternatives is known to make links to nonexistent objects. While this is
possible on *NIX, you can only link to existing objects on Windows.


In my case, /var/lib/alternatives/mta does exist, and its contents can be seen 
here:

Administrator@WIN-O968FBUMJDS /var/lib/alternatives
$ cat /var/lib/alternatives/mta
manual
/usr/sbin/sendmail
mta-mailq
/usr/bin/mailq
mta-newaliases
/usr/bin/newaliases
mta-sendmail
/usr/lib/sendmail

/usr/sbin/ssmtp.exe
0
/usr/sbin/ssmtp.exe
/usr/sbin/ssmtp.exe
/usr/sbin/ssmtp.exe

Administrator@WIN-O968FBUMJDS /var/lib/alternatives
$


On 2021-11-19 20:10, Doug Henderson wrote:
:

and make sure the ownership and permissions are identical to the other
files, e.g. in my case:

$ ls -l /var/lib/alternatives
total 8.0K
-rw-r--r--+ 1 Admin None  41 Jul 16 11:25 2to3
-rw-r--r--+ 1 Admin None  52 Jul 16 11:25 gnuplot
-rw-r--r--+ 1 Admin None 218 Jun 16 17:48 lua
-rw-r--r--+ 1 Admin None  77 Jun  5 08:46 pip3
-rw-r--r--+ 1 Admin None  85 Jun  5 08:46 pydoc3
-rw-r--r--+ 1 Admin None 110 Jun 16 07:34 python
-rw-r--r--+ 1 Admin None  89 Jun  5 08:46 python3
-rw-r--r--+ 1 Admin None  59 Jul 16 11:25 python3-config



Here's what I have:

$ ls -Fals /var/lib/alternatives
total 1
0 drwxr-xr-x+ 1 Administrator None   0 Nov 20 11:46 ./
0 drwxr-xr-x+ 1 Administrator None   0 Feb  5  2018 ../
0 -rw-r--r--  1 Administrator None   0 Apr  4  2013 .keep-alternatives
1 -rw-r--r--  1 Administrator None 200 Nov 20 11:46 mta


Thank you both for your help with this.

-- Kevin

--
Problem reports:  https://cygwin.com/problems.html
FAQ:  https://cygwin.com/faq/
Documentation:https://cygwin.com/docs.html
Unsubscribe info: https://cygwin.com/ml/#unsubscribe-simple


Re: [ANNOUNCEMENT] New: unison2.48+4.04.2, unison2.48+4.08.1 [test]

2021-11-20 Thread Ken Brown via Cygwin

On 9/8/2020 4:57 PM, Andrew Schulman via Cygwin-announce wrote:
[...]

You can install any number of these packages side-by-side. Separate
packages are needed because in order to synchronize your files, you have to
run compatible versions of Unison on the client and server. Two Unison
executables are compatible if and only if:

(1) They have the same first two numbers of the Unison version. For
example, all Unison versions 2.48.* are compatible with each other. But if
you try to use version 2.51.x to sync with a server running version 2.48.y,
Unison will issue an error message about incompatible versions and quit.

AND

(2) They were built with compatible versions of the OCaml compiler.


This old problem reared its ugly head again for me, but I found a simple 
solution that I'm passing on in case it's of use to others.


My situation is that for years I have been syncing my Cygwin system with a Linux 
system on which I had built unison 2.48.x with OCaml 4.04.x.  This is compatible 
with Cygwin's unison2.48+4.04.2.  But I just added a second Linux system that I 
want to keep in sync with my other systems, and this one comes with a Unison 
compatible with Cygwin's unison2.48+4.08.1.


It turns out that Unison's upstream maintainer is making Linux binaries 
available, built with various different versions of OCaml.  See, for examples, 
the Assets listed under v2.51.4 at


  https://github.com/bcpierce00/unison/releases

So if I install unison2.51 on Cygwin and install the appropriate binary in ~/bin 
on both of the Linux machines that I want to sync with, then everything works. 
For example, if I see


  $ unison -version
  unison version 2.51.4 (ocaml 4.12.0)

on Cygwin, then I know that I need to use 
unison-v2.51.4+ocaml-4.12.0+x86_64.linux.tar.gz on the Linux machines.


It's still annoying, but not as bad as it used to be.

Ken

--
Problem reports:  https://cygwin.com/problems.html
FAQ:  https://cygwin.com/faq/
Documentation:https://cygwin.com/docs.html
Unsubscribe info: https://cygwin.com/ml/#unsubscribe-simple


Re: ssmtp-config fails on step 6

2021-11-20 Thread Ken Brown via Cygwin

On 11/19/2021 7:07 PM, Andrey Repin via Cygwin wrote:

alternatives is known to make links to nonexistent objects. While this is
possible on *NIX, you can only link to existing objects on Windows.


You're talking about native Windows symlinks, not Cygwin symlinks.  The latter 
can point to nonexistent objects just as on *NIX.  This shouldn't fail unless 
the user has forced Cygwin to use native Windows symlinks by setting 
winsymlinks:nativestrict in the CYGWIN environment variable.


Kevin, have you done this?  If not, do you have any reason to think that 
alternatives failed to create symlinks?  What does 'ls -l 
/etc/alternatives/mta*' show?  On my system I see:


$ ls -l /etc/alternatives/mta*
lrwxrwxrwx 1 kbrown-admin None 19 2021-11-20 17:04 /etc/alternatives/mta -> 
/usr/sbin/ssmtp.exe*
lrwxrwxrwx 1 kbrown-admin None 19 2021-11-20 17:04 /etc/alternatives/mta-mailq 
-> /usr/sbin/ssmtp.exe*
lrwxrwxrwx 1 kbrown-admin None 19 2021-11-20 17:04 
/etc/alternatives/mta-newaliases -> /usr/sbin/ssmtp.exe*
lrwxrwxrwx 1 kbrown-admin None 19 2021-11-20 17:04 
/etc/alternatives/mta-sendmail -> /usr/sbin/ssmtp.exe*


I also have:

$ ls -l /usr/sbin/sendmail
lrwxrwxrwx 1 kbrown-admin None 21 2021-11-20 17:04 /usr/sbin/sendmail -> 
/etc/alternatives/mta*


and

$ /usr/sbin/alternatives.exe --display mta
mta - status is manual.
 link currently points to /usr/sbin/ssmtp.exe
/usr/sbin/ssmtp.exe - priority 0
 slave mta-sendmail: /usr/sbin/ssmtp.exe
 slave mta-mailq: /usr/sbin/ssmtp.exe
 slave mta-newaliases: /usr/sbin/ssmtp.exe
Current `best' version is /usr/sbin/ssmtp.exe.

Do you see something different on your system?

Last question: You said in your first post that "ssmtp-config fails silently" 
and that you isolated this to the alternatives command failing.  What do you 
mean when you say it failed?  Since it failed "silently", do you mean that 'echo 
$?' showed a nonzero error code?  Or do you mean that something is not working 
as expected?


Ken

--
Problem reports:  https://cygwin.com/problems.html
FAQ:  https://cygwin.com/faq/
Documentation:https://cygwin.com/docs.html
Unsubscribe info: https://cygwin.com/ml/#unsubscribe-simple


[ANNOUNCEMENT] Updated: cpuid 20211114

2021-11-20 Thread Brian Inglis
The following package has been upgraded in the Cygwin distribution:

* cpuid 2024

The program displays detailed information about the CPU(s) gathered from
the CPUID instruction, and also determines the exact model of CPU(s).

Whereas /proc/cpuinfo is like an abstract of the features important to
Linux in a system, cpuid is a standalone utility which writes a paper
expounding on every feature in each CPU's architecture and what it can
do, at about the one line per bit level.

It is updated and released frequently and appears current with Intel and
AMD information and supports other vendors' chips.

See the project home page for more information:

http://etallen.com/cpuid.html

For information about changes since the previous Cygwin release,
see below or /usr/share/doc/cpuid/ChangeLog after installation.


Changes

* Updated (uarch synth) for Bergamo to say "Zen 4c".
* Added (synth) decoding for Pentium Gold G5420 (0,6),(9,14),10, based
  on a sample from Andrey Rahmatullin. Previously, I knew only of
  variants with stepping 11.
* Renamed Golden Cove & derivatives physical properties to use new Intel
  7 branding. Likewise for the preliminary Redwood Cove & Granite Rapids
  and Intel 4.
* Added (synth) decoding for AMD Storm Peak, Lucienne A1 stepping, Van
  Gogh, Mero, Mendocino, Chagall, Badami, Rembrandt, Genoa, Raphael,
  Phoenix, Bergamo.
* Added (synth) decoding for Intel (0,6),(10,8), (0,6),(10,10),
  (0,6),(10,12), (0,6),(10,13), (0,6),(9,15).
* Added Cougar Mountain name to Intel Puma 7.
* Added Microsoft hypervisor+3/edx, hypervisor+7/eax, hypervisor+7/ecx,
  hypervisor+0xa/eax new fields.
* Updated (synth) decoding for Alder Lake-S to include Core variants.
* Added (synth) and (uarch) decoding for VIA/Centaur CHA (CNS).
* Added 7/1/eax AMX tile support, from Linux kernel.


-- 
Problem reports:  https://cygwin.com/problems.html
FAQ:  https://cygwin.com/faq/
Documentation:https://cygwin.com/docs.html
Unsubscribe info: https://cygwin.com/ml/#unsubscribe-simple


[ANNOUNCEMENT] Updated: curl, libcurl{4, -devel, -doc}, mingw64-{x86_64, i686}-curl 7.80

2021-11-20 Thread Cygwin curl Maintainer
The following packages have been upgraded in the Cygwin distribution:

* curl  7.80
* libcurl4  7.80
* libcurl-devel 7.80
* libcurl-doc   7.80
* mingw64-x86_64-curl   7.80
* mingw64-i686-curl 7.80

Command line tool and Library supporting transferring files with
URL syntax, using FTP, FTPS, HTTP, HTTPS, SCP, SFTP, TFTP, TELNET, DICT, and
FILE, SSL certificates, HTTP POST, HTTP PUT, FTP uploading, HTTP form
based upload, proxies, cookies, user+password authentication (Basic,
Digest, NTLM, Negotiate...), file transfer resume, proxy tunneling and a
busload of other useful tricks.

For more information see the project home page:

https://curl.se/

As there are multiple components and many changes each release please
see below or read /usr/share/doc/curl/RELEASE-NOTES after installation;
for complete details of changes since the previous Cygwin release see:

/usr/share/doc/curl/RELEASE-NOTES
or
https://curl.se/changes.html


curl and libcurl 7.80.0

 Public curl releases: 204
 Command line options: 243
 curl_easy_setopt() options:   294
 Public functions in libcurl:  86
 Contributors: 2533

This release includes the following changes:

 o CURLOPT_MAXLIFETIME_CONN: maximum allowed lifetime for conn reuse
 o CURLOPT_PREREQFUNCTION: add new callback
 o libssh2: add SHA256 fingerprint support
 o urlapi: add curl_url_strerror()
 o urlapi: support UNC paths in file: URLs on Windows
 o wolfssl: allow setting of groups/curves

This release includes the following known bugs:

* see /usr/share/doc/curl/KNOWN_BUGS (https://curl.se/docs/knownbugs.html)

This release includes the following bugfixes:

 o .github: retry macos "brew install" command on failure
 o aws-sigv4: make signature work when post data is binary
 o BINDINGS: URL updates
 o build: remove checks for WinSock 1
 o c-hyper: don't abort CONNECT responses early when auth-in-progress
 o c-hyper: make Curl_http propagate errors better
 o c-hyper: make CURLOPT_SUPPRESS_CONNECT_HEADERS work
 o c-hyper: make test 217 run
 o c-hyper: use hyper_request_set_uri_parts to make h2 better
 o checksrc: ignore preprocessor lines
 o CI/makefiles: introduce dedicated test target
 o ci: update Lift config to match requirements of curl build
 o cirrus: remove FreeBSD 11.4 from the matrix
 o cirrus: switch to openldap24-client
 o cleanup: constify unmodified static structs
 o cmake: add CURL_ENABLE_SSL option
 o cmake: fix error getting LOCATION property on non-imported target
 o CMake: restore support for SecureTransport on iOS
 o cmake: with OpenSSL, define OPENSSL_SUPPRESS_DEPRECATED
 o cmdline-opts: made the 'Added:' field mandatory
 o configure.ac: replace krb5-config with pkg-config
 o configure: when hyper is selected, deselect nghttp2
 o connect: use sysaddr_un from sys/un.h or custom-defined for windows
 o curl-confopts.m4:  remove --enable/disable-hidden-symbols
 o curl-openssl.m4: modify library order for openssl linking
 o curl-openssl: pass argument to sed single-quoted
 o curl.1: remove mentions of really old version changes
 o curl: actually append "-" to --range without number only
 o curl: correct grammar in generated libcurl code
 o curl: print help descriptions in an aligned right column
 o curl_gssapi: fix link error on macOS Monterey
 o curl_multi_socket_action.3: add a "RETURN VALUE" section
 o curl_ntlm_core: use OpenSSL only if DES is available
 o Curl_updateconninfo: store addresses for QUIC connections too
 o CURLOPT_ALTSVC_CTRL.3: mention conn reuse is preferred
 o CURLOPT_HSTSWRITEFUNCTION.3: using CURLOPT_HSTS_CTRL is required
 o CURLOPT_HTTPHEADER.3: add descripion for specific headers
 o docs/HTTP3: improve build instructions
 o docs/Makefile.am: repair 'make html'
 o docs: fix typo in CURLOPT_TRAILERFUNCTION example
 o docs: provide "RETURN VALUE" section for more func manpages
 o docs: reduce use of "very"
 o doh: remove experimental code for DoH with GET
 o examples/htmltidy: correct wrong printf() use
 o examples/imap-append: fix end-of-data check
 o ftp: make the MKD retry to retry once per directory
 o gen.pl: insert the current date and version in generated man page
 o gen.pl: replace leading single quotes with \(aq
 o http2: make getsock not wait for write if there's no remote window
 o HTTP3: fix the HTTP/3 Explained book link
 o http: fix Basic auth with empty name field in URL
 o http: reject HTTP response codes < 100
 o http: remove assert that breaks hyper
 o http: set content length earlier
 o http_proxy: make hyper CONNECT() return the correct error code
 o http_proxy: multiple CONNECT with hyper done better
 o hyper: disable test 1294 since hyper doesn't allow such crazy headers
 o hyper: does not support disabling CURLOPT_HTTP_TRANSFER_DECODING
 o hyper: pass the CONNECT line to the debug callback
 o imap: display quota information
 o INSTALL: update symbol hiding option
 o lib/mk-ca-bundle.pl: skip certs passed Not Valid After da

Re: possible snprintf() regression in 3.3.2

2021-11-20 Thread Tony Cook
On Thu, Nov 18, 2021 at 09:08:40PM +, Sam Edge via Cygwin wrote:
> I use newlib on embedded with threading libs that have predetermined
> fixed thread stack sizes. While we tend to have more RAM than in former
> times we also have multiple thread stacks. Use of alloca() or variable
> length automatic arrays makes me wince especially in code I might not be
> able to avoid calling which is often the case with XXXprintf() in
> third-party libraries' debug output. I'd usually rather take the
> performance hit from using heap instead of having to make all my stacks
> bigger.

A simple option would be to use an small auto fixed buffer for most
conversions, but use malloc() for %f formats for numbers greater in
magnitude than some limit, though it would also need to be adjusted
for the precision (ndigits here), since they take extra space.

This would avoid using the optional-to-implement VLA feature too.

Tony

-- 
Problem reports:  https://cygwin.com/problems.html
FAQ:  https://cygwin.com/faq/
Documentation:https://cygwin.com/docs.html
Unsubscribe info: https://cygwin.com/ml/#unsubscribe-simple