I absolutely agree with the general sentiment here and I apologize if this came off a bit rude. It was not my intention to step on anyone's toes or make someone's community contributions feel any less appreciated!

The simple reality here is that SlackBuilds in -current differ quite a bit from SlackBuilds on SBo, and for simple builds it's just easier to start over fresh than to mold an existing one into -current's format. For any more complicated builds though, I always check SBo and, as far as I know, everyone tries to make sure attributions go out to the right people in that case.

efivar should be bumped using the fetch-efivar script to fix the eMMC issue. It builds now once again with GCC 10.

Andrew.

kernel 5.10.4 will cause Xorg freeze when display is put on standby, https://mailman.alsa-project.org/pip...ry/178794.html
If no new kernel it out yet, would be better to include the patch.

Thanks very much for making clear that "the inner circle" cares about those contributions. That's really important even for simpler builds, as the motivation to spend time(and even some money) on contributions stays with this feedback.

When i added s-nail to SBo some years ago, it was basically Pat's mailx buildscript adapted to s-nail. I wrongly didn't keep the copyright line from Pat(as the buildname changed i wasn't sure), but a SBo maintainer added the copyright-line back.
So i wondered that the other way round that didn't happen. Admittedly, your build is not the same as mine.

I get the point. I offer to streamline SBo builds in case some more SBo builds find their way into -current, just can't guess when this happens.

One more nitpick, the manpage-compress-section in the -current dehydrated.SlackBuild is somewhat ancient.

5.10.5 is expected in 6 of january , you propose a patch for a 1 day ?

I think no success.

Wait to 5.10.5

Please llvm is under 11.0.1-rc2 , consider upgrade the libclc , when 11.0.1 goes released as stable, cause now libclc goes hand by hand with llvm version.

See libclc-11.0.1 is under rc2 like llvm

https://github.com/llvm/llvm-project/

Let Encrypt
 

Never could get certbot to run on slackware, so we used acme.sh, it's so simple and easy to use even the most basic of support staff can use it, its never failed a rego, its never failed a renew so its never caused websites errors due to the 3 month expiry - unlike certain other named proggies

acme.sh is a bash script, a damn powerful one.

I doubt Pat would consider adding it, but ... https://github.com/acmesh-official/acme.sh if he read this

Just a note.. I've run letsencrypt on half a dozen 14.2 and -current servers. The slackbuild works fine (first use sqg to build the queue)

Possible broken sym links:

* /usr/doc/cldr-emoji-annotation-37.0_13.0_0_2/COPYING
* /usr/doc/artikulate-20.12.0/COPYING
* /usr/doc/artikulate-20.12.0/COPYING.DOC
* /usr/man/man1/bind9-config.1.gz
* /usr/share/bash-completion/completions/runuser
* /usr/bin/texdoctk

Take a look at the GitHub project pages for Issues and Pull requests.
https://github.com/dehydrated-io/dehydrated
https://github.com/acmesh-official/acme.sh

I use acme.sh, which illustrates its beginner's friendliness. To renew a certificate:It always worked for me so far.

for 'articulate' it's probably, because 'COPYING and COPYING.DOC' are symbolic links in the source of articulate.

for 'runuser' bad link, this is a part of 'util-linux' package compiled with 'su' by default, but, su is removed in the SlackBuild, because the 'su' used in slackware is the 'su' of 'shadow' package, compile util-linux with the option --disable-su, solve this, but, it's probably a cosmetic problem.

for the others I did not look ;)

As apache+php-fpm becomes more common(e.g. diffrent php versions may be used in parallel), it would be a benefit to have /etc/rc.d/rc.php-fpm startet after /etc/rc.d/rc.httpd in /etc/rc.d/rc.M, rather than messing in /etc/rc.local.
Therfore a valid default php-fpm config needs to be present, e.g.:Or, just let /etc/php-fpm.d/www.conf.default there at it is, as 644 is the default for /etc/rc.d/rc.php-fpm and fpm-pools would need anyway to be configured by an admin.

Yes acme.sh is 100% reliable - unlike other methods

Although if you disagree with DoH, you should also use option --dnssleep 30 with new and renew
This option will force use your local resolver and not send your requests to a centralized internet via cloudfare and co.

Interesting, thats 24 hours ago and I have not yet got my email advising this (I always build php myself as I use/need specifics)