Firmware Selector - Packages with "-" in front?

I just went to download a snapshot for the GL.iNet MT2500 from the Firmware Selector, and when I expand the "Customize installed packages and/or first boot script" section, I see two packages that are preceded by a dash ("-"), then repeated later in the list (without dash):

-kmod-mt7915e -wpad-basic-mbedtls base-files busybox ca-bundle dnsmasq dropbear e2fsprogs f2fsck firewall4 fstools kmod-crypto-hw-safexcel kmod-gpio-button-hotplug kmod-leds-gpio kmod-mt7915e kmod-nft-offload kmod-phy-aquantia kmod-usb3 libc libgcc libustream-mbedtls logd luci mkf2fs mtd netifd nftables odhcp6c odhcpd-ipv6only opkg ppp ppp-mod-pppoe procd procd-seccomp procd-ujail uboot-envtools uci uclient-fetch urandom-seed urngd wpad-basic-mbedtls

What does this mean? Is there any reason for concern?

It means 'minus' -- basically it looks at the default manifest and then excludes the packages with the "-".

No, this is very likely done for a reason, and you'll notice that the following is removed:

But then

this is added.

In this case, it looks like what was removed is the same as what was added back, so not exactly sure why the build recipe is structured that way, but there should be no issue.


OK. That's along the lines of what I was thinking (I saw the add-back on both packages).

Is there anything to do with dependencies or ensuring a most-recent package that could explain removing them initially and then re-adding?

Honestly, I'm not positive. But I would guess no because it should pull the packages based on their names.

There are some games you have to play when you replace default packages like dnsmasq with dnsmasq-full, so the various tools all do it a bit differently.

auc and LuCI Attended Sysupgrade both just use ubus call rpc-sys packagelist and feed that result into the ASU server. The Firmware Selector lets you type in a list of packages, so you can directly modify it with the -pkg items that are sent to the ASU server. Once the ASU server gets it, then the device-specific image builder also does some stuff that I haven't dug into yet, but I think it applies its transformations with a consistent set of rules across all the targets.


This topic was automatically closed 10 days after the last reply. New replies are no longer allowed.