General Linux Chat and Small Questions v. I broke my Arch Install

[rilez]

Forcing Firefox windows together with the WM brings up some issues in practice. For example, what if the user [B]wants [/B]a new, seperate window? Or if they happen upon the old "new tab" function, then there's a confusion between tabs and windows. Especially if someone used the old shortcut to make a new "window", and all it did was make a tab. It felt like a really round about way to save space. So I'm trying something different:

[t]http://i.imgur.com/q2aBzjP.png[/t]

Everything occupies the same bar. I use KWin and HTitle to make exceptions for Firefox to ditch the window border, and to draw window buttons on the bar. Since I'm using the GNOME theme to test with there's no Maximize or Minimize, but those will be there too. You can still drag the window around, though I might toss a bit more padding on my FF theme to make it easier to do so.

I feel like this is the perfect way to merge UI elements without actually removing functionality. I'm not sure how else you could do it without it being confusing or hard to use

[editline]25th March 2014[/editline]

The scrollbar will also be a sort of overlay bar (like Unity) except not really. Transparent trough, but big enough to click. I would also like it to auto-hide if possible, like on Android/iOS. So you can just focus on content without the UI getting in the way.

[nikomo]

I feel the need, the need, for speed.
[img]http://i.imgur.com/D8oZe11.png[/img]

This is pretty great now that I have it set up.

[FPtje]

falco@falco-K53SV:~$ sudo apt-get remove apache2
Reading package lists... Done
Building dependency tree       
Reading state information... Done
The following extra packages will be installed:
  php5-cgi
Suggested packages:
  php-pear
The following packages will be REMOVED:
  apache2 libapache2-mod-php5
The following NEW packages will be installed:
  php5-cgi
0 upgraded, 1 newly installed, 2 to remove and 1 not upgraded.
Need to get 5591 kB of archives.
After this operation, 7789 kB of additional disk space will be used.
Do you want to continue [Y/n]? 

Remove one package get another in its place.

What the fuck am I supposed to do with php5-cgi?

[editline]26th March 2014[/editline]

falco@falco-K53SV:~$ sudo apt-get remove apache2 php5-cgi
Reading package lists... Done
Building dependency tree       
Reading state information... Done
The following packages were automatically installed and are no longer required:
  apache2-bin apache2-data libaprutil1-dbd-sqlite3 libaprutil1-ldap php5-cli php5-fpm php5-readline
Use 'apt-get autoremove' to remove them.
The following extra packages will be installed:
  php5-fpm
Suggested packages:
  php-pear
The following packages will be REMOVED:
  apache2 libapache2-mod-php5 php5 php5-cgi
The following NEW packages will be installed:
  php5-fpm
0 upgraded, 1 newly installed, 4 to remove and 1 not upgraded.
1 not fully installed or removed.
Need to get 2839 kB of archives.
After this operation, 19,3 MB disk space will be freed.
Do you want to continue [Y/n]? 

How long will this continue?

[mastersrp]

Alright, so after I've gotten all of the sed work done that I needed, I stumbled upon a problem which I'm not sure how to fix. This is going to present me as a complete UNIX newbie, but that's the way of things.
Anyway, I have this RSS feed parser that fetches any kind of news feed, and outputs it into what I'll refer to as an intermediate format. It's somewhat simple to read for humans, but I'm not sure how to parse it yet. I realize that this isn't the smartest of ways, but bear with me.
Anyway, two feed items may look like this:

Title: Hello world
Link: http://www.example.com
Description: A very world of hello.
It can have newlines too. Potentially links? Or other content, like HTML tags.
<a href="link">not really a link</a>.

Title: The second item
Link: www.google.com
Description: No parsing is required of the individual item sections. I'll handle that with some simple regex.
However, I'm not sure how to parse this.

Ideally, I'll need it to work with Busybox Ash. One such way would be

for f in $(magic code that iterates over each item); do
  parse( $f );
done

Any takers? I'm kind of at a loss here.

[editline]26th March 2014[/editline]

[QUOTE=FPtje;44359881]

falco@falco-K53SV:~$ sudo apt-get remove apache2
Reading package lists... Done
Building dependency tree       
Reading state information... Done
The following extra packages will be installed:
  php5-cgi
Suggested packages:
  php-pear
The following packages will be REMOVED:
  apache2 libapache2-mod-php5
The following NEW packages will be installed:
  php5-cgi
0 upgraded, 1 newly installed, 2 to remove and 1 not upgraded.
Need to get 5591 kB of archives.
After this operation, 7789 kB of additional disk space will be used.
Do you want to continue [Y/n]? 

Remove one package get another in its place.

What the fuck am I supposed to do with php5-cgi?

[editline]26th March 2014[/editline]

falco@falco-K53SV:~$ sudo apt-get remove apache2 php5-cgi
Reading package lists... Done
Building dependency tree       
Reading state information... Done
The following packages were automatically installed and are no longer required:
  apache2-bin apache2-data libaprutil1-dbd-sqlite3 libaprutil1-ldap php5-cli php5-fpm php5-readline
Use 'apt-get autoremove' to remove them.
The following extra packages will be installed:
  php5-fpm
Suggested packages:
  php-pear
The following packages will be REMOVED:
  apache2 libapache2-mod-php5 php5 php5-cgi
The following NEW packages will be installed:
  php5-fpm
0 upgraded, 1 newly installed, 4 to remove and 1 not upgraded.
1 not fully installed or removed.
Need to get 2839 kB of archives.
After this operation, 19,3 MB disk space will be freed.
Do you want to continue [Y/n]? 

How long will this continue?[/QUOTE]

I don't recall exactly how aptitude works, but can't you ask what kind of installed package depends on php5-cgi?

[nikomo]

You have some software that depends on PHP, so aptitude is just trying to keep everything functional whilst you rip apart the system.

[FPtje]

[QUOTE=mastersrp;44359938]
I don't recall exactly how aptitude works, but can't you ask what kind of installed package depends on php5-cgi?[/QUOTE]

It was more of a rant than me asking for help. I just uninstalled everything PHP related and shrugged it off.

Somehow it says that it needs to [B]install[/B] another package before it can remove that one package.
The joke is that I end up with more disk space being used after removing a package.

falco@falco-K53SV:~$ sudo apt-get remove apache2
...
The following NEW packages will be installed:
  php5-cgi
...
After this operation, 7789 kB of additional disk space will be used.

You can install packages with [I]apt-get remove[/I]. Count me confused if you can also remove packages with [I]apt-get install[/I].

[nikomo]

Imagine package "X"
X depends:
apache2 OR php5-cgi OR php5-fpm

apt-get remove apache2, we have package X installed, it has a dependency, and since the user isn't removing X, we have to keep it functional, so we'll install one of the other dependencies because the user wanted to remove the original dependency, not both the dependency and package X.

[deadeye536]

[QUOTE=FPtje;44360921]You can install packages with [I]apt-get remove[/I]. Count me confused if you can also remove packages with [I]apt-get install[/I].[/QUOTE]

If packages are in conflict, that's what its supposed to do.

[nikomo]

Package managers solve problems, they don't arbitrarily do what you tell them to. Unless you stick a million --force --yes-I-said-force --just-fucking-force-it --you-motherfucker-force-it at the end of it.

[larikang]

I feel like kind of a jerk.

I just changed this 600 word, 13 paragraph [URL="https://wiki.archlinux.org/index.php?title=REFInd&oldid=304016#A_dual_boot_setup_for_Archlinux_and_Windows_8.1_using_rEFInd"]opus[/URL] to [URL="https://wiki.archlinux.org/index.php/REFInd#Using_rEFInd_with_an_existing_UEFI_Windows_installation"]this[/URL]. I suppose the original author's enthusiasm is praise-worthy, but do you really need so many words to say "you don't have to do anything differently"? I always get nervous making huge edits on a wiki like this, but it seems justified in this case.

[nikomo]

[QUOTE=Larikang;44362982]I feel like kind of a jerk.

I just changed this 600 word, 13 paragraph [URL="https://wiki.archlinux.org/index.php?title=REFInd&oldid=304016#A_dual_boot_setup_for_Archlinux_and_Windows_8.1_using_rEFInd"]opus[/URL] to [URL="https://wiki.archlinux.org/index.php/REFInd#Using_rEFInd_with_an_existing_UEFI_Windows_installation"]this[/URL]. I suppose the original author's enthusiasm is praise-worthy, but do you really need so many words to say "you don't have to do anything differently"? I always get nervous making huge edits on a wiki like this, but it seems justified in this case.[/QUOTE]

I basically just skipped over the old wall of text, only way you'd get me to read that bullshit would be that I'm trying to solve a problem whilst using Arch and I have a problem related to that text.

[shodan]

[QUOTE=nikomo;44362510]Package managers solve problems, they don't arbitrarily do what you tell them to. Unless you stick a million --force --yes-I-said-force --just-fucking-force-it --you-motherfucker-force-it at the end of it.[/QUOTE]

apt-get purge libc6 --yes --force-yes

[Darkwater]

Forgot to mention, I installed arch on my laptop three days ago. Shit barely works, but I like having to fix everything. Really teaches me a lot about how everything works.

[shodan]

[QUOTE=Darkwater124;44368877]Forgot to mention, I installed arch on my laptop three days ago. Shit barely works, but I like having to fix everything. Really teaches me a lot about how everything works.[/QUOTE]

I find that arch is a great way to learn how everything works but is unusable as a stable environment to work in. That's why I prefer Fedora.
Not saying arch is bad tho, their repos/wiki are fucking godlike.

[reevezy67]

Arch is relatively stable if you don't mindlessly update. But I wouldn't use it for say production servers.

[Comfy]

[QUOTE=reevezy67;44369109]Arch is relatively stable if you don't mindlessly update. But I wouldn't use it for say production servers.[/QUOTE]

Anything bleeding edge on any kind of production system is just fucking dumb.

The only exception is when you really need a bleeding edge software (for example, the newest Ruby or python version), but then you would still have stable packages and only use the bleeding edge packages you need.

[vlad]

Arch has always been my distro of choice

[Squerl101]

Just putting in my 2 cents worth, but I think that there is no real "Best" Linux distro. I think that it depends on what you need. If you need lightweightness, use something simple like Lubuntu, and if you need something that you can customize to your own needs, use Arch.

If you don't give a Damn about any of these, use Linux Mint :)

[Darkwater]

[QUOTE=Squerl101;44369574]Just putting in my 2 cents worth, but I think that there is no real "Best" Linux distro. I think that it depends on what you need. If you need lightweightness, use something simple like Lubuntu, and if you need something that you can customize to your own needs, use Arch.

If you don't give a Damn about any of these, use Linux Mint :)[/QUOTE]

You could argue over the "best" distro for a specific need/goal, and you can always have a personal favorite.

[rilez]

Personally I've found Ubuntu/derivatives to break the most, especially when you start mixing repos. Edgers is hilariously broken, and the only way to get updated Nvidia drivers.

Arch is designed to be rolling, and easy to fix. Even though I have yet to have a problem with it... maybe its different on servers that have to do one specific task for a long time, but for desktop use its the most stable distro I've used

[editline]27th March 2014[/editline]

I also don't like how Ubuntu fragments Linux. Like their custom patched versions of certain GTK libraries for example. People build packages based on these, and then they don't work on the normal ones on other distros.

[mastersrp]

[QUOTE=mastersrp;44359938]Alright, so after I've gotten all of the sed work done that I needed, I stumbled upon a problem which I'm not sure how to fix. This is going to present me as a complete UNIX newbie, but that's the way of things.
Anyway, I have this RSS feed parser that fetches any kind of news feed, and outputs it into what I'll refer to as an intermediate format. It's somewhat simple to read for humans, but I'm not sure how to parse it yet. I realize that this isn't the smartest of ways, but bear with me.
Anyway, two feed items may look like this:

Title: Hello world
Link: http://www.example.com
Description: A very world of hello.
It can have newlines too. Potentially links? Or other content, like HTML tags.
<a href="link">not really a link</a>.

Title: The second item
Link: www.google.com
Description: No parsing is required of the individual item sections. I'll handle that with some simple regex.
However, I'm not sure how to parse this.

Ideally, I'll need it to work with Busybox Ash. One such way would be

for f in $(magic code that iterates over each item); do
  parse( $f );
done

Any takers? I'm kind of at a loss here.
[/QUOTE]

I solved it! For anyone curious, here's how I did it. First of all, I was wrong about the descriptions. They do NOT contain newlines at all. So that cleared that up, easing my development of this.
Second of all, those are actually RSS feeds parsed using [url=http://www.vanheusden.com/rsstail/]rsstail[/url].

# We need the newline character for cutting the strings.
# $1 is the URL of the feed we want to parse.
# $ITEMNUM is the amount of items in that feed, we want to parse.
# Using simple math, I determined that each RSS 'item' would be 4 lines, hence we tail the output of rsstail.
# We also run rsstail with the -1 parameter, to ensure that it doesn't continously fetch feeds, but just fetch the $count number of latest feeds.
# There Is Probably A Better Way To Do This®
NEWLINE="
"
for count in $(seq 1 $ITEMNUM); do
		RSS="$(rsstail -1 -p -d -l -n $count -u $1 | tail -n 4)"
		title="$(echo "$RSS" | cut -d "$NEWLINE" -f1)"
		title=${title:7}
		link="$(echo "$RSS" | cut -d "$NEWLINE" -f2)"
		link=${link:6}
		desc="$(echo "$RSS" | cut -d "$NEWLINE" -f3)"
		desc=${desc:13}
		pubdate="$(echo "$RSS" | cut -d "$NEWLINE" -f4)"		
		pubdate=${pubdate:10}
		# Do some magic tricks with the above values
	done

[larikang]

Done! The [url=https://wiki.archlinux.org/index.php/REFInd]rEFInd[/url] wiki page is about 9,000 characters shorter, with nothing of value lost.

[nikomo]

I'm so glad I don't have to fuck with EFI on my laptop, I have it set to BIOS-only mode, even though EFI is possible.

Also, got an idea with my avatar, now you guys have no excuse for not being up-to-date.

[rilez]

Efi is actually nice if you boot more than one OS

Puts the boot options right in your bios. And if you use Windows it boot s a bit faster

[editline]27th March 2014[/editline]

Also it uses GPT which is slightly better than MBR

[nikomo]

I know there's technological improvements in EFI, quite a lot of them, but the entire fucking thing is fairly fucked up, with manufacturers basically building them so that they'll ship it if Windows boots, without any care for standards.

I'll switch to it when it matures and I have hardware that ships with a proper implementation.

[rilez]

Secure boot is shit, yeah. Manufactures that don't let you turn it off can shove off. I have it off on mine, but still use EFI

[lavacano]

[QUOTE=nikomo;44372648]I'm so glad I don't have to fuck with EFI on my laptop, I have it set to BIOS-only mode, even though EFI is possible.

Also, got an idea with my avatar, now you guys have no excuse for not being up-to-date.[/QUOTE]

Are you using some sort of automated script to update your avatar?

[rilez]

Why the fuck is the new version of GTK so inconsistent. If you guys remember, I was complaining about their new title bars and how some things were in the GNOME Shell appmenu, and some things weren't. They fixed that, so some applications like Gedit handle this perfectly:

[img]http://i.imgur.com/BVOnw4q.png[/img]

All the options are now moved to the little gear icon, including preferences. Great, I can include this on other distros!

But then some other fucker decided to do THIS with context menus:

[img]http://i.imgur.com/GRmnmN6.png[/img]

Why in the HELL are these buttons in the title bar now?!?! Why is cancel spelled out and on the left, when all it does is close the window (which is on the right)?!

They've done this with some confirmation dialogs... but others look like this:

[img]http://i.imgur.com/Y0SdWHH.png[/img]

These look great... so why don't all the confirms do this. The buttons should be below the stuff you're reading. So you don't have to look back up to navigate through shit. It's so inconsistent

[nikomo]

[QUOTE=lavacano;44373445]Are you using some sort of automated script to update your avatar?[/QUOTE] Nope, but I'm thinking about making one.It's not much of a bother to maintain it manually, there's an rc release like every 2 and 3/4 weeks, if I remember correctly, so that's not much of a bother, and stable releases are also fairly rare.

[larikang]

I've been using UEFI exclusively for about three years and I haven't had too much trouble with it. The biggest difficulty has always been lack of documentation but that's getting better of course. Also Windows 7. UEFI with Windows 7 made me want to die.

It's really nice and simple as long as you
* don't buy a motherboard with shitty UEFI support (i.e. read reviews)
* can run up-to-date software
* and use a distro with EFISTUB in the kernel by default

Especially with rEFInd (which I just learned a whole lot about) and Arch, everything basically works out of the box.