Brief note on Manjaro and libraries, and install of Arch, then back to Manjaro

Distrohopping has not been my way, of late.  For a few years I have leaned heavily toward Arch Linux's infrastructure.  When stymied by Arch, I have several times thrown up my hands and installed Manjaro.

Manjaro is fine with me.  I am using Manjaro i3 Edition of late.  For over a year, I think.  I'm happy with i3.  Especially on a laptop, a few years old, i3 lag-freekeeps me lag-free.

This week I was furious when I learned that Manjaro (hence Arch) does not support the use of /usr/local.  I'm not ready to write a long piece about this, but it's pretty much the only big deal I have encountered with Arch, requiring an intervention to use it in my accustomed manner, my muscle memory workflow.

Well, maybe Muscle Memory is a little too much to say abot this glitch.  It's a PITA, though.   I installed a library in /usr/local/lib, and discovered that a certain program I had installed would not run, due to the need to take extra measured to ensure that /usr/local/lib would be visible to the linker.

The program was tideeditor, from the xtide suite of programs.  I have been self compiling these programs for years, without encountering this problem.  In about 1993 or 1994, I discovered Debian, or maybe 1995, after using Slackware for some years.  I learned from my experience with Debian the value of the /usr/local subtree: one thereby does not touch the files in the distribution itself.  I can install programs, do anything in /usr/local/ and it is mine, not part of the infrastructure provided by the "maintainers."

This week, after over 25 years of using GNU/Linux, I learned about this problem with /usr/local.  I went overboard, and decided to install Arch.  I was in a rush and apparently made some mistakes---not to mention that my storage is filling up on my Thinkpad Yoga.  The upshot was a rat's nest of errors.  I was unable to use Emacs. 

I had attempted to modify the instructions for a video for installing archlinux in 10 minutes using systemd-boot as the boot manager, while following the Archlinux Installation documentation.  It worked, but I seem to have skipped or messed up some steps, at least twice.  Although the new machine booted fine, I was unable to use Emacs as I am accustomed.

So after fiddling around a full day, and then some, I finalliy reached for Manjaro i3 Edition.

This is a masterpiece of guru magic.  Very well thought out.  A conky on the desktop screen has the important keystrokes on display in any open desktop. 
This installation is working extremely well, as I have modified it over time to fit my needs.  But not least of the fine features of this Manjaro i3 is the setup of the boot manager (which is grub.  That's fine.).  I was able to pick up the previous Manjaro right where I had left off.  And---this is what I like---now that the new Manjaro boot manager set up has been booted through and into the old install, on a different partition, it will remember this and boot to it every time, as long as I want.  That's a stroke of good magick.

Manjaro installed from a flash drive in perhaps 10-15 minutes til first reboot.  Few questions asked.

I follow this simple rule, which leaves me pain free: ALWAYS USE A SEPARATE /home PARTITION.  In any install of a distro, one can (in my experience) always find a way to specify that partition during the install process.  Step 2: I use dropbox for work that needs to be portable, including some config files.

Leaving me time to write the short bit.

By the way...

the program I was installing was tideeditor.  The library is libtcd.  I wrote a hasty email off to David Flater, the developer of Xtide and friends.  He answered with cut and pasted FAQs.  To his credit, he has always answered such inquiries from me, even though I have so often failed to check the FAQs before emailing.

Even more is this to his credit because, as I discovered when re-installing libtcd, a long message at the end, after typing "sudo make install," he has included a long explanation of this very problem---that the library is installed in /usr/local/lib, and that therefore some further steps may be necessary to make the library visible to programs.  WOW!  It's on me.

The best fix is to include a file in /etc/ld.conf.d/ pointing to /usr/local/lib.

And one question remains: why did this not happen on any earlier installs?

Useful Helpful Expedient Help for Programs on GNU/Linux

A Quick List:

• tldr :: (tl;dr == "too long didn't read") I installed as "tealdeer"
• eg :: gives examples for a given program
• GNU help2man ::  I haven't used, but it makes a man page from "--help" listing
• cheat :: cheat sheet for a given program

Where I'm coming from: 
  One of the things I like the best about GNU is the Man Pages.  It was, I think, a mandate: anyone can write a utility for a Unix-like Operating System, a system which is made up of multitudinous little programs that work together.  Today, it is becoming more common NOT to follow that convention.

  The man page format is something else, however: it seems to assume that the GNU/Linux user (and any Unix-like OS user) is able to read that crazy listing of arcane specs about a program.

TLDR solves my problem, which is the inconvenience of reading through a lengthy man page to figure out how to run a program.

Toward a HOWTO : Modifying cb2Bib to enable the use of the Annotation field natively

[I am posting this, even though it is is a month or two old, and not finished, with the intention of finishing it soon.  A new version of Cb2Bib was recently released.  It is still possible to use a patch to incorporate these modifictions.  AED] 

This Micro-HOWTO describes a method for modifying the source code of Cb3Bib, to make it possible to easily generate annotations and print PDFs of the database with comments or, if you like, annotations.  This is the one thing I have wanted most from Cb2Bib, Pere Constans's amazing reference database program.  More information, and the program itself, can be found on the website.  Most GNU/Linux distribution provides a package for Cb2Bib.

The instructions I am providing are rudimentary.  For now I will provide a patch for Cb2bib's source code, and a revised *bst for LaTeX's BibTeX component.  This is not necessarily easy, but it works.  At some point, I hope to present more thorough instructions.  I will not be able, unfortunately, to provide assistance in setting this up; I have barely been able to get it running myself.

Here's the workflow for literature searches, to generate a reference list for the libraries.

1. Search for the references using Google Scholar
2. Start up cb2bib (it's easier, but not essential, to do it before saving a citation.
3. Having previously set up Google Scholar to produce bibliographic citations in the BibTeX format 
4. Click on the Bibtex link underneath a useful item in the research results.
5. With cb2Bib running, the GUI text input widgets may fill in, if you are fairly luck.
6. If the Author, Title, Etc, text widgets are populated, select a *.bib database file, and save (click on the icon with a Floppy Dist).  Don't worry: this is not a complicated file; it can be edited using a text editor.  Emacs's Bibtex Mode is awesome for this.
7. Open a new window (I think this is an editor window) by clicking on the icon with a pencil on it, to the far right at the bottom.
8. Go to this new window frame, open "Files" and select "Postprocess and show ..", proceed.
9. A pdf will be displayed.  For me it is displayed either by Evince ("DocumentViewer") or Okular.  Whatever is the default.

This workflow is streamlined and easy.

Perhaps even MORE magical is the way Cb2Bib handles a selections from any source, in guiding the generation of a record for the database.  I will not discuss this.  I should mention that Pere Constans has spent a considerable amount of time developing an "Annote" capacity; I have not gotten my head around it, however useful it may be for others.

What is missing for me is a means to type in a comment or annotation, and the automatic inclusion of this comment in the printed PDF.  A trivial benefit would be to present the Call Number of a book. An annotated bibliography is the obvious target for this kind of work flow.

Pere has done a great service to me, in providing instructions how to modify the code to obtain this result. Now, on page 2 of the input screens is found a space to type an "Annotation", where "Annote" used to be.  Through modification of a BibTeX .bst file, to print the Annotation field.

First, the source code must be patched.  I have had to keep my eye out for some minor glitches that may come up during the implementation of the revised program.  I found it necessary to delete all files left behind by a previous copy of Cb2Bib, including files in the ~/.config directory under MOLSpaces.

Patching cb2bib source code

  

Linux Adventure: Manjaro, i3, and Ranger with notes on Arch linux

I have started running the i3 window manager on my Linux boxes.  I have installed Manjaro on both.  One of them is running Manjaro's i3 community edition; the other one, I think, started with an xfce4 version of Manjaro, to which I added the i3 window manager. 

It's been an adventure.  It's been this way for over 2 weeks now, and I'm not sure I'll go back to XFCE4 or Gnome.  I have distro-hopped for a good many years, and convinced myself of the advantages of various options along the way.  XFCE4 has gotten to be a favorite over the past "few" years.  It's pretty intuitive, and it doesn't get in my way much.  That's not much to say, but this isn't a review of XFCE4.  I would say, though, that if I do decide to abandon i3, XFCE4 would probably be a first step. 

i3 is something different: a tiling window manager.  I have tried a couple of these, and generally given up in a few minutes, due to the excessive amount of tweaking and steep learning curves of, for example, OpenBox.  i3, however, was pretty easy to use out of the box: I couldn't do everything in the world with it, but I could do just about anything, one way or another.  I reached some limit after a few tries, and abandoned it for a while, but I have come back, and challenged myself to learn enough to make myself dangerous. 

Manjaro is a compromise.  Arch is my goto OS, but my lack of boot manager chops generally abandons me somewhere along the line.  I was able to install Arch to my self-contructed box, Beast, but used a bootctl configuration.  Bootctl is much easier to configure on a new machine than grub, for my money.  I dislike grub very much,  But when it comes to multiple boots, I am out of my depth, so I installed Ubuntu, which does a decent job with grub out of the box, and leaves my machine easily bootable onto the Arch install on the machine.  Manjaro also does a nice job of installing. 

It has a peculiarity that has been noticed by others: once Manjaro has been installed, a manjaro splash screen is seemingly indelibly ingrained into the system, even when another OS like Ubuntu is installed over it.  I suspect that the developer(s) of Manjaro are quite knowledgeable about the fine points.  This bugs me, but I live with it.  Because Manjaro just works. 

It is an Arch Linux system with some differences, that allows me to use the AUR "repository" that makes Arch so amazingly able to easily install almost any software out there that conforms to basic source code packaging tenets of Unix. 

I am in awe of the cleanness of Arch.  The developers keep the components up to date remarkably well.  I have one Arch install that has been abandoned because of a mistake of my own installing a software package that required the Wine windoze emulator, and the result was that the desktop manager doesn't work anymore; and somehow the system seems to be corrupted.  I have to use startx to start XFCE4 or Gnome.  Ideally I would reinstall Arch. 

But I know enough to make myself dangerous; not enough to administer a GNU/Linux - Unix system at the lowest level.  I take shortcuts, so I suppose it is inevitable that I would be bitten.  But in general, I can go a long, long time just upgrading the Arch system without any issues that I know about.  I know about the precautions against deviating from the official repository, just like I know I should read every click-license.  But life is too short.

Today, I experienced something to remind me of how awesome Arch is: during an update of my Manjaro box, I started receiving a message that an update could not be performed because a certain file already existed on my filesystem.  I tried several things.  Then I remembered: I should do a search.  I use Google, with some regrets.  Duck-Duck-Go would probably work or other search engines.  But Google is an awesome search engine, as it has been since I started using it in about 1993, when it was just being developed at Stanford on Linux boxes. 

I searched the name of the file and "Arch" and "update".  A few key words usually work.  And up popped at the top of the search a message stating that upgrading requires manual intervention.  The name of the file was given, so I deleted it as suggested, and .....  Voila!  the update ran without a hitch.

This happens many times.  Well, not MANY, but it happens.  The secret is to compose a proper search, when one runs into a problem. 

Manjaro is an Arch Linux system, with some differences, most of which I don't know about.  It's easy to install, vs. Arch's somewhat more difficult installation---especially installing the boot manager.  I haven't noticed any problems with Wifi hardware for a while, excpety for an iMac, and that doesn't count as enything but domestic terrorism. 

Once it's installed it IS an arch system, with a few differences, some of which I understand.  The repos are different, so that updates are more hands on by the higher echelon developers of Manjaro.  The kernel is not as up to date, but it's reasonably recent.  I think if I understood, I could easily adopt a newer and crazier kernel, and I might try one of these days to bring the kernel up to the level of the NVME I have installed.  Most of the software on Arch repos is found on Manjaro's, but M's are not as up to date, and perhaps are under some micro-control, version wise.

Youart is a package manager with extended capabilities to deal with AUR packages.  This is already installed on Manjaro, requiring little setup to start using.  AUR. Arch User Repository (I think).

More to follow.

Mounting Google Drive on a GNU/Linux system

First, the following post worked, with the proviso/modification that I am not using Ubuntu.  I installed google-drive-ocamlfuse as an AUR build, and followed the simple steps from the post. 

Wonderful...

http://www.omgubuntu.co.uk/2017/04/mount-google-drive-ocamlfuse-linux

1. Install google-drive-ocamlfuse
2. run $ google-drive-ocamlfuse
3. $ mkdir ~/googledrive
4. get a cold drink ready
5. run $ google-drive-ocamlfuse ~/googledrive
6. open up a file manager!  (I used caja)

 
According to this post, one can open files with standard tools on GNU/Linux, unlike the web version of Google Drive.

But the following is a bit of a problem for someone who, like me, is forgetful:

1. When you are done, run this:
   

   $ fusermount -u ~/google-drive

WISHLIST: Scavenging titles from PDFs for use as filenames

Some Possibly Interesting New Finds (For Me)

• Several Programming Tools 

  I am not a programmer, even though I play around with code a little bit, from time to time, and some of my main tools involve writing code (gri).  BUT some of the best tools for my money (usually no money at all) have often turned out to be hacker level tools.  This leads to some problems.  But here's the deal: I learned enough computer science to be dangerous to myself, and, of course, to make myself useful TO MYSELF.  I took an electronic engineering course in summer school in 1982: Computer Architechture.  That was an amazing course.  We did some trivial programming in Assembly Language and learned the basics of what a computer it, on the hardware level.  I also took a Math course, for fun: Fortran.  The young Graduate STudent who taught the computer architecture course said this to me: you now understand what computers can do, and can make yourself useful in your work.  This is true. 
  
  In about 1986 I had been collecting notes on animal names in several dialects in the E. Caroline Islands, Chuuk and nearby.  I desired to develop a digital database of some kind, and even while working on an island with very rudimentary infrastructure and other resources, I was able to get started thank you to my Mother, who knew about this project and funded a Laptop for my work.  I was working for 11,000 to 14,000 a year, in 1987 dollars.  I learned about the Free Software Foundation through Info World.  Don't ask my why: it was free, I guess.  I had a subscription, and read it from cover to cover.  That was in the halcyon days of snail mail, when letters could take weeks from the US. 
  
  I needed a programmer's editor.  This is the first tool I needed, for two reasons: programmer's editors, unlike word processors, did not insert obscure and hidden codes in the files.  They produced ASCII files, where, in a very primitive sense, What You See is What You Get.   So those proprietary behemoths would not do.  But neither did I have the funds or the inside knowledge of what kind of program I needed.  A Linguist at the University of Hawaii, Robert Hsu, sent a demo disk for an editor, and it worked well.  It was MultiEdit, and it would have cost 300.00 to get a full on copy of the program.  That was more than half of my bi-weekly take home pay!  I needed the facility to enter simple diacritics over vowels.  (Even today, in 2017, it is not trivial to do so in this browser, and I won't try).   As nice as that program seemed, to get the manual to learn how to do this would require the purchase of a license. 
  
  In InfoWorld on week, about 6 or 8 pages in, was a small article about the Free Software Foundation.  Knowing nothing about their work, seeing the name nevertheless conjured up visions: was it possible that I could get a programmer's editor, of some description, from the FSF?  I wrote a letter. 
  
  Some time later, I received a small box with, IIRC, 13 disks, those 3-1/2" disks in plastic that were used in that time, with software of various kinds from the Free Software Foundation, developed by the GNU Project.  This insignificant looking little package was life-changing for me.  And, even today, I still am confounded when I realize that the Free Software Foundation gave me for free the tools to do what I needed to do, without so much as a request for a donation.  The "learning curve" was substantial.  But every thing I needed to do was possible with these tools, and the subsequent tools that were available as Free Software. 
  
  Notice I did not say Open Source Software.  That idea came later.  To be sure, Free Software source code is open.  Right now is not the time to reiterate the details or history of the Free Software Movement: for that, please follow this link for a bit of history about the GNU Computing System.
  
  The software in that box many programs that were especially compiled for the Windows Operating System.  The most important was Emacs.  I use Emacs to this day.  Also included were a number of utilities for manipulating text files.  I was thrilled!   A superior sort program for sorting my data; the program gptx, grep, find: all of these were crucially useful.  And these are but a few. 
  
  Over the months and years to come I relied upon this software to generate the data files.  Through the suggestion of Robert Hsu and other linguists from UH, I learned of the Band Format to generate free form data bases.  Over time, I came to use TeX and LaTeX to generate publishable output.  These were and are ALL Free Software: Free as in Liberty, in that I can use them without violating any copyright or patents; and usually Free in terms of money as well.  My work was truly enabled by this gift from the Free Software Foundation.
  
  From the start, these tools were not easy to use, but unlike almost all proprietary software I have used, I could use the full horsepower, not some limited powers as dumbed down by the developers. 
  
  I started receiving the GNUS Bull, a periodical newsletter from the FSF.  This was early on.  At some point, perhaps in 1992 or 1993, the GNUS Bull printed an article about free UNIX-like operating systems: FreeBSD and Linux.  Some time after, I travelled to Guam for medical reasons.  While I was there, I made arrangements to download several disks for the Slackware Linux distribution.  This was the beginning of a long dependence upon Free Software.
  
  Immediately, my computer was faster.  Multitasking was a real process, not a faux multitasking as on Windows 3.0. or 3.1, where multiple tasks could be queued up, but only one at a time could use the CPU.  
  
  Over the years I have tried to explain my preference and  reliance on Free Software to my friends and associates.   I am saddened looking back, to realize that few if any of my friends took this seriously.   I am saddened that I have not been able to share the tools that changed my life.   Often I have been met with derision, as the butt ofa  joke, for not being more open minded to the expensive software tools that often were provided by the schools I worked for.  Software companies worked very hard to engage schools with their proprietary systems.  It was a matter of prestige for School Districts in Micronesia, where money was scarce, and millions of dollars could have been saved... 
  
  The tools are much better, from my perspective.  Perhaps I failed in not teaching computer literacy or computer science.  I did engage a numbe rof students, like one on Saipan, who learned to install Linux on some out of date systems we were able to scrounge, or on new computers we built to showcase the concept of saving money through use of Free Software. 
  
  So my toolkit includes less glamorous programs than those of many of my colleagues and friends.  But as time has progressed, some of the great Free programs have come into greater use. 
  
  I now want to mention a few more recen ttools that I have found useful.  These tools are not mainstream advance wave programs.  Their utility is great, in many ways exceeding the facility of fancier GUI tools. 
  

  Ranger

  Purportedly a vim-based file manager, this text based utility is extremely useful.  Potentially: IF I can sort the wheat from the chaff.  Today, I managed to delete many files in one fell swoop by not understanding the usage.  I don't know whether this was my fault.  The good side: I don't even remember what files I lost: good, because I don't have to worry about it.

  Silver Searcher 

  A grep replacement I think that works better than almost anything else I've found to search my extensive org-mode main folder.  

  Ripgrep

  I think an even better silver searcher than Ag. 

  
  

  i3 Tiling Window Manager

  This is the first tiling WM that I have been able to get my head 1/2 way around.  The bad part is I haven't yet understood what it is I cannot do.  

  
  

  XFCE4

  My goto Window Manager when other experiments lead down rabbitholes or blind alleys.  GNOME 3 is extremely interesting except for that it is, for me, almost unuseable.  That's bizaare.  Issues:

  I need a menu.

  Why the heck do I need to push the mouse to the upper left edge of the display to make the desktop switcher images appear on the right?!  

  
  

  
  

Tide graph experiment: seeking a colorblind friendly palette

This is a first try.  I am working on a graph of height of tide as a function of (x) clock time.

This time, I have used the "Juxtapose" web tool to compare a graph (work in progress) with a simulation in The Gimp of what this image would or might look like to a person who is somewhat  colorblind with Deuteranomaly (apparently weak green vision).  For this simulation I have used the Color Vision Deficiency python plugin for the Gimp.
We can be pretty sure that these figures do not look like this to colorblind people; but I can see that my intense, contrasting color palette does not work to discriminate the different lines for persons with color deficient vision.

The color deficit is Deuteranomaly weak green. Pull the slider to the right to see the original image; slide to the left to see the simulation of what we may think a person with this deficit in color vision might see.
Scroll further down to see another type of colorblindness.
The objective will be to produce colorblind-friendly graphs. I found this on the blog of the website with the slider maker: https://knightlab.northwestern.edu/2016/07/18/three-tools-to-help-you-make-colorblind-friendly-graphics/index.html

1.  http://colorbrewer2.org
2.  Chroma
3. Checking: other ways to visualize/simulate your work

•  Color Oracle (available on Archlinux AUR)

                  [Depends on Java 6, not officially supported by Arch]

• Sim-Daltonism(iOS or MacOS only)

My most useful applications

Emacs

 

Photo Management

 My criteria

1. Quickly sorting photos
2. Tagging is quick and flexible
3. Tags and comments saved to file metadata
4. Basic cropping and renaming simple and easy
5. Opens from File manager
6. Printing (Gthumbs is nominally best so far: prints multiple marked photos on a sheet)
7. Does not save in a non-standard way, or to some bizaare kind of database; uses file system paradigm; photos are saved as files that are easily copyable; files and storage paradigm are accessible from other photo management, etc., software

Photo Mgmt Packages I am evaluating

1. XnView: seems at first glance to do most of the above
2. GThumb lacks some functiionality, but best for printing several photos to one sheet.
3. nomacs: lots of bells, seems quck, interesting
4. NOT Apple products
5. Geeqie:
6. Ristretto (xfce4) 
7. Viewnior

File Management

1. Files:  Standard Gnome FM (was Nautilus)
2. Caja: Does dual pane; works well with Dropbox; not shown on other menus
3. PCManFM: has advantages, don't remember what
4. Thunar: (xfce4) seems mature, cool icon, many tweaks available
5. Dolphin: Does dual pane
6. Ranger: quick and interesting, but vim bindings
7. Emacs dired still is the goto for general wandering
8. MC

 Photo and Graphics Editing and viewing

1. The Gimp
2. Inkscape
3. Gv for postscript
4. 
   

My Toolbox

Many of these are reason enough to stay with Arch Linux.  It's a bit hairy in Ubuntu---without even getting into thoughts about the philosophy of the distro--to keep all of these up to date

1. cb2Bib: Bibtex refernce management a la extraordinaire
2. gri: graphing
3. Xtide
4. Xephem
5. TexLive (This can be installed from Arch AUR; upstream for other distros)
6. xfce4 Terminal and Konsole both allow random background colors for new windows.  I want more control, but it helps alot. 
7. dictd and whatever possible dictionaries.  Ubuntu/Debian gives the most complete selection of dictionaries.  
8. 
   

Window Managers etc

1. Gnome is ok, too much overhead, oversimplified.
2. KDE is hell, but has many of the best apps (highest overhead).  some apprent incompatibilities, on Gentoo, updates always were mangled when trying to update libraries for KDE that conflicted with other software in various ways.  
3. XFCE4: simple and fast, works ok
4. i3: Attractive, and gaps may be even better

Browsers and other Internet tools

1. Firefox: still too heavy and cumbersome.
2. Google Earth 
3. Dropbox
4. Transmission
5. Conkeror (nice (emacs key bindings). 

 SYSTEM Tools

1. Top
2. Htop w/ temperature sensor patch
3. psensors
4. some applets or indicators
5. GParted
6. pavucontrol / alsamixer 
7. NetworkManager / nm-applet or whatever
8. Youart rocks!  I also install Packer.  
9. Orage Calendar
10. evince 
11. okular is excellent for many purposes for PDFs, but I've encounted issues, maybe with printing?  In 2017, Evince (Document Viewer) is working fine.
12. Some random AUR gui tool
13. 
    

Progress report on The Beast

A short note about the OS. 

I have settled on Arch for this machine.  Ubuntu 17.04 has a partition but the new Arch system's boot setup does not give any dual boot access.  That's fine with me for now, and I suppose later this could be solved in any number of ways. 

Boot management on GNU/Linux systems is an arcane field.  For myself, I only ever cared to know enough about it to get a system running.  I hailed GRUB when it first appeared, as it seemed a human readable/writable form.  But soon the wizards towed it into the wood of obfuscation; it was not longer a pleasure to use.  With UEFI, I, at least, faced just one more unmanageable layer of intractability.  Arch required to fiddle with boot managers, GRUB 2 in particular; for me, the confusion often left me grabbing for Ubuntu or other distros that are easy to set up. 

There is something to be said for easy to set up.  I can have an Ubuntu system up and running in less than 30 minutes, with the new speedy machine I did it in maybe 5.  To install the core of my software needs would take part of a day, and to get it polished often took longer.  The cost of Ubuntu, in my terms, was the need to compile several of the more important programs to support my work flow, and generally to remember them and keep them up to date.  It was and is a pleasure to just go to work.

Arch Linux, once installed, is ideal: almost every program I need is available either as one of the large number of blessed programs on the repositories, or as a PKGBUILD on AUR, the Arch User Repository.  A PKGBUILD is a script to download the upstream source code, build the object, and install it.  It is easily maintained, for example using yaourt.  Maybe I've just been lucky: almost every program I wish to try is on AUR.  At last count I think this was around 44000.  Debian stable may have more in terms of numbers, but it doesn't have some of the packages I need.  PPAs exist for a few, but they easily go out of date.  I can easily edit a PKGBUILD to use newer versions as they become available. 

Bootctl

This time around I resolved to install Arch, and I had a few hours to do it.  In the long run it would save time.  Less fiddling.  I watched a youtube video on installing Arch in 10 minutes, and learned about a new boot manager: bootctl.  Without going into detail, this method pretty much worked for me, to get Arch installed in a very short time.  I'm not going to go into detail.  I did have to work some wrinkles out, but the instructions on the Arch Wiki and perhaps other sites were straightforward.

The Arch Way

I don't know whether anyone will read this, but I want to comment on the arrogance of Arch forum rats about "the Arch Way."  Debian's mailing list used to be the gold standard for trashing noobies.  Gentoo developers were much more helpful and tolerant, and the docs were perfection, and have never been equalled in over 10 years.  Responders on Arch forums and perhaps mailing lists are unforgiving with noobies who want to take shortcuts to install Arch.  It's a big deal.  Arch devs will not produce a liveCD installation.  There are a number of alternate approaches, and I've tried many ---because of my grubophobia, and, truth be told, the problems of UEFI/EFI, and actually some confusion---now, I think, at least partly resolved in the Arch installation tutorials about how to mount the EFI partition. 

In the past I have successfully used Antergos, and converted to pure Arch; the last time I tried, this was no longer useful.  I installed Arch on an iMac using "archanywhere," a script, but it has failed on my laptop.  For me.   Some other approaches did not work.  Once in a while, I was actually able to install Arch using the Arch Way.  More recently, I installed Manjaro i3 edition on my laptop.  It's a beautiful system.  Easy to install.  But it does something almost criminal to capture the system boot process, so that other installs often fail to install their own GRUB, but install through Manjaros.  I was able to convert using a 10 step process (provided by Mr. Google) to Arch.  But the conversion was partly abortive: new kernels never got properly set up, during the frequent installed. 

What I want to say about the Arch Way is this: contrary to the claims of the Arch Way adherents, installing in the Arch Way, using the command line, does not teach much about the inner workings of the system: at least anymore.  Gentoo installs, almost impossible on ordinary hardware in my opinion, do teach a lot.  Linux From Scratch does.  Arch uses faily complex scripts to do most of the heavy lifting.  I was surprized how easy the installation was.  It's getting "better."  I am, in some sense, though, an interloper, and I have never contributed a PKGBUILD.  Perhaps I will do so soon: I wish to install more dictionaries for dictd.  Debian has many more, and I have more or less a recipe for doing this when I have time to sit down an focus on it.

Hardware notes

I'm not going to get into much detail about this.

1. NVMe storage is not operating at full speed, but it is very fast.  
2. Scrolling and moving the mouse leave a trace in the audio: speakers exhibit a background buzz/hum, not so bad though.  Ideas I've seen (this is an FAQ online): PSU (Corsair CX650 Bronze) or inteference with the integrated audio.  Logitech wifi mouse.
3. temperatures are very cool: I see CPU maxes of 59C, and the corse are running at 26/27C as I type.  I feel not heat at any time when I put my hand on the box.  psensor is great for monitoring temps.  Htop has a version on Arch already with patches to display temperature on the front page.
4. Good choice department: The Ultrastar 6.0Gb/s 64MB Cache 2TB 7200RPM drive is pretty fast for an HDD.  
5. This case has no provision for an Optical Drive.  Trolls online make comments that they are no longer a thing in the new age, but I disagree.  The brackets for HDD and SSDs are something new, and cables provided wth the MB and PSU do not make it easy to access them.
6.  The Motherboard seems to operate well with GNU/Linux.  The UEFI BIOS is nice to work with.  It would be great if the utilities were available for GNU/Linux.  [C was developed for cross platform compatibility, but Micro$oft did everything possible to end around standardization.  Apple comes closer with an embrace of the Unix idea, but builds kinks into the little bits, to make them difficult or impossible to use outside the Apple infrastructure.  One shudders to think what has been the cost to the global economy of this proprietary locking of technology.  Apple put some nice pieces on CUPS, developed in the Free World, bought CUPS, and does not give back to the free world upon which their implementation is based. ]
7. The logitech wifi mouse does not work well even three feet away from the receiver, probably due to interference.  When a USB drive was plugged in, this was expecially noticeable.  Perhaps I need to clean up the wiring, untie some lines to avoid crosstalk?  (what do I know).
8. I use Dropbox.  It is central to being table to set up a new machine quickly in a familiar way.  Most of my ongoing work is on the Dropbox, so once it's installed, I am right at home.  It doesn't take so long to get a machine or a new linux install going, with Arch.  I need to collect a list of all installed package as I was wont to do with Ubuntu/Debian some years ago.  
9. KUDOs to Arch devs, who do amazing work.

New Pure Linux Computer

I am too stoked about the result of this project to spend much time writing about it. This will be brief.

I have assembled a machine that is blazing fast, due to a trifecta of parts:

1. A super fast CPU: i7-7700K 
2. A futuristic Storage device: Samsung 960 EVO NVMe 500GB
3. Moderately fast DDR4 RAM: 16 GB of 3000MHz Corsair Vengeance 

Also included are

1. a GTX 1050 4GB Graphics Processor
2. An ASRock Z270 Killer SLI/ac Motherboard
3. A Refurbished 2TB Hitachi Ultrastar 6.0 SATA with 64MB of cache

The ASRock Motherboard was not going to be my first choice.  It has basic features (believe it or not) compared to some others.  It does have integtrated Intel Wifi and Bluetooth adaptors and an Intel integrated HD Graphics adaptor.
What was nice, the Graphics and the Wifi worked out of the box with Ubuntu GNU/Linux.  That was nice.  There remains a bit of wonderment about flashing red LEDs.  These have much been written about.  Perhaps they are the glitz?