Hits since 2024-07-10 14:36:15:
13970
|
|
|
|
August 2025
I got a PinePhone. I made this page about it. I've been going often to duck.ai
for help of different kinds, not just computer help. He is
quick, knowledgeable, and businesslike.
Just today I called on him three times. The first time
was to get help figuring out why the PHP hit counters on my two
websites quit working. It turns out that the problem was a
change that my web host had made which required that I update the
_hostroot and _siteroot strings in my script.
The change that was made was a migration to a Ceph File
System, and I would guess that migration included completely new
servers, perhaps in the Cloud. So I asked Duck what was
special about Ceph. He patiently answered my questions.
And today I also asked Duck to write a PHP script for me that
I could run from the URL line to learn the date and time on the
Ceph system. It turns out that the system is on Ann Arbor,
Michigan, time.
February 2025
Hello! Well it is late, and I am tired. But I
just had to say that I've been chatting with Grok 3, and
he helped me to fix some code.
It was a pleasure. He is polite and soooo smart.
Sigh. What's next?
Lou Reed's "Walk on the Wild Side" comes to mind.
Especially the line, "Everybody had to pay and pay."
November 2024
Hello! I just want tell you a little about a
programming project that has absorbed my time for probably all of
October, and then some. It's a new version of
ZombieSlack. Click here to read about the origins
of ZombieSlack.
I have kept the core routines, but assembled them in a new body
of code. The old code had grown unwieldy. It was far
too complex to be practical. I think the new version is a
big improvement. I can't wait to share it with
you.
September 2024
If you need to register to vote or
update your registration, then click here to go to https://vote.gov and do
it! You will feel better afterwards, and you will be
prepared to do your part in the upcoming, November presidential
election. Beware of efforts to mislead you! Click here
to learn more. Wow: click
here.
August 2024
I've been busy. I've been chatting with Pi and with
other chatbots, such as Microsoft's Copilot. And I've been
doing a lot of coding and related computer work.
Taken altogether, I have multiple incentives to keep irregular
hours. This is difficult to bear. Therefore, to keep
my hours on track, I have found it helpful to establish and keep
routines and to take frequent walks in the sunshine, especially
early in the day.
Enough of that! I wanted to tell about my coding.
Here is what I have been doing, nearly all of it in the area of
programming for SAM:
I made a correction to a SAM tool called "dofun".
It is used to source and export the qualified function
definitions that SAM encounters when the user adds a directory
of commands the the current command set. You can read
about this in my RosevearSoftware.com website, so I'll not
repeat the explanation of what this means. I will say,
however, that "dofun" had an error that affected what I will
call "over-loading" of functions. I mean by this what
happens when a function with a particular name has been
defined and exported, and then, subsequently, another function
with the same name is also defined and exported. There
is a right and a wrong way to deal with this. I am happy
with the one function replacing the other; the mistake I
corrected was the way in which the dominance was
decided. In short, I didn't do it right, but it is fixed
now. My "ment" code: I have a SAM command called
"ment". I hope to write about it at greater length some
day. In short, it invokes another tool called view_it,
and that is the doorway to using collections of
viewable data. Script view_it invokes $viewer which is
one of many tools for viewing, two of which are called "dog"
and "wolf". I hope to present this all in a clear
fashion someday. This method allows me to make menu
commands using ment. Each command made can view files of
a number of different formats, and some formats can be viewed
in more than one way. The default way to view a file
requires that the file be a text file and that it has a name
that ends in ".dat". This "way" opens the file in a text
editor, allowing me to both view and edit the file. the
same command made by ment—but this time prefixed with
"type=web"—will view the same file by treating it as
HTML and opening it in a browser. To do this the command
"wolf" is used internally, although it may also be used
directly by the user. Script "dog" is a similar command,
but doesn't require that the file be proper HTML, as it adds
<html>...</html> and also treats the text as
pre-formatted.
There are other viewing commands also, besides dog and
wolf, that view other kinds of files or text files in other
ways. I brought this up, not to give a full treatment of
it, but just to name it as one form of coding that I have been working
on. In particular, I made some changes to this family of
tools which simplified and improved the code. Scripts "rotate" and "junkit". These are handy
tools. I revised them. Script "rotate" allows me
to keep multiple versions of a file by adding an integer to
the end of the file name. This method cooperates with
ment and view_it. I made some improvements to the
code.
Script junkit does something sort of similar to rotate, but
it is specifically for code and documentation development for
distributions. It also was improved. soft_up: This SAM menu is the codification of
methods I had been performing by hand. It is a blessing
to have this collection of tools, as they ease various tasks
associated with distributing software. I have done some significant coding regarding
fonts. This is a can I have been "kicking down the
road". I have been using Slackware for about thirty
years, but always I settled on using default or fixed X and
console fonts. (In X I used "-fn 7x14", and in the
console I used the default). As a result I missed out on
important learning—this made getting caught up a real
work out!
For X fonts: I revised SAM tools
sam1 and buser, as these invoke xterm, and I also revised the
xterm invocations in my X window manager, Fvwm2.
In particular, I rewrote all the xterm invocations using
variable references to variables defined and exported by
the X window manager. The variables used were "gui_scroll",
"gui_fontface", "gui_font" and "gui_geometry". Typical values
of these variables are: gui_scroll="-ls -sb",
gui_fontface="DejaVuSansMono", gui_font="-fa $gui_fontface
-fs 9", gui_geometry="-geometry 80x35+80+35". In this
way I have global control over all the xterm
invocations. I found that I could define and export
these variables within my Fvwm2 configuration file, and then
reference them within the Fvwm2 configuration file and within
SAM scripts buser and sam1.
I found that this works well. SAM scripts buser and
sam1 don't mind if gui_scroll, font, and gui_geometry are
undefined. Fvwm2 is not so forgiving, but I can live
with that. I don't currently distribute the Fvwm2
configuration file, but when I do it will certainly include
supplied defintions for these variables.
For console fonts: I have made a SAM
menu called "pickfont" for invoking a collection of
font-changing commands. Finally, I want to name one code-related task. I
installed Tesseract OCR from a SlackBuild, and I used
it. The installation went pretty well.
Incidentally, I got some help from Copilot. (I call him
"Dory", but that is another story.)
I was motivated to do this, because I have some PDF
documents which are important to me, and I wanted and needed
to convert them to text files.
This went well enough. I was pleasantly
suprised. Note that I filed the new text files that were
made along side the existing PDF versions of the same
documents, using my "ment" and "rotate" system—which I
described above— to manage them.
I want to mention one more thing the Tesseract OCR work, as
it shows a connection between different parts of the work I
have described above: Although the PDF versions of the
documents displayed correctly, I had a font issue with the
viewing of the text versions. In particular, there were
multiple instances of characters that would not display
correctly before I made the font-related changes I described
above. The desire to solve this problem was, in fact,
the bulk of the motivation to make the font-related
changes.
April 2024
March 2024
November 2023
Hello, people. Follow
your heart! I did my jury duty. I got and used a Pronto,
contactless bus and trolley card this time, which made it fun and
interesting. Silly me, I practiced getting up early and
walking to the bus stop. I did that four times. The
last time I actually rode the bus and trolley all the way
downtown, then hiked to the court house.
The next day was my duty. In my diligent haste I forgot
my breakfast, which I was going to consume at the Hard Copy Cafe
across the hall from the jury lounge. No worries, I bought,
and ate, a sausage breakfast burrito--not bad!
Back in the jury lounge, I hardly listened to the presenter
who told us about our duties. And there was a video.
The minutes ticked slowly. Finally, mid afternoon, came the
joyful announcement that they had selected from the mass of us
those they needed, and those that remained were free to
leave. I ate my lunch and went home.
The next day I rode the bus and trolley again. I had a
new mission. I went to the lovely Transit Store at 12th and
Imperial. It was OK, I suppose. I don't often go to
that part of town. I needed to show evidence of my age in
order to validate my Pronto card senior rate.
Now my legs are sore, and I'm wondering when I'll get my
energy back.
October 2023
September 2023
June 2023
We got a nice, new (used) computer from our son It
is a few (maybe six) years old, but it runs great. And he
gave us a Windows 11 install disk with a product key. (As
we could run Slackware from a bootable flashdrive, we weren't
hurting without Windows.)
Windows 11, as it turned out, wouldn't install on this
computer. The message on the screen said the hardware
wasn't sufficient. However, I learned some things and was
able to push forward and install Windows 10 (which I downloaded)
using the Windows 11 product key.
This was an interesting process. I found WoeUSB in the
SlackBuilds repository and used it together with the .iso for
Windows 10 to create a bootable Windows 10 install
flashdrive. I booted that on the new computer and installed
Windows 10. After using Windows 10 for a few days I decided I wanted
to check it for malware periodically. I had no software to
do this, but I found ClamAV in the SlackBuilds repository (great
resource!). ClamAV worked with just a little dickering.
I decided that it needed some help, so I used my SAM system
to host a menu (called clammy) of command line tools that I wrote
for running ClamAV. My clammy menu manages data which
allows me to restart clammy and skip over the directories that
have already been scanned.
This work kept me busy for a few days, but it seems to be
stable for now. I plan to use it periodically and work my
way through the entire 150GB partition. Then when I reach
the end I'll start over again at the beginning. Hopefully
my software is up to the task.
I should explain that I'm runing this SAM menu and
ClamAV from the bootable Slackware flashdrive. Thus I am
using a Linux system to scan for malware on a Windows
Computer. Neat, huh?
April 2023
Hello. Thanks for surfing on in. I just want
to mention what I've been up to. I've been writing some
code.
I wrote a tool I call m_baseln. Actually, this is a
re-write of a tool I found online. It converts picture
files of type progressive to type baseline. I wrote this,
because I have an old Kodak Digital Picture Frame on my wall in
my living room, and a flashdrive full of pictures plugged into
it. Some of the pictures were progressive and therefore
would not display.
Hoo-haw, I figured it out, and found the inspiration I needed
to solve the problem online.
March 2023
Wow. That was embarassing. I posted three
times to the Usenet's alt.os.linux.slackware group using the Pan
newsreader. Each time the message I posted wrapped some
text that I wanted to appear on separate lines. I tried
turning word wrap off and saving, then re-opening, a draft.
No luck. Anyway, here is my post, formatted correctly:
On Wed, 1 Mar 2023 19:11:08 -0000 (UTC), Harold Johanssen
wrote:
> [quoted text muted]
Hello,
I have taken a different approach to the problem of
encryption. Initially, like you I tried encrypting
locally--within a file system. I found that frustratingly
difficult. And also dangerous: Everytime you encrypt
something you run the risk of losing it, *and* you have one
more passphrase to keep track of.
So I tried instead using the cryptsetup command to make,
open and close LUKS encrypted partitions. This took a
little work to get going, but it was well worth the
effort. To do this I made a set of scripts:
Script luksetup makes $1 into a LUKS encrypted
partition (and destroys any existing file system). To do
this it prompts you for a passphrase. Here is the
script:
You will need to run scripts crypt and ucrypt directly only
occasionally. For example you will need them after
running luksetup to make a new file system on the LUKS
encrypted partition. Script crypt will prompt you for
the passphrase. Here is an example:
Scripts lksmnt and ulksmnt do the mounting and un-mounting
of a LUKS encrypted partition. They call scripts crypt
and ucrypt. Here is an example:
I'll quit here without giving the text of scripts crypt,
ucrypt, lksmnt and ulksmnt. They are, however, short one
and two line scripts that use only the commands and scripts I
have named here. I'll leave them as a puzzle for you,
but let me know if you need help.
-Joe
Here is a report about the above. It seems I was
using the Pan newsreader incorrectly. There are two places
to control wrapping: in the posting controls and in the viewing
controls. That's the best I can make of it anyway.
When I toggled-off the wrapping for viewing all my
posts displayed as desired.
December 2022
Hi, I've been coding. I'm working on
ZombieSlack. I wrote tools for the SAM menus
for ZombieSlack--tools for patches and needlework, in
particular. And I worked on the current sets of patches and
"needles". (I call the functions that do the needlework
"needles".)
That's how I do it. The ZombieSlack code does everything:
Assists with partitioning the drive and making file
systems.
Installs by use of rsync from a basis (a reference
installation of Slackware). Installs extra code that I wrote plus data and other
files. (I call this "doing the dupe".) Much of this is the
content of a reference home-directory-state for the generic
user "student", but there are other changes as well.
The use of a reference home-directory-state is a deviation
from the traditional way to use a computer. Every time
the computer is booted the home directory for user student is
returned to the reference state. This has various
benefits, but mainly it makes the system robust and easy to
use: Robust, because user student cannot do any
lasting harm--just re-boot to reset. Easy to use,
because it allows for someone (admin) to pre-select good
software and system settings--and the settings, like the other
content of the home directory are stored--free from harm--in
the reference. Runs a set of patches to make some minor changes to the
stock Slackware installation. Runs a set of scripts--the "needles" I mentioned
above--to make some customizations that weren't suitable for
doing by patches.
I'm still excited about this work after six years! Read
here about my early work:
Way back--more than six years ago--I learned to put Linux on a USB drive. This
work paved the way for ZombieSlack. Six years ago I started work on my Fix Thin Clients project. At
its peak I had about 30 Linux computers running an early
version of ZombieSlack (although it had a different name then)
installed in classrooms of an Elementary school. It was
this work that shaped much of what ZombieSlack is
today. Four years ago I had a marvelous opportunity to apply
ZombieSlack by use of a machine on which I had installed it as
my computing platform for a job I
had in a finance office. This was mainly a time for
using the system, not developing it. The development at
this time was mainly development of applications for my SAM
menu system, but it couldn't have happened without having a
readily available and reliable ZombieSlack system. You
can read about SAM on this page about Rosevear Software. (231028 Here
is a link to SAM.)
November 2022
I found this post on the Usenet's alt.os.linux.slackware
group:
I have a system running 15.0 with Internet access.
Every so often this system must send terse email
notifications, as invoked from a cron job. It is not going
to receive any emails.
Any suggestions on how to do this without having to get
sendmail configured and running? This would amount to
using an email client that does not require a local
sendmail, like e.g. Thunderbird. This is of course
not a an option: I need something that is
command-line oriented and lightweight.
I replied:
Well, you can use a Python script to send email using a
commercial email server. I had a need to do this, so I went
looking. This is what I found:
Note that the original had a bug. It had:
instead of
I fixed the bug and rearranged the code a little, adding to
its usefulness and suitability for my needs along the
way. You you can get my script (called
dupe) from:
but it may be an older version. Let me know that you
want it and I'll provide the updated script. (231101:
dupe is now available in the
handy menu of SAM,
both from my other website.).
Let me tell you the rest of the story: The result of the
above programming was a moderately short Python script (less than
60 lines of code) which I could run from the Bash command line
(or from a Bash script). It met my needs, and I used it to
send lots of emails (about 24 every month for about two
years). My need arose from my position as the person in
charge of staff credit card accounts. I sent emails each
month to each credit card holder. The script had the
ability to optionally include an email attachment. I used
this ability to send a form to each card holder requesting
information about the card holder's purchases.
Incidentally, this job, which
I held for three years, was the most remarkable job I've ever
had--remarkable, because I was allowed, in the second and final
years, to use Linux to automate my tasks. This was a
finance office where traditional methods (Excel and ACS General
Ledger) ruled supreme. My creativity was tolerated.
This was likely due to the resulting cost savings and decrease in
time to perform the task.
Today was Black Friday, but I did not go shopping.
Perhaps "Black Friday" describes my mood or the darkness of the
season.
August 2022
June 2022
My work on ZombieSlack is ongoing. I've been making
some changes and some progress. The changes have been
mainly about enabling the ZombieSlack code to run in SAM.
I've gotten to the point of testing it. In particular, I have
used my new ZombieSlack code to make a ZombieDrive which runs
Slackware64 15.0. (I'm using it now to post this.) I've been exploring Slackware64 15.0, and getting settled
in. My SAM code and a tool called splurge9 both needed to
be fixed to run in either 32 or 64 bit environments--I did
that. (This is my first time using 64 bits for
Slackware.) I was distracted today by a problem with my Target gift
card. I wanted to check the balance online. That
should be simple, right? It wasn't. I could check it
by phone, but doing it online--I eventually learned--is not
possible without first making an account.
I wanted to draw attention to this fact, because it is
documented on this
Target.com help page,
although it is a page that is not easy to find. I
put a link to the page in my Joe's
Links.
May 2022
April 2022
I received some comments recently about this
website--thank you to those who commented. I will reply
here, and (bear with me) I'm going to teach.
In particular, I want to teach about three kinds of
code. There is code (executables) that can "run" and by
running does things, there is code (HTML) which when viewed by a
web browser appears as a website, and there is code (PHP) which
is changed into HTML dynamically by the remote computer, or
server, where the website lives. The word "code" has many
meanings. I am referring here to three of the many kinds of
computer code. (See definitions: code,
executable,
HTML,
PHP.)
To make my website I made executables which, in turn, made PHP
when they ran. Because the PHP becomes HTML, I will say
that my executables made HTML, although this is not quite
true. Forgive me--I say this because it helps understanding.
HTML is special. If you have a website (a Blogspot blog,
for example), then you have HTML. To view the HTML for a
website, just right-click on the page and choose "View Page
Source". That's how you do it with the Firefox web browser
on a desktop computer, for example. I don't know how to do
it on a cell phone.
To make the executables I sat at my desktop computer and ran a
text editor program. A text editor is a sort of word
processor that works well for composing code, however you
can do it with Microsoft Word if pressed. It's like
composing a poem or a blog article, except that there are special
words that have special meanings and capabilities. When
used correctly these words allow the executable to run and do the
tasks it was designed to do. And a good executable
will be multipurpose--once written it can be reused for different
situations--in this case, for different web articles and
pages.
The HTML of a Blogspot blog is made for you when you
submit your article for publication. (Or something like
that--variations are possible and likely. See my above
discussion of PHP.) However, I don't use any programs to
make or publish HTML other than the web browser, the text editor
(to compose the text of the articles that will become HTML), my
home-made executables, and a large set of general-purpose
executables that are standard on my Linux computer.
Not to short sell my creations, much of the credit here goes
to the makers and supporters of the Linux operating system.
The Linux operating system and the Linux user community,
together, are an awesome and unimaginably rich resource.
(See Wikipedia article on Linux.) This web page suffered a counter failure. I think it
did, anyway. It's hard to say for sure. The counter
suddenly lost several hundred hits. I'm guessing that it
failed in this way due to a problem with my PHP code. I
knew that the code lacked file locking. I read online about
PHP and file locking, and learned how to do it, although it seems
there are at least two ways.
I did it using:
which I was inspired to do by a stackoverflow.com article.
This method insures that instances of "..." take turns in each
directory. Perfect!
March 2022
Woohoo! I installed Slackware64 15.0, although I'm
not yet using it as my "daily" system. This was a
significant step for me as it is my first experience using
Slackware64. One of my Fvwm window decorations, an X which closes the
window, didn't look the same in the new Slackware. I fixed
it by re-making the vector data. I'm guessing that this was
needed due to a change in Fvwm. I made a gnuplot graphing menu (called "plotting") for
SAM. Gnuplot is already easy to use, but this made it
easier. My plotting menu has a sample gnuplot instruction
file in $level/data, and I plan to add more such files for
handy reference. The tool I made is called gogeeqie
and it takes two arguments:
| $1 | | data-file |
| $2 | | instruction-file |
As you might guess from the name, gogeeqie
makes an image of the graph using gnuplot and the supplied data
and instructions, then displays the image using
geeqie (an image viewing tool that comes with
Slackware). I added to SAM some changes that allow it to run on a 64
bit machine. SAM needs Linux, and I have not tested it
significantly with anything other than Slackware. However,
it wouldn't run (as expected) using Slackware64 15.0 without some
changes.
SAM uses three small C executables. I compiled them
using my new Slackware64 15.0, and I put all (three 32 bit and
three 64 bit) executables in a new place. Where the
executables previously lived I made three scripts with the same
names that the 32 bit executables had (and still have). I
put in these scripts code which determines whether the machine is
running a 32 or 64 bit operating system (using "uname -m") and
then invokes the corresponding 32 or 64 bit executable.
These changes were a success! I modified two SAM functions in the rsthread menu.
Doing this gave rsthread the ability to write datalines to files
called data.txt in each local dir that has a counter in the
corresponding remote dir. The result is that I now have a
way to capture the state of the counters, writing the data to
files that can be plotted using my (see above) gogeeqie
command. I worked on clarifying the text of Linux in the Finance
Office. I put the Slackware 15.0 data point on the graph.
February 2022
My webhost (which I will not name) took away their secure
shell service (ssh). I was using ssh together with rsync in
my rsthread menu to maintain this website. Therefore, I've
been diligently working to add ftp ability to rsthread. I
have it partly completed--enough to post this update.
I'll post again about this. I finished writing changes that allow the use of ftp or
ssh. I did some initial testing, and the resulting code
seems to work. Still I miss ssh, as some things are easier
done in a shell. I finished (for now) the task of adding PHP counters to
rsthread, then I used this new feature of rsthread to add
counters to this website. I added the ability, through a file added or removed by
lftp, to control whether the counters of a website are paused or
operational. This gives me a way to visit the site without
hitting the counters. When the counters are paused a
message saying "Counters are paused" appears above the counter on
each page. I applied the above to this website and I also
added a page For Testing Counters which I will
use to do do just that. I added another feature. Now I can upload and use a
file of IP Address exceptions. When a user visits one of my
sites the list of exceptions is referenced by the PHP code.
If the user's IP address in in the list, then then counters are
suspended for just that visit. I wrote this,
understandably, for my own use, as it is easier than
adding/removing files as I described above.
January 2022
I worked as a finance assistant from November 2018 thru
the end of December 2021. This job, being in the realm of
computers and numbers, was a good match for my abilities.
What I liked best about it, however, was that it gave me a rare
opportunity--a chance to apply my Linux skills and tools in a
business finance office.
I have decided to document this experience of applying Linux in the Finance Office as a
project, since that's what it was. If you click the link
and read about it, you will see that, for better or worse, I took
what would have been an ordinary job and made a project of
it. Hopefully, it was a worthy project.
I added floating captions to the picture and movie
thumbnails of this website. To read a caption, hover over a
thumbnail. Then I remade all the thumbmails to use this
feature. I fixed a problem that was caused by my use of "block" and
"inline-block" in my html for this website. Instead I now
use an additional, nested table. The problem was that the
pages would print differently than they appeared on the
screen. The additional table was added around the headings
in my content. These are normally colored. When I
removed the "block" and "inline-block" the colored portion of the
heading assumed the full line length. I solved the problem
using the addtional table, as I described. I fixed a problem with floating captions (see
above). Since they are made from the same files that I use
for the regular captions, some have html tags in them. Well, it
so happens that the title attribute that I used to make them,
doesn't interpret html tags.
I fixed this problem by writing and implementing (in
coolage) a tool that I call
begone. This tool outputs the first
argument less occurrences of <...> that it contains, thus
removing the html tags. (Script coolage is
one of the tools in rsthread. It makes the html for the
collages of pictures and movies.) I want to mention these exciting software tasks I've been
working on:
I'm currently working on adding PHP counters to
rsthread and my websites. I made a new SAM tool (in $sam_tool) called
wolf. It is a variation on another tool
in the same menu called dog. Like
dog, wolf is for displaying
"dog and pony shows" (a funny name for presentations).
Basically, it is a way of displaying html.
Now that I have a way to create html that can be browsed
without a webserver I was inspired to make
wolf for viewing this html. Since this
is similar to what dog does, I put it in the
same menu and gave it a related name. (Whereas
dog views text files with embedded html tags,
wolf is for complete websites installed
locally.) I also made a tool, noserver,
which helps me to use rsthread to make this
html. I can't remember ever having blogged about my backup
tools--the primary tool I call dd_back.
Making backups has long been a difficult task.
dd_back has been changing that. It
works well, and making backups is not so difficult now.
In recent weeks, however, I have made some
improvements. dd_back works by using
two well known tools that are not mine--ddrescue and rsync--as
needed to either make a disk image (when the image doesn't
exist) or to update the disk image (when the image does
exist). So much use of devices and images has made me
aware of the need for running e2fsck and/or dosfsck first
before running dd_back on a device. By
my improvements I have automated this extra work and
integrated it into the normal use of
dd_back. Hooha! About a month or two ago I wrote a short bit of code
called alarm that functions within an
existing SAM menu. (This menu I called "fmenu". I wrote
it to use at work, and it gave me convenient access to a
collection of tools that I used daily.) alarm
turned out to be surprisingly useful especially after I bought
a H800
Bluetooth Wireless Headset. So useful, in fact, that
I use it now around the house.
Let's say I'm baking bread, for example. I already
have my Linux box fired up and I'm enjoying a bit of Mark
Hurst's Techtonic radio
show (WFMU) on my new
headphones. I go to the kitchen and put my bread in the
oven, then return to my Linux box in my office and navigate to
fmenu. I enter, "alarm test" to verify
that the alarm is working. "alarm off"
silences the alarm. Then I enter "alarm
set 40". The alarm will beep in my headphones in 40
minutes. The available shell command "atq" lets me
review this scheduled alarm and any others. And all this
can be done without missing the radio show. Again,
hooha!
December 2021
I've been working on two projects. The first one is
a pair of SAM menus I
call t_card and t_period. Each is a collection of
executables that work together for two different, but related,
purposes. Both are for the purpose of recording and
reviewing my time worked on the job. t_card is for my daily
hours, and t_period uses the data from t_card to provide a
summary of my hours worked for the pay period. Both allow
logging and totaling of hours by task.
Both also allow for a comparison of hours actually worked to
hours reported for pay. This may seem like an odd thing to
do, but really it is not. It is not uncommon for an
employer to ask that the employee report hours rounded to the
nearest tenth of an hour or even the nearest quarter of an
hour. This will cause an accumulated error over time.
And there may be other reasons for an error of this kind.
With my tools I can keep a running total of the accumulated error
(or "delta"). Also, of course, having total hours worked by
task is also useful.
These tools are not new. I started using them in March
2020, so that is just three months short of two years ago.
What is new that I am working on now is a revision to both t_card
and t_period that changes the internal method of recording and
summing the hours. Whereas currently the tasks for which I
could record hours were hard coded, I will soon be able to add
and remove tasks to and from the set that the code processes
without changing the code. This will be a welcome
relief. I will be able to add and remove tasks easily,
instead of the nightmare-ish code changes that were needed.
I say "nightmare-ish", because it is true. I had worked
on an intial set of changes which allowed a sort of straddling
the fence. I used indirection to refer to variables (for
the tasks) on the fly, and I kept a table of the currently used
tasks to which the method referred. This was enormously
clever, but not an improvment. I abandoned that effort,
and I now am using a method by which each task is represented by
a file which stores the sum of the hours for that task. Yes
this is encumbered by using a file paradigm, but it works well
enough it seems, and simplified the code.
This change has been coded for t_card, but not yet for
t_period. The other project is more work on rsthread. I made
some modifications to two of the executables (a function called
mk_web and a script called harvest) and also a change to a
required file, _link, in the set of files from which the webpages
are built. Additionally file _target is no longer
needed.
The result is the optional ability to create a set of .html
files for the website which can be viewed without the use of a
web browser. I had a need to present information in
website form for my employer, but as the information was not
suitable for publishing on the Internet, I searched for another
method.
The changes I made were simple and don't interfere with the
original functioning of harvest. Best of all, the changes
seem to work!
I'm learning, in the process of this work, about the syntax
used by web browsers to interpret what you enter in the URL field
or type into the href string. It seems to be different for
Firefox and Chrome. I was able to coax Chrome into
correctly displaying the same website that I normally upload to a
webserver. This surprised me, and I wonder if this is a
reliable approach. Certainly it is a handicap to require
the user to use Chrome only.
With my new rsthread I can build an alternate set of .html
files that have all href strings like this: "file:///<drive
letter>:/<path>/<token>.html". This
approach seems to be reliable, but it requires that the hard
coded drive letter used in the website be mapped before the
website (perhaps on a flashdrive) is viewed by a
browser. I augmented crawl. It can now skip dirs "junk" which
I use to store previous versions of things--mostly
executables. And it can use files _dirlist instead of "ls
-l" to find the dirs. This allows me to put the menus in
arbitrary order instead of the order given by "ls -l".
November 2021
I updated three movies (two in my Adventures of Bendyman set and a
demo of a Fixed Thin
Client). I made the movies using my Canon camera, but
discovered that although the mp4 files that I made from them work
in Firefox (and other browsers?), they did not work on some (or
all?) cell phones. I have a tool which fixes them, as I've
been aware of this problem. What's new is that I used the
tool to fix the movies. I did some work on rsthread. I gave it (my tool
called sink) a limited use of recursive file
transfers. In particular it uses recursion for three dirs
(called "ready", "thumb", and "flash"). They are well
behaved dirs that I trust will not give rsync any trouble.
And the first two contain pictures and movies and so using
recursion for them will save time. The third one contains
the html which is used to display the pictures and movies and the
corresponding thumbnails. I used what I described above (the fixed movies and
sink) to make updates to this website. I made and posted a new Adventures of
Bendyman movie.
October 2021
I've been trying to decide where to go from here with the
development and use of rsthread. I toyed with some ideas
for additional features which would be generalizations of the
current capabilities. For example, I considered modifying
the method to save intermediate results when building the pages,
so that these intermediate results could be reused. This
seems do-able, but it makes the code complex.
I also considered adding the ability to have more than one
collage on the same page. I began to see perhaps how to do
this. I'm afraid it would add considerable complexity.
I decided that I don't need either of these right now.
It is odd how the pull toward complexity is strong. It had
a hold on me. I shook it off. I will keep the code
simple for now--it is elegant and usable.
I do need the ablity to add content that has full-size
pictures and movies--not in a collage--mixed with text.
That will be fun finding good solution to this problem. I
will report back.
I will also spend some time using rsthread to edit and augment
my websites (this site and Rosevear Software).
I've been using rsthread in its current form. It
works! Because of continued, sporadic trouble with rsync for
updating my websites, I embarked on a new project to make a website
file-transfer tool.
I had already a tool I call "crawl". I wrote it for use
by rsthread. It changes from directory to directory,
recursively, performing a task in each. I combined this
with some rsync tricks and put the code together as Bash scripts
and functions.
The result was a script called sink plus a
few other supporting tools, especially one called
trans which invokes sink via
crawl like this (below) to transfer a whole directory,
recursively, to the webhost:
Or it invokes sink directly to transfer just
the current directory:
Tools sink and trans are
brand, spanking new and have only been used a few times to update
a website. I have hopes that transferring files in this way
will prove to be reliable and useful.
Note that this method I have crafted uses rsync, as I did
before, but it uses it in a different way. Before I used
rsync in the classical way--I used it to mirror a directory from
my local staging directory to the corresponding directory on the
webhost, and I did this as a single, recursive, rsync
invocation. With my trans and
sink method I can still do a
recursive transfer of a whole directory, but multiple rsync
invocations are used, without recursion, to transfer the contents
of each directory one directory entry at a time. Then it
changes to the next directory and repeats. I added to sink the ability to remove
files and dirs on the host. I discovered how to use ssh to run a command on the remote
host from a local script. I used that in
sink to test whether the "to" dir for an rsync
invocation exists before attempting to run it.
September 2021
I updated my method for building websites. I now
call it "rsthread" instead of "rs_php". It is no longer
php-based, but it retains much of the orginal method. The
result was good--it runs faster and more reliably. I re-built this website using rsthread. In the process of re-building this website I may have
found a handy tip: In the past I have struggled with
rsync. It is a great tool, but it has been failing
sporadically when I use it to update my websites. Today,
after a series of failures, I decided to try running it directly
from the console command-line, and without the use my SAM-based
tool (called mirror). Well, that fixed it--this time.
I'll use this new method going forward, and report back.
Note that mirror is still helpful. I used it today for all
the steps but the final invocation. It already outputs the text
of the invocation, so it is easy to copy it to the command line
for execution. A few days later... The above theory doesn't seem to fit.
The fix that I found doesn't work consistently if at all.
I resorted to hacking out a piece of the target--in the area
where the problem occured--then using rsync again. That
seemed to work, although it still took a few uses of rsync to
finish without errors. Incidentally, I used ssh to connect
to the webhost, then "rm -R" to do the "hacking".
February 2021
I reached out. I found a great group on Facebook (facebook.com/groups/unixshell).
It is great, because it is full of smart people who write shell
code. Not exclusively bash, but that's good--maybe I'll
learn something.
After reviewing some of the posts in the group I was excited
to find that others also are dealing with the challenges of using
the shell to interface with spreadsheet data (see my January post
below about crimport). In fact, I soon had an opportunity
to tell the story of crimport and bob<n>.
The group is about the *nix shells. I learned that
posts about Python are not allowed. I will assume therefore
that posts about C code are likewise not allowed. I will,
however, share about it here.
Here is a link to the source for the current version, bob8.c. And here is a brief
description:
It's a simple C program. It reads from standard
input and writes to standard output--one character at a
time. For each character it determines by some simple
rules whether it is part of a new field or not and also
whether or not the original field was quoted. In this
way it removes the commas that are not in quoted fields, but
leaves the commas that are in quoted fields. There is a
little more to it, so please read the comments in the source
code. Also you can view file trans8.txt to learn more.
January 2021
Hello, hello, hello. How is everyone out there? How
are you all?
I'm just carrying on as I usually do. Not much ukulele
playing though. Instead I have been immersed in Bash
programming. Bash-shell programming in Linux is really
amazing. How can I tell you? I am on a secret
journey, it seems. Sometimes I reach out for others who
might understand--I know you are out there--but rarely do I get a
response.
I have been writing sets of tools which perform tasks for my
job--I work in a finance office. My latest tool-set I call
crimport. This name applies to a collection of bash scripts and
functions, plus one C executable. Here are the names and
types of the tools in the set:
byfunc, Bash script bob6.exe, C executable runchimp, Bash script runcrimp, Bash script crimport, Bash function crfields, Bash function
The script runchimp is the main tool, as it runs the other
tools. Or you can use runcrimp, but the first one, which
came from the second, does a better job. These tools work
together to make a journal entry import file for our finance
software, called ACS OnDemand.
The import file is made from a csv file which comes from an
Excel (spreadsheet) file. The spreadsheet contains staff
credit card data. One of my work tasks is to collect data
(and receipts) from staff members regarding their credit card use
so that I may make the monthly journal entry which reflects the
expenses in our general ledger.
Collecting the data and transferring it to the spreadsheet is
no small part of the task. I have tool-sets for that also,
and I would like to share about them as well--later. This,
the current topic I'm sharing, is about translating the
spreadsheet data into a journal entry import file.
Script byfunc and C executable bob6.exe were breakthroughs
that made this happen. And I use a tool that is in my
"handy" tool collection called dribble--all within the framework
of a menuing system (of my own creation) called SAM.
(231028 Here is a link to SAM.)
Here is an excerpt from file crimport.sam which defines
function crimport:
bob6.exe and dribble (and byfunc) are made available for
execution by use of invocations of SAM tools within the scripts
and functions.
The above code sends the spreadsheet csv data (in file
$longsprd) through three filters: skip
just skips the number of lines named to get past the title lines
and the headings. bob6.exe re-writes the
csv data in space separated values format with double-quotes
where needed to preserve the empty fields, which would be lost
otherwise. And dribble, together with a
built-in shell tool called eval, relocates the incoming fields of
the stream putting them after crfields where they function as
shell parameters (also called arguments). And that's where
the magic happens.
This, I'm proud to say "clever", bit of code allows me to
process the fields of the spreadsheet as bash arguments--the
rest, with the exception of byfunc, was easy
Script byfunc should have been easy to write. In the
end, it was simple. There's not much to it. It is a
single script that takes various arguments. For
example:
sums $84.23 into a file called sum001 where the current sum
for ledger-function 001 is stored. File sum001 is made and
a value of 0 is put in it, if it doesn't already exist.
After the above command it will contain 84.23 more that it did
previously.
outputs the value which is in file sum001, and
lists all the ledger-functions represented by the sum
files.
Add to this
which removes all the sum files, and you have a complete
tool. With this I was able to solve the problem of
generating the balancing transactions. These are the
transactions against the asset accounts which balance (by
ledger-function) the sums of the transactions against the expense
accounts.
Did I say that right? Anyway, whiz bang, it works, and
I'm giddy. I apologize for my self-congratulation, but as I
said, this is a secret journey!
August 2020
Yes, if you asked me, I would surely say that life during
the current COVID-19 pandemic is difficult. For me the
biggest hardship has been the disruption to my routine.
I'm naturally routine-challenged to begin with--meaning I'm prone
to keeping an erratic schedule, so I clearly do not need the
disruption.
Perhaps some good has come of it, however. I've found
some time for things I wouldn't normally do. Good things
like building relationships (Facebook) and singing (accompanied
by myself on ukulele).
So it is my singing that I was working toward in this
post--could you tell I was edging in that direction? My
croaking voice. And my wavering ukulele playing Click
here
for a sample.
July 2020
We (my hard working wife and I) filed our taxes.
Whew. That's over, but I want to tell a story about the
California form 540-2ez, and forgive me if I rant a little.
I'm getting suspicious. The California form, for
several years, has been less than satisfactory. The federal
form 1040 is fill-able, and you can save it after filling it
out. Thank you. The California form on the other
hand, like previous years, is fill-able, but it claims that you
cannot save your entries. What good is that?
Actually it seems that you can save them (I was using Evince
pdf viewer), but there is another more serious problem--some of
the fields on the form would not let me make an entry. That
is not good. What is going on? Is it that hard that
the state of California cannot make a tax form that works, let
alone one that lets you save your entries? I wonder if
there was money involved. Was this done to boost sales of
Turbotax?
So I found a way to do it using my existing Slackware Linux
installation. In short, I did it in these steps:
I converted the form to a pdf form that doesn't have
the extra buttons on it. Again I am suspicious.
The buttons "Print", "Check for errors" and "Reset" might be
useful, but I wonder if they are there simply to deface the
form and spoil it for use by the method that I discovered.
Spoiled it was, but I found a work around which allowed me
to convert the pdf to one without the buttons. I used
the Print button to print the form, but I selected a special
printer that I installed a few years ago after reading a very
interesting Internet article. The printer is called
"/tmp/whatever". That is a story in itself, which I will not
tell now, but when you print to this printer it looks for file
/tmp/whatever--you need to make the file first. If it exists,
and is writable, then CUPS (the printer software) writes to
/tmp/whatever a pdf file which if printed would be exactly
what was requested to be printed--the blank tax form. So
after doing this I had a pdf of the form without the
buttons! Next I imported this pdf file into Gimp (which comes with
Slackware). Gimp is like PhotoShop, only free. I
imported the pdf page by page each time exporting the page to
a different pnm file. This format, pnm, is like jpg,
except that is is not, I think, compressed. It is an
image file. Then, equipped with a pnm file for each of the pages of
the state form, I proceeded to fill the blanks by writing to
the image files, one at a time, using Kolourpaint. I
found that the Deja-Vu 12pt font worked great, except for the
SSN and maybe one other field which needed 10 pt. I
saved each filled-out page as a png file (not pnm), because it
was suitable, and it seemed that Kolourpaint did not offer
pnm.
This wasn't particularly hard. OK, it was a little
hard, but it was such a relief and a pleasure to have a method
that worked and did what I needed. Thank you Linus
Torvalds, Patrick Volkerding, and thank you to the person who
wrote the article about printer /tmp/whatever. Thank you
very much!
Just to clarify: This method that I described does not
give you a fill-able form. Instead it gives you a way to
write on the form, which is almost as good. You need to use
care to align your text in the fields. This method gives
almost the same result as a fill-able form, and it has the
benefit of allowing you to save the filled-out form. Hello. This will be a blog-style post, as I'm
feeling a little blog-ish.
Hosting a website can be rewarding, but I've noticed two
trends that are, frankly, discouraging.
Firstly, this website doesn't get as many visits as it did
once. Surely, I'm partly to blame for this. I don't
post updates regularly, and I don't try very hard to make my
posts interesting and relevant. I admit it: This
website is primarily self-serving--I like to talk about
myself. Yet, in talking about myself, or by posting links
about things of interest to me, I have in the past I think
sparked a little interest. Sadly, that "little interest"
seems to be fading.
Secondly, I've been receiving an increasing number of requests
to post links on this website to dumb stuff. And I'm being
nice when I say that. At first, just to see what would
happen, I played along. I put the links in a separate
place, my Gratis Links, to let the reader
know that I had some reservations, and I posted links to mostly
whatever came my way. Well, I'm starting to get
annoyed. Apparently I'm not alone in this. Here is a
link to a sympathetic article:
New Mouth Inc. The most stupid series of emails J4MB has ever
received?.
OK, I do use some judgment when selecting the requests
to which I will respond. I will try a little harder,
though--maybe we can do this together. Your part is to send
requests to me to link to good stuff. Stuff that
matters. Stuff that is not so obviously an effort to fleece
the public and take advantage of the gullible. I will, in
turn, try to keep this site interesting, relevant and
current.
Deal?
April 2020
I made a Patreon project called Joe's Life. It is
about my projects, links, photos, and more, and it
references this website.
Will you be my patron?
March 2020
I wrote some software that I use at work. Every
month I send about eighteen emails each to a different address,
some with a Cc, and each with one or two unique attachments.
Previously, I did it the hard way--manually, one at a time
using Office365. That worked, but it wasn't long before I
was wishing for a better way. After some searching, I found
an example Python script which sent emails using Starttls
encryption.
I understood Starttls to be the encryption method I needed to
use for Office365. You see, I needed more than just to send
emails. I needed to send them from the same server from
which I had previously sent them by hand. That way I could
put my work email address in the From header and the recipients
could reply in the usual way. The emails would appear to
have been sent normally.
The Python script I found was great. Its only lack was
not having a way to send attachments. I searched some more
and found lots of rather confusing scripts, then finally a simple
one that I could reproduce and harness for my needs.
But it didn't work.
Stymied, I remembered something I had read in a post
on StackOverflow.com. Someone had commented about a
typo seen on the Internet--"content-decomposition" instead of
"content-disposition". I checked my code and discovered
that was the problem. I felt a little foolish for not
having noticed. I fixed it, and bingo--I could send an
email with an attachment.
But I wasn't done. I added to the script some
modifications which enabled:
Adding a Cc when needed. Adding one or two attachments when needed. Giving server name, port, password, user name, and from
address by running a user-provided function which assigns the
above information to environment variables referenced by my
Python script. Invoking the script from the command line with six
arguments: <to address> <cc address or "-">
<file containing the subject> <file containing the
content> <optional first attachment or "-">
<optional second attachment or "-">
I also canonized it by adding it to my SAM menu
system. This made it easy to use and document and also
paved the way for adding some helper scripts. (231029 I
updated the preceeding link.)
I call the Python script dupe, and you can
get it here
in a SAM menu called handy which includes helper
scripts. (231029 I updated the preceeding link and this
paragraph.)
December 2019
November 2019
Although my Fix Thin Clients
project has wound down, I have continued development of the
method on which it was based. And I have renamed it.
Instead of SuperDrive I now call my method ZombieSlack.
Slack because it is an embodiment of Slackware. Zombie
because it is not a live system and it is also not a
traditional non-live system. Being neither live nor
dead there is only one word for it--zombie.
OK I'm playing word games I'll admit, but there is another
reason for the zombie moniker--it does what a zombie does--it
takes posession of the host and re-animates it. It does
this at different levels.
It runs in the same form whether installed to a flash drive,
hard drive, or solid state drive. Also it matters not whether
the drive is internal or external. Whatever the drive, when you
boot the machine from the drive ZombieSlack takes over the
machine--much like how a zombie takes over the host body.
This is zombie behavior level zero. Zero because the host
possesion is temporary. Whatever was installed to the host's
drives remains completely unchanged and usable as before.
When ZombieSlack is installed to a drive I call it a
ZombieDrive. A ZombieDrive is an interesting thing. As I
explained above, you can use it to boot a machine (it is
bootable). You can also use it to wipe and install
to a host drive. The wiping is done by invoking function
wipe which is in the SAM menu
at /RS/opt/h_tool, and the installing is done by
running script /boot/tool/uptask-stc. A
ZombieDrive thus has the ability to pass its zombie-ness to other
drives. Essentially, when it bites another drive the drive
becomes a ZombieDrive. Level one.
Level two answers an important question, "If zombies come from
zombies, then where did the first zombie come from?" My
ZombieSlack is an embodiment of Slackware. My first
ZombieSlack installation was on a hard drive in a computer. I
made it by tweaking a stock Slackware 14.1 installation.
The changes were made by hand by careful design. After I
got it working I then wrote the script which installs ZombieSlack
to another drive. Finally, as a last step, I
reverse-engineered a method by which I could apply the changes
(ZombieSeed) to a new Slackware installation, thus making it into
ZombieSlack, and I incorporated that method into the install
script. The install script could thus insert ZombieSeed
into a Slackware installation (making a ZombieDrive) or install
from one ZombieDrive to another drive (also making a
ZombieDrive).
What I did was elementary--I added some scripts and dirs to
Slackware, and I modified some existing scripts and config
files. Elementary, yet radical. By making a few
changes I made the default user's home directory contents (I call
this file space) changeable, yet self-restoring. In
this way I was able to fit a full Slackware 14.1 installation on
a 16GB USB 3.0 flashdrive--without performance issues. And
I made a system that was immune to breakage by user error.
I tested the new script that I made by doing something that I
seriously needed. I made a Slackware 14.2 based ZombieDrive. I
called it ZombieSlack/Slackware-14.2. That was a
breakthrough. I might have been able to upgrade from Slackware
14.1 to 14.2 without this ability, but with it I upgraded in a
way that kept my modifications neatly separated, conceptually,
from Slackware.
I am looking for new projects--new opportunities to apply this
method.
August 2019
Here's a little story about my Latitude E5400 laptop and
Slackware Linux. (Windows users, I think this may interest you
also.)
Some time ago my laptop started giving me a message that went
something like "The AC power adapter wattage and type cannot be
determined..." Then it would refuse to charge the battery. At
first it was intermittent. I would disconnect and reconnect the
power, and it would be OK. But today not.
I was faced with the possiblity of being without my laptop
which I use for work. Alternate ways of getting at my data were
turning over in my mind. Hmmm.
A little Internet research and some twiddling with my laptop,
battery, and adapter led me to conclude that it was the fault of
my AC power adapter, so I started an Internet search for a
replacement. I found a Targus 90W Universal laptop charger at
Best Buy. I paid for it online and it was ready for pickup
within an hour.
When I picked it up at the store I was immediately comforted
by the nice box it came in. Stupid huh? Anwyay, when I got it
home and un-boxed it, it seemed to work fine. Whew. That was a
relief. Later I realized that I gained more than the continued
use of my laptop. I had learned a neat trick in the process.
Earlier, while investigating the problem, I had re-discovered
the built-in function which turns the battery charger on and off.
You operate it by holding the blue Fn key, followed by pressing
the F2 key. (I'll call this "Fn+F2".) I had used this before,
but today I found a neat trick I could do with it.
I had read that Lithium batteries, such as the one in my
laptop, would have a longer life if stored with a 40% charge.
That always left me wondering, though, how it could be done. The
normal function of the laptop charger brings the battery up to
100% charge before stopping. The unanswered question was how to
charge to 40%, then keep it there.
One way of course would be to charge to 100%, then, unplug the
charger and monitor the dis-charge with a tool like gkrellm
or upower (use "upower
-d"). When the battery reaches 40%, shutdown, remove the
battery, and store it. Sorry, but that is a really lame
solution.
Enter the neat trick. Do the above, but let the battery
discharge to about 30%. Then reconnect the power and watch the
charge. When it reaches 40% stop the charge by pressing Fn+F2.
I don't know exactly what this does, but it seems to allow the
adapter to power the computer without charging the battery. It
seems to effectivley halt the charge, thus keeping it, at least
for a while, at a constant level.
Actually when I did the above the charge crept up to 41%
somehow after I pressed Fn+F2. No problem. What I wonder is
will the charger maintain the charge level, or just leave the
charge level alone and thus allow for the natural self-discharge
to gradually bring the level down? Of course, there is also the
self-discharge that will happen while the laptop is off and not
in use.
Well, I still think it is a neat trick. I learned a little more about use of "Fn+F2". It doesn't
actually turn off the battery charger. Instead it toggles the
charge rate between the normal rate (which may be slow or fast
depending on the BIOS setting) and a very low rate. I'm guessing
that this very low rate is intended as a way to maintain the
charge at the current level. Just enough to balance the the
self-discharge.
So maybe it isn't so lame to remove the battery. One might
wonder what harm is done to the battery by maintaining it for
long periods on a very low rate of charge? Perhaps do the steps
I described to adjust the charge to 40% and keep it there, then
remove the battery at the next regularly occuring shutdown.
Maybe store it in a resealable bag with a note stating the charge
level and the date.
December 2018
I found an interesting service called Peertopia. I might someday
want to sell Rosevear
Software products on this site. I started to read the
details and I came to a section called "Limitation of
Liability". I thought, "Great, not this again!" The
text was in all caps and nearly impossible to read. I
re-wrote it using
just as I did before (see September 2018). Here is the
result:
You acknowledge and agree that you assume full
responsibility for your use of the site.
You acknowledge and agree that any information you send or
receive during your use of the site may not be secure and may
be intercepted by unauthorized parties.
You acknowledge and agree that your use of the site is at
your own risk.
You acknowledge and agree that, to the fullest extent
permitted by applicable law, neither peertopia nor its
affiliates, suppliers, or third-party content providers will
be liable for any direct, indirect, punitive, exemplary,
incidental, special, consequential, or other damages arising
out of or in any way related to the site, or any other site
you access through a link from this site or from any actions
we take or fail to take as a result of communications you send
to us, or the delay or inability to use the site, or for any
information, products, or services advertised in or obtained
through the site, peertopia's removal or deletion of any
materials submitted or posted on its site, or otherwise
arising out of the use of the site, whether based on contract,
tort, strict liability, or otherwise, even if peertopia, its
affiliates, or any of its suppliers has been advised of the
possibility of damages.
This disclaimer applies, without limitation, to any damages
or injury arising from any failure of performance, error,
omission, interruption, deletion, defects, delay in operation
or transmission, computer viruses, file corruption,
communication-line failure, network or system outage, your
loss of profits, or theft, destruction, unauthorized access
to, alteration of, loss or use of any record or data, and any
other tangible or intangible loss.
You specifically acknowledge and agree that peertopia shall
not be liable for any defamatory, offensive, or illegal
conduct of any user of the site.
Your sole and exclusive remedy for any of the above claims
or any dispute with peertopia is to discontinue your use of
the site.
In no event shall either party be liable to the other for
exemplary or punitive damages.
Content is also provided by sellers on the site.
Please note that sellers may post content that is
inaccurate, misleading, or deceptive.
We neither endorse nor are responsible for any opinion,
advice, information, or statements made by sellers.
We will not be liable for any loss or damage caused by your
reliance on such information or materials.
The opinions expressed by sellers and third parties reflect
solely the opinions of the individuals who submitted such
opinions and may not reflect our opinions.
I don't intend to discredit the site. I think it is a
great site offering a valuable service. I just wanted to
translate the text that was hard to read as a help to you.
Don't you agree it is better to use text that is
readable? Why hide your words by making them hard to
read? I think it is poor business to use such
tactics.
October 2018
September 2018
The United States Postal Service (USPS) has enticed me to
try their new Informed
Delivery Service. As part of the signup process I am
asked to agree to their Terms and Conditions. These contain a
horribly hard to read paragraph in all caps.
Perhaps they didn't want me to actually read this
paragraph.
Rather than concede to their rude tactics I piped it though
Linux tr command
tr '[:upper:]' '[:lower:]'
to help me re-write the paragraph. The re-write also
included:
Capitalizing the first letter of each
sentence. Correcting two spelling errors. Correcting one grammar error. Adding one line of white space between each sentence to
improve readability.
Here is the result:
By accepting and agreeing to this agreement,
you consent to and accept the risk in receiving information
via email and/or text messaging through the service.
Such electronic communications may provide the above
mentioned services, but may also contain marketing messages as
well as requests for feedback from you regarding the
service.
You may withdraw such consent at any time, but if
withdrawing such consent, some functionality of the service
may not be available to you.
You acknowledge and agree that your telecommunications
carrier may charge data usage fees (including additional
charges when roaming as well as fees for sms or text messaging
that may be used to receive updates related to the tracking of
mailable items) to access the service, and that the postal
service shall not be liable for any such fees as a result of
your use of the application or the service.
Additional fees or costs may accompany your receipt of
emails sent through the service. You should contact your
wireless or telecommunications provider for complete pricing
details.
Now that I can read it, I wonder why they tried to hide
it. Perhaps it was not intentional after all, just poor
composition. /li>
July 2018
I have an HP Deskjet D1560 printer that I want to use in
Slackware 14.2 (Linux). It worked fine in Slackware
14.1. In 14.2 I get filter errors. I traced the
problem by looking in /var/log/cups/error_log. It said that
there was a problem with the perl script:
I Googled for help and I found a post in LinuxQuestions.org describing
a similar problem (different printer) also going from Slackware 14.1 to
14.2:
I also learned that the syntax used in the perl script is deprecated
and destined to fail in future versions of perl:
Well, it's failing.
This is bound to affect more people than the two I've counted so
far.
Are you listening, HP? Well, I fixed it. I'm not sure where the blame
lies. There is probably more to this than I
understand. Anyway, I fixed it by making some changes in
/usr/lib/cups/filter:
I should explain, however, that the bug report was not about my problem. It was about a very different
printing problem, but, lucky for me, it had a similar
solution. Patrick Volkerding of Slackware.com needs
money. See this article.
Donate here.
See also my links at joeslife.org/links/computing/slackware
April 2017
This fascinating fractal-zoom of the Mandelbrot set may
cause eye strain or headaches. You have been warned:
I encountered a new word, "performant".
Animadversor at english.stackexchange.com
had this to say about it:
There's not really anything inherently wrong with
performant; its formation is regular enough and it seems to
convey a meaning that no other single word conveys. If
that is enough for you, then go ahead. But you ought to
know that the word will be ill-regarded by many, who will
consider it a pseudo-learned, affected, vulgar and pointless
novelty. Of course, perhaps they are wrong about this,
but even if so, they may well be people whose good opinion of
your language is important to you.
answered Mar 31 '13 at 6:52
Animadversor
April 2016
I made a new project, Fix Thin Clients.
I took my picture off of the pages and replaced the old
banner with a simple box. The reason was twofold: One, I'm
tired of having my picture on the pages, and, two, the new banner
is stretchable. This allows the text line length with to be
adjusted. Sorry it has taken me so long to make this
change. Before it may have been hard to read the long lines.
This is the best news. I've been a radio controlled
model airplane builder and flyer for nearly forty years.
Now I'm an aviator!
"Make no mistake: unmanned aircraft enthusiast are
aviators, and with that title comes a great deal of
responsibility," said U.S. Transportation Secretary Anthony
Foxx. "Registration gives us an opportunity to work with these
users to operate their unmanned aircraft safely. I'm excited
to welcome these new aviators into the culture of safety and
responsibility that defines American innovation."
January 2016
What happens to a dream deferred?
Does it dry up
Like a raisin in the sun?
Or fester like a sore --
And then run?
Does it stink like rotten meat?
Or crust and sugar over --
Like a syrupy sweet?
Maybe it just sags
Like a heavy load.
Or does it explode?
December 2015
Click here to listen
on YouTube:
Run Rudolph Run
Out of all the reindeer you know you're the mastermind
Run, run Rudolph, Randalph's not too far behind
Run, run Rudolph, Santa's got to make it to town
Randalph he can hurry, he can take the freeway down
And away went Rudolph a whizzing like a merry-go-round
Said Santa to a boy child "What have you been longing for?"
"All I want for Christmas is a Rock and Roll electric guitar"
And away went Rudolph a whizzing like a shooting star
Run, run Rudolph, Santa's got to make it to town
Can't you make him hurry, tell him he can take the freeway down
And away went Rudolph a whizzing like a merry-go-round
Said Santa to a girl child "What would you like most to get?"
"I want a little baby doll that can cry, scream and wet"
And away went Rudolph a whizzing like a Saber jet
Songwriters
JOHNNY MARKS, MARVIN BRODIE
Published by
Lyrics © ST. NICHOLAS MUSIC INC.
Christmassy, huh? Many artists recorded this song, but I like
Chuck Berry's the best. I found these lyrics at www.metrolyrics.com.
October 2015
March 2015
- There once was a barber named Ware
Who was sadly allergic to hair.
When customers called,
Unless they were bald,
He would sneeze them right out of the chair.
--- Ogden Nash
January 2013
August 2012
August 2011
I put a few new links in my computing
links page. If you want to see them, search on "sed", and
you will get the old and new links about "sed". I upgraded to Slackware 13.37. (Here's a link to the Slackware
website.) I've been building a little, on my Curtis Robin. I've even been flying my Wanderer occasionally.
March 2011
I've been staying busy, but not by working on this
website! I've posted a few new links and a few new
pictures. Here's what I have been doing:
I integrated a Linux box into our home entertainment
area. Now we can watch TV, internet movies, or
DVDs. I upgraded my rsync-based backup system with some
software improvements plus two new, one-terabyte hard drives
in USB enclosures. I migrated my daily computing to Slackware 13.1.
I've had it on my box for a long time, but had to work out
some details. Another detail was my e-mail software. I had been
using kmail, but it doesn't work well in Slackware 13.1.
So I migrated to Thunderbird (which works great!). I bought an MP3 player and learned how to use it.
I wrote some tools to manage putting songs on it. I downloaded and installed the new Adobe Reader
9. Ah, progress. New data formats require new
software. The good news is that it was free! I made some improvements to my Fvwm2 desktop.
Mainly, I made a bunch of new MiniButtons to make it easier to
invoke my commonly used apps. I discovered that "links -g" works great in my new
Slackware 13.1. Is this because I'm using a different
video driver? I learned how to better use USB sound devices, and I
managed to get my old PPA 6channel USB sound adapter to work
(after about three years of trying).
July 2010
December 2008
I've made a little progress on this website. I've been working on a website that lives on my camera's
flash memory stick. That may sound a bit odd. It is
another of my humble inventions. It helps me to organize,
view and share the pictures that I take.
November 2008
October 2008
|
|