• 2023
  • Jun
  • 3

Dayton 2023: Stuff I drug home.

I made good on my promise to not drag home too much junk this year, but a few boat anchors did manage to make their way in my car. Both of them had an immediate use, but the one that was going to go in to the rack didn’t work. Not a big deal, but still…it gives me a change to re-engineer some stuff. The rest of the items were parts for later use, whenever the need arises.

d231_stuff.jpg

Just some random bits I picked up. An old LED display (made in the USA!) from Rat Shack, and a 1971 Allied/Radio Shack catalog. The catalog is fascinating, the prices that some of the things cost was incredible, and it’s interesting seeing some of the Shack’s staples like the Weather Cube. It’s fun to look through it and see things I wish I could still buy.

d232_parts.jpg

A vendor was selling older Jim-Pak parts, as well as a handful of things from other vendors. Some were $1 per, others were 2 or 4 per $1. I picked a selection of parts, as well as some of those oddball fuses old meters used for their current sections.

d233_audio.jpg

This is a unit that was meant to monitor the audio feeds in a studio. At it’s heart, it’s two “channels” with a center channel. Unfortunately, it didn’t work. Examination reveals a lot of white corrosion on the input board, and it looks like some water was spilled and lay on the board for a long period of time. That’s not really a big deal, I’m thinking a couple of LM386 ICs can fix this guy right up. I’m not interested in studio accuracy, I want something for my rack that I can connect to the plugboard for audio testing.

d234_meter.jpg

According to the vendor, this works. Well…it didn’t. Mostly. I picked it up for parts since many of these are being torn down for the displays so they can become clocks. :( The device itself had a hard life, evidenced by the replacement screws in the bottom and the considerable cosmetic damage to the unit itself. It had all of it’s feet and bail, so that’s quite nice.

An examination on the inside revealed many repairs, some boards that had left their slots, and a missing fuse. Bringing the device up on power revealed that it indeed did not work, the master clock was not present so the device wouldn’t sample. One of the tubes was also partially bad, but it had the incredibly rare polarity indicating tube, probably worth the price of the device itself.

While cleaning the card edges I found that the clock card had a transistor leg lifted, so someone was troubleshooting the device when it was put aside. As this was supposed to be a parts unit, I’m not sure how far I want to go fixing it - if at all - but it will be the subject of a future post detailing the examination.

d235_cb.jpg

This is one of Radio Shack’s oddball devices from the CB radio era. It’s a crystal-controlled device that listens to Channel 9, which was commonly used as the emergency channel It was designed for antenna passthrough, and has it’s own speaker. If I can find a schematic for this, I may see if I can re-tune it to Channel 19, since there’s still a little activity there.

d236_cable.jpg

A friend of mine said he wanted a 2-pin cable for a battery box project he made. Said he needed an extension, so I found this. It’s not going to get warm anytime soon.

d237_buckle.jpg

This was a CB-era piece in the bins with the Jim-Pack parts. I don’t know why I got this other than it’s funny. Rip-off characters, anyone?

The slim haul wasn’t because I didn’t see things I wanted. I did. A lot of it. There were a couple of pieces that I saw that I wanted to bring home, but didn’t have a need for - a WWV receiver and some other devices among them. I figured if they were there Sunday, I’d bargain, but they weren’t. That’s me being told that I don’t need them, leave them where you found them.

(But you know very well if i see that receiver again I may pick it up because I’m being told it’s still here so you know you want it!)

Overall, I think I got some good stuff. The meter may make an interesting project, should I decide to repair it, but I think it’s just parts with a good plug-in for my other devices.

Next show is Butler, PA, and then nothing for a few month. Stay tuned for some show posts from that one, and some detail posts on a few of the devices I’ve picked up over the last year.

  • 2023
  • Apr
  • 23

Returning a Kaypro 2 to it’s original owner… but,

I’ve had this Kaypro 2 Luggable Computer since about 1993, I believe. I purchased it from a fellow college student with the disclaimer that if I ever needed to get rid of it, he would like it back. Since it’s been 30 years, I figure it’s time for it to go back home since it’s not doing anything here at the pygg pen.

kaypro2.jpg

A little history…

Kaypro, aka Non-Linear Systems, is credited with manufacturing one of the first widely available DVMs in the world. Using a relay driven wheatstone bridge, it was a concept that would have been difficult earlier, but transistors made it easy enough (if expensive) to make. NLS continued the test equipment line, making voltmeters, panel meters, scopes, and other items. Non-Linear Systems still exists as an entity, making panel voltmeters as usual. Find them here https://nonlinearsystems.com/.

In the 80s, however, NLS decided to get into the nascent home computing market with their entry, the Kaycomp. Named after the founder Alan Kay, these Ferguson Big Board clones rapidly gained a following due to their simple robustness and the fact that you received an entire library of software with them - OS (CP/M), Word processing, spreadsheet, programming, games - the machine was ready to go out of the box. Businesses loved them, and they were reasonable enough that home users bought them. Kaypro went on to make several offerings including hard drive models, and even a PC based model on the luggable form factor.

Kaypro eventually succumbed to that which kills all PC makers. PCs became commodity devices, and their IBM-PC based offerings weren’t any different (but they did manage to get a 286 machine out the door before IBM released the AT!) - and they released some machines that had crippling issues. The 90s saw bankruptcy and closure of the computer manufacturing, with test equipment continuing.

The Kaypro name has been resurrected a few times over the years, but has never stuck, and has faded to legends of a world before the IBM compatible devices took over.

And now, the main(board) feature.

This machine suffered a failure some years ago where the 8048 keyboard decoder went bad, got hot, and destroyed the relatively narrow +5V supply line. I, um, fixed it (yeah) years ago, but did a terrible job. The first order of business was to fix what younger me did. That was as easy as removing the wire fix, cleaning everything up, and re-placing it with some hot glue to keep things in place. Easy enough, and that’s done.

mainboard.jpg

(Just as a note, the 8048 was a common microcontroller used as a keyboard decoder, and often had a mask-program ROM onboard for the decoder portion. Replacing it was impossible unless you had a pull from another machine handy.)

Next thing was to give things a good cleaning. The connectors that I could access got a squirt of DeOxit for metal-to-metal contacts, and the scratchy brightness control got a squirt of DeOxit for controls. All good, and the brightness pot now feels nice and smooth. A dust removal and the inside looks good.

Of note here is the mainboard ROM is populated with a Micro Cornucopia Pro-Monitor ROM. Micro C, as it was fondly called, was a magazine/tech house dedicated to the Big Board machines, and later it’s clones like the Kaypro and Xerox devices. Micro Cornucopia manufactured several enhanced ROMs for these machines, and this one is no different. I believe this one allows the use of DSDD drives (instead of SSDD,) as well as provides the ability for 4 floppies. While it looks like I or someone may have modified it to do such, I can’t remember if this had the ROM onboard when I got it, or if I put it there from one of my other machines. It will simply remain lost to the mysteries of time.

Also of note: There’s a keyboard in here. I do not remember putting that in, but again, it must have come from one of the many junkers I purchased over the years. Another mystery.

But now, the but…

A quick inspection didn’t show any noticeable physical damage.

powersupply.jpg

A few quick checks didn’t reveal any shorts or other electrical problems. So, plug into a fused outlet and…

drives.jpg

The Kaypro “Grrrrrrk!” of the drives spinning up greets me. All LEDs are lit. The drives are running. That’s good - the drives on this spin constantly and load the head on and off the disk during access. It’s a way of getting quicker access, but you’re spinning the disks all the time.

However, there’s a problem. Not surprising after sitting for 20+ years.

screen.jpg

I know it’s hard to see, and I apologize. The screen has multiple issues. The first, that of the characters are partially formed. The second, and the problem which prevented me from troubleshooting the first is the screen is dim and goes dark.

Since the brightness is controlled by an analog pot, this is probably indicative of something in the power supply failing. I turned it off immediately so as not to destroy potentially unobtainable parts.

We can theorize about the malformed characters without power.

Kaypro2_Video_Small.jpg

For a bigger copy of this, click here: https://files.catbox.moe/pgldav.png

Video is generated with some RAM, a character EPROM, and a bunch of TTL glue. There are several things that strike me as potentially being bad here. Those are (but, not limited to of course!):

  • A bad character EPROM. These can fade over the years.
  • Bad RAM.
  • A bad TTL part.
  • One of the socketed chips just needs pulled and reseated.
  • Data from the monitor EPROM is bad, for the same reason as above.
  • An interconnect problem between the mainboard and CRT.

To be fair, this would probably be easy enough to track down with a logic probe or scope, but I’m unwilling to run it for any length of time because of the screen going black. I don’t have any CRT device troubleshooting tools left, and I don’t want to destroy the device to a point where I can’t get a part.

I’m not sure how the original owner wants to use if (if at all,) so it stays in the state it’s in.

  • 2023
  • Mar
  • 19

Random board shot - a commercial weather radio.

It’s always interesting to open a device and see ancient chips you’ve never run across before. The LM389 is a LM386 with three transistors for RF, IF, and 2nd IF on the substrate. This ancient commercial weather radio uses one for it’s audio final, if I want to keep it in operation it’s probably best to find a couple of spares along with the MC3357 IF chip next door. This device was made in the USA, even the circuit board was made right here in Ohio as evidenced by the GE Textolite “T” in the substrate of the board.

It’s well built, but the electrolytic capacitors with the long leads surprise me. I would think for a device with RF floating around inside, you’d want to keep those short.

It needs a new power cord, the old one is short and worn, as well as some DeOxit sprayed in the controls and switches. I may also replace the power supply filter capacitor, but that’s only because it was laying against the regulator for untold years and has a nice discolored spot on it.

More on this device later…

comwrad.jpg

  • 2023
  • Mar
  • 12

A General Electric T-125A “All American Five” Radio, Part III

This is part 3, and the final part of working with this radio. It didn’t turn out as good as I’d hoped, some of the other capacitors are leaky - and some of those are in a couplet. The customer just wanted it to make voices come from the air, so I’ve left it at that for now.

The original diagnosis I came up with was bad filters, and that was the case. The original electrolytic capacitors were bad, and were simply acting as opens instead of capacitors. The device in questions was a dual section, wax sealed tubular device, consisting of a 50uF and a 30uF part. I chose to replace them with Cornell-Dublier 47uF capacitors, which should be good for another 50 years. The across-the-line capacitor also got replaced with a Vishay 0.047uF @ 600VDC device. Just as good as the original without the bumblebomb problems.

The parts, fresh from Mouser:

parts.jpg

What we’re working with. The black tubular part in front, and the metal cased tube in back are the suspects. If you look in the middle of the board, you’ll see a reddish-brown rectangle That’s the couplet, and was a cost-saving multi-part unit. When they go bad you rebuild what’s in them…

what.jpg

Other problems reared their head as it was being taken apart. The biggest is the age of the plastics. The knob used a metal band on the plastic to tension it on the shaft of the tuning capacitor. Pulling this knob from the shaft left the tension band on the shaft (it had little ears that were probably used during manufacture as an install aid,) and the plastic is so brittle and dried out that it just crumbles to dust. There was enough left to put it back on, so I broke the ears on the band off and replaced the tension band and said “Don’t touch this again” to myself.

knob.jpg

The other bad thing, and something I didn’t notice before, is the speaker had a tear. The paper was so rotten it felt like wet tissue and just kept tearing itself. Well, it still works, and this can be replaced easily enough if desired.

speaker.jpg

So, on to the repair. It was a simple matter to desolder the parts, the heavy copper on the board let it’s old solder go with a 35W iron and some wick. The parts pulled out easily and the board cleaned up well.

burnt.jpg

You’ll notice the board is discolored. That’s where the 50C5 power amp tube sits, and it gets toasty. The solder joints are exhibiting potential cracks, so they all got touched up.

The parts I installed were much smaller than their vintage counterparts, so they got bent leads to sit down on the board, and were just placed in the board. I wasn’t going for a beauty contest here, they need to be in there solid and electrically correct. You’ll notice the one electrolytic is offset a bit, I realized after I put one in that the banana slicer had one section larger than the other. Oops…that other part goes in the parts bin.

line.jpg

filters.jpg

For the filters, I just jumped the two together with a loop, looped a piece of bare wire around the common and soldered it all together. The original capacitor was full of old wax and wasn’t being rebuilt.

It was all back together…but there was still no joy. The filters took care of the buzz, and the output was nice and strong - then it started to fade and get fuzzy as something started leaking. One of the coupling capacitors or other small filters on the plates is probably leaky and drawing things out of tolerance, or a tube is bad. On a whim, I tested a few of the tubes, and they seem ok, so it’s probably one of the capacitors. With some being in that couplet, that’s more of a job that a “I’ll take a look” can cover.

testok.jpg

A short clip of the radio playing. When it’s first turned on, it’s nice and strong but fades quickly. Sorry about the download, Youtube has decided that I need to verify my identity once again, and I’m done with that game.

https://privateemail … 9e11/1/8/NjE/NjEvMTA

It goes back the owner this week, if they want more repairs that’s a different proposition. Some of the younger members of that family are starting to get into electronics, perhaps this is a good project 15 years down the road.

Beyond that, it’s really not cost-effective to repair something like this for someone else unless they understand it’s basically a replace everything deal.

  • 2023
  • Feb
  • 7

A General Electric T-125A “All American Five” Radio

This is a radio a co-worker rescued from a barn sale, and wanted to know if it could be made to work again. The T-125A was manufactured between 1958 and 1962 (or 1963, depending on who you believe.) It still has the civil defense markings on the dial, so this radio was probably one of GE’s last radios to have this marking.

001_front.jpg

096_cd.jpg

The radio apparently spent the majority of it’s life in a woodshop, and it shows. While the case plastic is in excellent condition with no chips or cracks, it’s stained with various wood coloring products. Back and original cord are present and intact, although the cord is getting stiff. It, however, exhibits no cracking so it will get left as-is for now. The inside of the radio shows you where it was for most of it’s life, being coated with wood dust.

002_dirty.jpg

Since these are simple inside, there wasn’t any need to grab an isolation transformer or variac, just plug the thing in. The worst that would happen would be the across-the-line capacitor would pop, so…

It exhibits exactly what you’d expect. A loud 60Hz hum, so the filters are bad. That was completely expected. It’s a dual section filter consisting of a 50uF and a 30uF capacitor, so I’m simply going to get two 47uF @ 150V caps and replace it. The across-the-line cap is .047uF at 600, it will be replaced with something similar. The only other suspect item would be the coupling capacitor between the detector and the power amplifier, so it will probably get replaced as well. The grid capacitors could be eyed with suspicion, but I’m going to leave those alone unless they show problems. One is in a couplet, which would require rebuilding the device if it’s bad.

The owner simply wants it to play long enough to go “Hey, that’s cool.”

A quick cleaning with a paintbrush and some canned air got rid of most of the dirt:

098_clean.jpg

099_dirt.jpg

So it’s ready to repair. I’ll pick up some parts at the next few hamfests I attend.

One interesting thing about this radio is the 50C5 power amp tube. It has different color writing, so I assume it’s a (genuine GE) replacement tube. It still has something interesting about it - the getter is black. I can’t say I’ve ever seen one like this without the tube being red-plated. I have to assume this radio was on for most of it’s life. Pretty cool.

097_black.jpg

Off to a show for parts. Check back later for results on this radio.

The GE T-125A schematic.
Courtesy of radiomuseum.org
https://privateemail … 729ec/1/8/NjQ/NjQvOQ

  • 2023
  • Jan
  • 23

StatusNet: Cleaning up the main page.

This is the second (and last?) part of my cleaning up StatusNet. This deals with removing some of the main page identifiers. While this isn’t hardening, it obscures the doorway just a bit, and hopefully drive-bys will just look at it as a not very worthwhile target. It also removes links to things that aren’t what they were when the service started, so there’s less chance for a user to click on something that may now lead to a malicious site.

This does remove some of the boilerplate regarding the GNU license status, among other things. While I don’t like to do this, StatusNet is so old that there are probably no new installs happening, as it won’t run on modern PHP. If you’re running this commercially, you may be obligated to leave those license notifications alone, so use your own judgement.

***
Here’s the original assets, and what we’re going to change:

001.png

002.png

The first thing I like to do is get rid of the dead link to status.net, and the link to wikipedia. The first isn’t necessary because status.net is just a garbage page full of SEO grabber links, and the second both because I don’t like some of Wikipedia’s practices and you hopefully know what microblogging is now. The term itself has passed into the legend of Web 1.0, so it’s not even a relevant term these days.

(I assume you’re comfortable with your webserver and know where things are. No warranties are expressed or implied, it’s up to you to make sure you’re doing this right.)

Start by making a copy of websiteroot/actions/public.php, and then open the original for editing. (Always back up your original, just in case!). Move down to the very bottom of the file. You’ll see this:

003.png

And when we’re done, it will look something like this:

009.png

To change the header box in question, edit the ’showAnonymousMessage’ function to say whatever it is you want. Use this as a guide, essentially you’re getting rid of all of the links and creating a static header that shows regardless of the other options chosen for the site. That’s why it’s duplicated in the ‘if’ block.

Since the site in question is TheDoghouse, %%site.name%% will show that, followed by ‘Woof!’ - you don’t have to use the variable here, and you can put anything you want in the block as long as it can be printed. I chose to use a simple message that will show “This is TheDoghouse. Woof!” on the main screen where the description about StatusNet and the service description lived. The variable originally used was left intact for this step.

The next edit is just to reduce clutter. In the block directly above the one we just edited is the sidebar links. I edit those out so, while they still are there, the function to actually print them on the screen is commented out with #. This isn’t necessary and doesn’t really offer anything, it’s just something I like to do. The base skin for statusnet will show the same size box, but other skins will reduce the amount of space used based on how much or little you comment out.

Here’s what it looks like by default:

007.png

This gets rid of the “Popular”, “Groups”, and “Featured” links on the main page. While they are still available, to me and my private site it’s just unusued cruft. So away they go. I didn’t delete them in case I want to re-implement them, the function block simply doesn’t show what has been placed in the variables. Similarly with the site name, I left the if/else case alone so later edits will be less painful if certain functions need to be restored. But that’s why you made the copy, right? You did make a copy, right?

But there’s nothing preventing you from putting whatever you want there. Just replace the variable with whatever you like and let the show function print it for you. (I suggest you check out some basic php stuff before going crazy!)

Save your edits, and check that permissions and owner of public.php are correct. Since you created a copy, it’s probably now root, so change that back to whatever your webserver uses as it’s username. load your site up, and you should see your changes.

004.png

The next, last, and other thing I like to get rid of is the footer, and all of it’s links. This shows up on every page regardless, even the main page. While again, this does contain license information, many of the links are dead and the version number of the service is present. As before, there are probably no new installs happening, so having this information isn’t terribly useful. It exposes things that could make it (ever-so-slightly) easier to probe for entry, so it goes away. Again - this isn’t hardening, it’s simply covering up the door so the curious passer-by hopefully doesn’t see it. The microblogging term stays here, but be careful of your license requirements.

For this, navigate to websiteroot/lib/action.php. Make a copy, and edit the original. Navigate to showStatusNetLicense, and you’ll see this:

005.png

What gets shown is dependent on if you have the “Brought By” line filled out in your config. But I don’t care about that, as before I’ve left the if/else cases alone so they can be reverted easier, if necessary. All of the “runs the” information is gone, as is the license, and the service shows the same information regardless of config options.

010.png

The last thing I like to get rid of is the set of links present at the bottom of every page. My site is private, so some of those are not useful. The rest are dead links, and like the now defunct status.net page, may lead to something unsavory. Those are fairly easy to get rid of, navigate to the showSecondaryNav function in action.php and simply comment out the block.

Here’s the original:

006.png

And my edits:

011.png

Everything is commented out except the first few and last few lines, which give the function something to do when called. Other than that, everything else is ignored.

Save your edits, make sure permissions are correct, and check your site.

008.png

That’s it. Your main page now does not show much information about what this is, other than StatusNet.

Troubleshooting: If you’re having some issues where you are now getting a 500 error, check your permissions. That’s probably going to be the main issue here. Also check that you’ve used the correct syntax in the code, variables are always enclosed with %% %% and lines must end with a semicolon; Beyond that, copy your original file over the edits and try again. I found out by trial and error, so don’t be afraid to play with stuff even if you don’t understand 100% of what it’s doing.

There are a few other things you can do inside the service to remove unwanted things like maps (that don’t work,) but I’ll cover those in a separate article.

  • 2023
  • Jan
  • 7

The TechTran 980 standalone 5 1/4” disk drive.

So what is this thing?

One of the devices I’ve had kicking around the shop for years are these TechTran 900 series standalone disk drives. These were the main storage medium for a large test platform at a former employer. They would provide commands to a computerized telco switching frame in the form of responses to a “Read next record” command. When the project finally ended, I asked the project engineer for the drive. Turns out there were 4 units, with manuals and service information.

many.jpg

fronton.jpg

So, it’s a disk drive?

These have two modes. First is a “File” mode, which works like you’d expect from a mass storage device. You create a file, write to it, then read it back. It also has a “Record” mode - instead of writing files with a directory, these store records, much like a punch card would store a piece of information that may or may not be part of something larger. You would write and read records - bits of text - in sequential order.

The unit does this via a RS-232C interface, using text commands. The device can operate in standalone or passthru mode, and has all the normal serial settings you’d expect - except everything is done via switches that control what is literally a box full of 74LS chips.

back.jpg

chips.jpg

How does it work?

Because the device was expected to be continually available while in operation, the manufacturer set these so the disk spins at all times, and the device loads the head on to the media whenever an operation is requested. This has the effect of only requiring the time to drop the head and wait for the record of interest to arrive at the head before the operation is completed. Normally, a drive would spin up, perhaps drop the head on the disk after it was spinning, seek the correct record in the directory, and then move to read it. These spent the majority of their life reading sequences, while the time savings was small it added up to all the years the unit was in operation. File operations, of course, required seeking the record in a directory, but there was still a small time savings in not having to spin the drive up each time.

This made for a very noisy device, as the fan was spinning, the drive was spinning, and the head was being loaded on and off the media surface over and over (with an accompanying clunk of the head load solenoid) each time. Not something you’d want in a quiet space, but the noise of the test frames generally overrode any noise this thing made.

Really?

Kind of. I’m digging into the service manual more because it has a theory of operation. It looks like the record mode does create “files,” but not quite in the same way as file mode. More on that as I understand it.

I have to say I’m somewhat surprised that these things still come up. I figured the belts would have long turned to goo at this point, and the fact they’ve been sitting since about 2000 probably hasn’t done them any good either. They had a long life, being put into operation in the early 1980s and running up to early 2000 or so. But spin up they did, attesting to the fact they were well engineered (aka cost a lot of 1980 dollars) and designed to last for a long time.

The one in the pictures, a Model 980, is going to be the subject of an attempt to make it talk to something modern, which shouldn’t be that difficult as it speaks serial. Set your baud rate to 1200 and stand by for further posts in this series.

Deleted the manuals for now, will package them up again soon!

  • 2022
  • Oct
  • 17

Deleting bad files in Windows

Windows can be quite the turd at times. For example, while it enforces a 255 character limit on a file (both path and name,) it will still happily allow programs like Chrome to save filenames that can be overly long, or even full of invalid characters. You’re left with a file that you can’t delete - or even rename - because it’s invalid to Windows.

The easiest way around this, assuming you have admin access to the computer in question, is to install 7zip.

7zip is a cross-platform archiver that supports it’s own native format, supports archive/unarchive of a number of other formats, and can unarchive just about anything under the sun. (Sorry CP/M’ers, I think SQ is too old for even this program to handle!)

Download 7zip from it’s homepage, 7-zip.org

Once you install 7zip, open the file manager interface that comes with the program and navigate to your bad file. You can delete it right from the interface, as it ignores standard windows conventions and does what it’s instructed - which is what it should do!

I use 7zip for pretty much all archiving purposes, even letting it overwrite the native zipper in Windows. It’s just that useful.

7ziplogo.png

  • 2022
  • Oct
  • 11

StatusNet: Sending email instead of SMS with a new post?

What are you talking about?

If you’ve been around for a while, you’ll remember the birth of Twitter and the open-source clones that came out shortly after. While those clones are still useful, the Internet has become weaponized and services that worked because we had more trust now no longer work as intended.

This article is about StatusNet and it’s ability to send a SMS whenever someone posts a status to a timeline you follow. It uses the email gateway provided by most carriers to deliver a message to you - that is, it simply sends an email to “yourphonenumber@yourphonecompany.com.” In theory, anyway. These gateways now have a very suspicious eye when receiving messages containing links or other words that hurt the carrier’s feelings. They’re all but useless these days, and part of that is due to the ease at which spam can be delivered to your phone.

StatusNet in itself still does the same thing it did back when the last releases were made. We’re still using it for this reason, and because it’s visually more appealing than the lotsawhitespace Gnu Social that followed. It’s probably not something that you should expose to the outside world unless it’s on a low value server, however, as it is over 10 years old.

Yeah, so?

As mentioned, StatusNet can send an SMS any time your timeline is poked. However, this is just an email. Why not modify that so instead of sending an SMS instantly after a status is posted, it sends you an email instead?

That’s perfectly doable, and it’s all but transparent to the StatusNet instance.

This involves poking around in and editing your StatusNet database. In particular, we’re interested in the sms_carrier table. I’m using phpmyadmin because it’s easy to make these changes, but you’re free to use what tool you like. I assume you installed the instance yourself - if not, this may not make sense to you. I’m not going to talk about how and where databases and other files are in this post - if you’re not sure where those are then this isn’t going to be of any use to you.

Open your database by selecting whatever you called it, then navigate to the sms_carrier table. We’re going to make two changes for each of the entries we want to add to our instance. Pick the row you want to edit, and click “Edit.” You’ll be on the screen with all of the things you can change about this entry. (You can also do inline edit, but I like to bring up the full edit screen.)

There are two ways you can do this. Either put a complete email address in as an entry, or use the built-in variable and patterns to allow multiple users with the same domain to enter addresses.

If you want to simply add a complete email address:

The first thing we want to do is give it a useful name: In the “name” field, under the “Value” column (last one) delete what’s there and type something that is meaningful to you. Next, in “email_pattern,” delete what’s in the “Value” column and type your complete email address. Do not modify any other value or function. When you’re happy with the changes, hit “Go” to save it. The StatusNet code will happily ignore the “number” you gave it and send to the email address provided. If something was messed up, hit the “Reset” button.

change.png

If you’re running a multi-user instance, I don’t suggest doing this. Any person using the instance can select your email. They can’t verify it because they don’t have access to it, but they can still select it and try to attach it to their account. A better method is to use the patterns provided by the original programming.

If you want to use the patterns:

You still want to give it a useful name, and for this example we’ll use “gmail.” Delete what’s in the “name” field under the “Value” column and type “gmail.” In the “email_pattern” field, we need to make sure we give it the proper information. You’ll see something like “%s@phonecompany.com” - replace the “phonecompany.com” with “gmail.com” - you should have “%s@gmail.com.” This will allow StatusNet to fill in the “number” you give it later. Hit “Go” and save your changes, or “Reset” to clear and start over.

gmailphp.png

With both of these methods, make sure all of your changes are fully saved, and exit the tool you are using. Now we’ll go to the StatusNet instance in question and log in.

How do I get that email?

When you are logged in, click Connect at the top. You’ll get a screen showing your connected applications, of which you probably don’t have any. There will be two options to the side, SMS and Connections. Click SMS.

connect.png

sms.png

You have two options here, depending on how you set up your database changes:

You added a complete address.

pick the name you gave the address in the “Mobile Carrier” dropdown menu, type anything you want in the “SMS phone number” box (it doesn’t care because we deleted that variable) and hit “Add.” It will process for a bit, then you’ll end up at the screen asking for your confirmation code.

You added a domain only.

In this case, chose the domain you want in the “Mobile Carrier” dropdown menu and type the username portion of your address in the “SMS phone number” box. For example, if your address is “mygmailaddress@gmail.com,” pick “gmail” (assuming you called it that,) and type “mygmailaddress” in the phone number box. Hit add, and you’ll end up at the screen asking for your confirmation code.

gmail.png

If you did everything correctly, you’ll have an email waiting for you with a confirmation code in it. Paste that in the box, hit “Confirm,” and you’re done. Now you can go back to Connect > SMS and select the “Send me notices” checkbox under “SMS Preferences” to receive a notification every time someone posts to you or on your timeline.

number.png

confirm.png

If you don’t receive an email, check that your spam filter hasn’t trapped it due to the name or sender. It’s kind of unusual to get an email for an SMS confirmation, so mail systems may say “No.”

Both of these steps assume you have an email sender set up on your machine, it’s configured properly, the relay or service you’re using is configured properly, and your receiving email isn’t overly aggressive about looking for spam. If not, this isn’t going to work (obviously!)

It works?

So now, whenever someone posts a message that is on a timeline you’re following, you should get an email notification.

itworks.png

status_email.png

It works! There’s still a chance that your email provider can trap the message as spam, especially if there are links that look spammy. Whitelist the email these are being sent from, and you shouldn’t have any problems.

StatusNet itself is getting pretty old, and there aren’t many instances left out there. If you’re like me, and have a couple of internal systems under your control, this can restore some of the functionality lost over the past 10 years.

This is part 1 of a 2 part series about bringing StatusNet a little bit into the modern era. The next and final article will be about removing some of the identifying information to provide a little less of an attack surface should someone happen to run across your instance.

Now tell the little blue bird to get lost. We have some statuses to post!

  • 2022
  • Oct
  • 9

The SwitchBOARd, a general purpose I/O device using the ESP-03.

When I was spinning up a number of ESP-01 devices, and before I found out how much RF noise they put out, I started exploring some of the other devices that were available in the ESP family. Specifically, the ESP-02 (A SMT version of the ESP-01, hard to get) and the ESP-03.

This board was designed to be a general purpose I/O board where you could read or write three individual bits and monitor those bits for control or notification purposes. Specifically, this was designed as a door open/closed device for garages. The ESP-03 was chosen because it was small, had a built-in ceramic antenna, and had the required number of I/O pins. It also looks kind of cool.

The board featured an onboard programming port and switch, a regulated power supply with a temperature monitor for the regulator, and indicators for all of the I/O. If desired, a passthru for the power supply could be placed, as to provide power for another board of this series without needing a regulator and extra supply.

Ultimately, I never could get the ESP-03 to take a program properly, and had just started to investigate why when I found out how bad these devices were at 320MHz. This board is one of two prototypes I built, the other having been sacrificed to the prototype gods during troubleshooting. Some other WiFi devices have shown up on the market, so this board may live again someday.

The boar outline found it’s way on to other projects, including this blog - and this is where the main page gets it’s name: Red Boar Design.

switchboard.jpg