Carleton Weather

That’s right – we are emulating the Apple ][ computer, running the ancient weather program, and “printing” tables of old weather data to text files on the lab iMac. Here’s what we did:

Setting up the connection
The Apple ][ serial card cable has a 25 pin female end. We hooked that up to a 25 pin male adapter, which we hooked up to a 9 pin male adapter, which connected to a 9 pin female null modem cable, which in turn we were finally able to hook up to our Keyspan 9 pin male serial port/USB adapter (plugged into the Mac USB port). We then booted the Apple ][ and started up ADTPro (version 1.1.9) on the Mac and were ready to go.

Sending ProDOS and ADTPro to the Apple
After starting up the Apple ][, we hit “control+reset” to get to the command prompt. On the Mac end, we selected the “Serial” option in the ADTPro window, kept the default highest baud port settings when prompted, then under the “Bootstrapping” menu we selected “ProDOS” and “Send ProDOS”. We were then prompted (on the Mac screen) to type the following into the Apple command line: IN#2, return, then command and A, and 14B, return. We could then click “OK” in the ADTPro dialogue box on the Mac, and the ProDOS kernel started streaming to the Apple. When this was done, we could do the same thing for the ADTPro program (“Bootstrapping”>”ProDOS”>”Send ADTPro Serial Client”). The steps were the same for this data transfer, and afterwards we had the ADTPro program installed on the Apple ][ atop the ProDOS kernel. We made a copy of the program on a 5.25″ floppy so that we can reload it whenever we need to – we typed “F” on the Apple to format a blank disk, “R” for receive and “ADTPro-1.1.9.DSK” in the filename prompt, then pressed return and selected our formatted disk as the transfer destination.

Getting weather files on the Mac
Back on the ADTPro main screen on the Apple, we typed “S” for send. When prompted we provided a filename with a .DSK extension and pressed return. We inserted the disk we wanted to copy into one of the disk drives and scanned the drives. We selected the disk we wanted to send to the Mac, and the data from the disk was transferred to the Mac as a .DSK file. Once we had a weather data file and the weather utility program on the Mac as .DSKs, we ran the Virtual ii Apple ][ emulator and inserted the weather utility disk file into the virtual drive. The program then ran on the virtual Apple ][ as it would on the real one. Now in Virtual ii we clicked on the “setup” icon and under “Apple][+” clicked “Slot Connections”, then “Grappler+card”, and then deselected the Epson printer and instead set the connection to “Text printer”. After this, we could choose to print a weather table with the weather utility program, select a monthly data disk as we would on the Apple ][, and the data is printed to a beautiful tab-delimited text file on the desktop rather than to a pdf from the virtual Epson. We also selected “Maximum speed” in the Machine menu so that printing out a month’s worth of data to a text file only takes perhaps 30 seconds. Very cool!

This is absolutely the neatest thing ever. Tomorrow I’ll bring in a camera and take some pictures of our setup to post here.

-Katherine

After toiling with the plotting scripts for the past two days, it now works! Technically I still have to update the daily and monthly scripts, but I know exactly what I have to change. It was a combination of old functions and me mis-naming a field in my database. Oops…

-Will

I am recording Doug’s account of our latest approach to obtaining the Apple ][ data here, for our own reference and the benefit of others who might want to do something similar. I think he says it better than I can.

“I talked with Troy Barkmeier in ITS, and he suggested a slightly different tack for the project. He suggested we move all the data and software programs to a modern computer (and showed me some ways to go about that), and then use an Apple ][ emulator to run the utility program and translate the data disks (which will be files on the modern computer).
Katherine downloaded and installed an Apple ][ emulator (Virtual ][), and got it up and running.
She also downloaded and installed a program (ADT Pro) for getting the two computers connected.
I brought over a Keyspan USB-Serial adapter and we loaded the drivers for that on the iMac. (We wouldn’t need this adapter if we decided to transfer things to a PC.)
I found an Apple Super Serial Card (SSC) in dead storage, along with its manual. The ADT Pro website helped us set up the DIP switches (which we did) for the initial transfer (which we haven’t done yet).
We still need a serial cable (25-pin to 9-pin) and we’ll have to figure out if it’s a straight-through or null-modem cable (there’s a HW switch on the Apple SSC that needs to be set according to the type of cable being used). Ill bet you have a cable that will work for this.
Tomorrow morning, we can try again to get the old and new talking to each other.”

-Katherine

After reading through all the documentation for Python’s time and datetime libraries, I’m pretty sure the timestamp system I used in the program I was talking about was of my own invention. That may be what we want to do for the database too – I think it would be easy, more comprehensive and more reliable than UTC timestamps, whose limitations depend on the platform but seem to be 1970-2038 on most systems. On another note, the USB/parallel port cable arrived today! Also, from what I’ve read on sites documenting RealTerm’s capabilities, it should actually be able to capture data from the USB port since it sends data via essentially the same type of serial stream as a serial port. I guess we’ll find out when we try it out on Monday!

Katherine

After spending the morning working with mySQL and getting a near-replica of the current database structure set up, I got the almanac to work! I started CSS-ifying it, but I have only gotten 1/3 of the way done with that. Finishing that should not take too long. In the next couple of days I will be able to move on to the page that displays the graphs, and then the site will be next to done.

-Will

After a lot of reading and re-reading the code, I figured out what the problem with the display was. There was a commenting out being added by php which was never closed, which threw off a bunch of the subsequent ones. I was hoping that would solve most of the issues I was having, but now all the data is displaying in the wrong places. Unless on may 2nd, 2010 the high temp was 2010… i havent solved that issue yet, i’ll see what I can do tomorrow.

-Will

In the time since my last post I got the current working version of the new website from Will, so I tweaked the mobile stylesheet for that version of the homepage. This version is actually better suited to mobile styling because there are fewer links, so they can be made bigger and easier to read – they were kind of squashed before. I also have the links working in this version. Take a look here: http://137.22.213.193/weather/6-20Draft/index.php
Right now I’m trying to figure out how to resize the logo image to make the text more readable – all the resizing method’s I’ve tried only stretch the blue background.

-Katherine

Ran into a snag with getting the pages to communicate with my SQL database, but it is sort of working right now. Not a lot to report otherwise.

-Will

After almost a week of messing around with CSS features, as well as discovering that the mobile stylesheet is actually layered on the plain old screen stylesheet (at first I thought the stylesheet wasn’t working again!), I now have a pretty readable working version of the homepage for a mobile device. Again, the url ishttp://137.22.213.193/weather/WillCode/index.php. Compare the computer screen version to the layout you get on your mobile phone, and give me feedback on the aesthetics and navigability! The links don’t actually work on this version, but can you tell they’re buttons? Can you read them? Is it easy to click them precisely? Is the current weather information readable? Let me know!

-Katherine

Couldn’t really think of a clever title. I got all the scripts seperated and not conflicting with one another, next step is finishing the displaying of the info and then finally hooking the whole puppy up to a my sql database and seeing if it works with that component. Today was (hopefully) the last of the really tedious boring stuff for now.

-Will