In the 1980′s I heard a rumor that IBM was the largest producer of printed material in the world. I guess that could have been true, because where I worked with MVS there was a large room filled with all sorts of manuals, including a full time librarian. Many people also ordered their own and it was kind of a badge of honor if you had a string of IBM manuals on a shelf at your desk.
Later we used BookManager, where manuals came on a CD in a format that seemed only known to the BookManager program you needed to install on Windows. I didn’t like the format because the designers had taken the original manuals and divided them up into (what they probably thought) were logical sections. For example, I would be looking at an OPEN macro and see the general syntax, but where are the details that follow? That’s on another page, and often I’d have to backtrack to a higher level to find that other page. I often just went to the printed manual instead.
Then along came pdf versions of the printed manuals, probably right about the time IBM stopped printing things on paper. Finding things is difficult because you first have to find the right manual, and then use ctrl-F to search from (generally) the top down until you maybe find what you want. But still, once I get to that spot I found them easier to use than the BookManager format. These pdf files were available via the web, right alongside another option which kind of simulated the BookManager split-up format. I always went straight for the pdf files and downloaded what I needed to my PC.
Later came Knowledge Center which had more features I guess, but I didn’t really use it much because it also resembled the BookManager format. Plus, as happens often, someone decided they would pile everything into one clump, so if you’re searching for a JES2 keyword, you could easily end up in a section totally unrelated to JES2. I seem to remember BookManager was like that too.
Disclaimer: A lot of what’s above is just my view and may be way off
So Why Am I Here?
There’s a chance I’ll be working on a z/OS 2.3 upgrade for an IBM customer, so I thought I might as well look at the latest IBM manuals and download some to my PC. I think the last time I did this (large scale downloads) was z/OS 1.3 so it’s time to look at all the changes IBM has been making to their manual web site. Google sent me here:
There are some things that look pretty cool here, but the thing I heard about that I’m really interested in is the link near the top of that page:
With that one link, we can download a zip file that contains the entire library, plus an index.html file that translates IBM’s difficult-to-use filenames to a readable title. After unzipping I ended up with files like this:
Notice that all the files ended up in a single directory. For example, you’ll find a Comm Manager book in the same directory as an MVS Macro manual. At first that sounds like a mess, but it allows cross links between books. So let’s keep it that way. The index.html can tell us where things are:
Some things I think are odd about the index.html
- You click on the pdf file name instead of the title, that confuses me a bit
- All sections are always displayed even though I may want to concentrate on one section
- Some of the links point to external web pages (i.e. Principles of Operation). I’m not sure why they did this.
- Some of the links point to things not found in the downloaded zip file (i.e. most of the Cryptography section). Was that done on purpose to keep secrets? The 2.3 Program Directory also has a link but seems to be missing.
- In Chrome (my html file default) when I click on a link it opens the PDF file within the browser. That’s not what I want, I want a new Acrobat window. There is probably a way around this in Chrome, but I’ll skip that because I have other plans.
What I would like is to have all the files in a local directory, such as c:\manuals\zos23, and then have a bunch of Windows shortcuts defined in various grouping directories so I can find things first by group, then by manual title… something like this:
So I’m thinking, what if I write a script that will read through the index.html file supplied by IBM, grab the group (i.e. BDT, BookManager, etc.), the book title, and the file name, and then create a bunch of shortcuts that look like the display above?
Turns out this was not too hard to do, but I had to learn just enough VBS scripting to get by. If you want to try the script, here it is: (you will need to copy the text to create_shortcuts.vbs in order to run – see instructions in the code).
That’s it… good luck