Herb Grosch Color Slides

Frank da Cruz
Columbia University
28 February 2011

Cleaning out my office today I found a long-forgotten CD that Herb Grosch sent me on April 18, 2004, containing his scanned color slides, 1947-1998, mainly of his travels and adventures and home life, not so much computer-history related, but there might be some surprises. Some of these shots were to go into the third edition of his book, "Computer, Bit Slices of a Life" but he never got to finish that project.

The CD contains 116 directories ("folders" in modern parlance), each one with JPGs of the slides, some of them a bit dirty, about 2000 unique slides in all. The directories have descriptive names. I have no other information about the pictures.

To make the Web-based albums, I used Kermit 95 on Windows to upload the contents of the CD to the Unix server where the website files are:

cd f:
send */*

C-Kermit on the server creates each directory dynamically, as the files arrive, but replaces spaces by underscores because spaces in file- and directory names are A Bad Idea, and then uploads the images in binary mode to that directory.

There is already a Kermit script (“photoalbum”) to create a photo album from a directory full of JPGs. I wrote a one-off script to use photoalbum to create a photo album in each of the 116 subdirectories, taking the title of the index page of each album from the name that Herb gave to the directory, setting the permissions for public read access, lowercasing all the names (because some were .jpg and others were .JPG, which can be confusing), and then running the photoalbum script to make the album for that directory, all in a loop.

It took several hours to do all 116 directories because each photo had to be resized twice (using the Image Magick suite on Unix), and Herb had made two copies of each photo, one in the appropriate directory and another in the “All1” (See all at once) directory. Here is Herb's cover sheet for the slides:



Hello !

These albums are a collection of more than two thousand pictures, selected from a scanned collection of about 4400 35mm slides that were taken from 1947 through 1982. The first, as you can see from their poor quality, were taken with a pre-war Argus. The square ones that come later were taken with a Robot. The best ones were done with a rangefinder f/2.8 Nikon (you will see the brown camera bag in a few shots; it also contained an excellent telephoto lens and a wide-angle one, flash equipment, an exposure meter, extra film, and an altimeter).

Automatic ranging and exposure, and Kodacolor negative film, were not yet attractive. I never had a medium-format camera. Only four rolls of film have lost color.

The slides were stored for many years in rotary projection holders with 100 slots. The initial labels were Rnn-xx, with nn going consecutively from 01 through 44 and xx usually going from 01 through 99 with occasional gaps [100 was coded as 99+].

In transferring selected pictures from reels to album folders I used the nn-xx coding. The original scans were 1024 by 768 pixels, and are retained in a huge 16GB record. In what you have here, however, the files have been compressed to .jpg format and grouped in albums of 99 small folders, whose names give information about the contents. The files within each folder, and the folders within the album, are in chronological order.

I suggest that viewing by Filmstrip or thumbnail methods is useful to zero in on a few pictures, but just picking a folder and doing a slideshow or clicking consecutively is quite feasible, since most collections are short. For anyone who wants longer runs, a click on m9f opens a cluster of ten folders from m0 through m9. To run the whole 99 folders of an album as one show, click ALL1.

This Summer I will pick individual items from the albums, and add scans from other media, to key into the text of COMPUTER . Those will be on a CD with a different name, but I mention it here so you can anticipate such duplication. These albums are for visual pleasure. I hope you enjoy them.


The non-standard numbering of five folders 201-205 is due to the late discovery, in forming the decifolder 29f, that I had not selected any pictures from Reel 05. This has been corrected, and 29f and the centifolder ALL1 are complete.