On Wed, 26 Oct 2022 14:55:05 +0000, Seymour J Metz wrote: > ... >Downloading and parsing multiple index.html files should work for some but not >all of the manuals that I've downloaded. > How are the forms codes useful? Ordering hardcopies? Citations?
Wandering only slightly OT, I notice that when I download and unzip the 0.8 GB archive from <https://www-40.ibm.com/servers/resourcelink/svc00100.nsf/pages/zOSV2R5Library?OpenDocument>, the metadata times are reasonably distributed over a couple years with some clustering around the ends of odd-numbered months. However, the file times (assuming UTC) are distributed through the several minutes the download took. This fits the astonishing assumption that the PDFs are rendered on-the-fly, or perhaps just copied from a repository, during the download! I'd find it more useful if the file timestamps matched the metadata. But I have a script (using "pdfinfo") to adjust those. I'd also be grateful if the timestamp of the .zip archive as reported in the HTTP headers truly represented the last update, or if the site supplied an ETag so I could poll for updates and download conditionally. -- gil ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
