Sixteenth, a major improvement in rebfpt.prg - itt now skips the first 32 bytes of a disastered table, simply assumes that the beginning of the first records is on the first byte after the list of fields. Because why read the header at all, when the table is already screwed up.
Working on genhlp... gen_hlp.prg (or the hlps version) bears a timestamp left by upitig2.prg on its generated routine, which was to be used to enter the key to the menu branch being inserted. Though I guess the whole menu.dbf would be inserted at once (it being a hierarchical table), so I guess the purpose of this little form was to enter the starting point in the hierarchy. The keys didn't really have to be fully hierarchical, ie. the menu key is not a rip field; there was also no parent key field. Instead, there was (which is Sale's trick) a child key in the parent record, which didn't really have to be a key. If it began with MN it was, if not it was a procedure to do, i.e. the record was a leaf. Quite simple and also allowed for more than one place to call the same branch of the menu (which then confuses the users, as they try to memorize where they found this or that, and finding the same in two places is disorienting).
For a number of years we had a generated menu in DBA; one of the first things I did in GenerAll was to build the menu on the fly from menu.dbf, which helped a lot in PolC, as the menu was built after login, so depending on who logged in, some portions of the menu wouldn't be just greyed out - they'd simply not appear at all. The menu was dynamic, well at least at the moment the user logged in. It didn't change afterwards.
Avai was in the little building owned by Bangro on my side of Žitni. Few months later, IIRC, we moved to their big building by the west side overpass. I sort of remember it was in the spring, but that's all. Perhaps spring 1997? Nope, there's a zip on 28th with a version of assets, containing even a the.log, metadata for catal6.prg, code generated by rebfpt.prg all on that date, and with that volunteer girl's nickname in the zip's filename. She didn't stay long, and left before we moved. So the move should be in spring.
Lena developed a fine little confusion, which then stayed in the house slang for several years. I don't know who tried to explain what an anđeo* is, which she memorized as anđel anyway (the shift of final l into o is specific to serbocroatian and lots of people never get it right anyway, see footnote), and for her the word meant bare child's butt. All the way to „you can see her anđel“.
Which is where I got the idea that this may have been a contributing factor to pedophilia, because so many guys never had a chance to see a proper nice ass or any other ass, unless in church, on those pictures. So it was always a child's ass, no larger sizes in stock. Who knows how many guys grew up with those images carved into their heads.
----
* so yup, anđeo is angel. This -eo and generally transformation of the final -al, -el, -il, -ul into -ao, -eo, -io, -uo is a rule with several kinks, which probably confuses anyone learning the languages, and in many cases even those who are born in it. It was automatically applied to any words which came into the language until about XX century and then it stopped - the raincoat is still a mantil, not mantio. And it's not sticking along through the cases, the -l returns in slanted cases, so a rooster is petao, but declenses as petla, petlu. The words with the -lac suffix (one of the six or nine that mean the doer of an action) are the worst, in nominative it's -lac (gledalac, posetilac), in genitive it's -oca (gledaoca, posetioca), in genitive plural it's -laca (gledalaca, posetilaca)... and then a newspaper says „pred očima gledaoca“ (in the spectator's eyes) which means there was only one viewer on the stadium.
23-XII-2019 - 16-V-2026