To work again, over Horgoš. Although there's also something on 7th of march, that doesn't sound plausible - the passport stamps are often hard to read, smudges and stamping over each other and sparse ink.
The listall.prg was edited this day, to deal with embedded carriage returns in text fields.
Note in rebfpt.prg, „close the table just in case, it's triggered by Pack memo sometimes. Tries to use missing .fpt upon closing.“... so I was trying already then to make the rebuild a part of the app, i.e. that it should check for a bunch of most used tables to see if they needed rebuilding, and would offer to do so right away. I know it did work like that a couple of years later, just didn't expect to find traces of that so early.
Went to Pest for two days, because m$ organized something, and Szoftex is, as it were, an authorized dealer. For accomodations, Mihály told Ula to find the cheapest possible, which she did - some remodeled workers' hotel, where the bachelor grunts worked. Hospital type bunk beds, but new, well repainted concrete covered with woven rags, the bathroom a recent add-on, minimal. It had its own restaurat, likewise not looking too good, but the grub was to burst from. Some insane optimism got into us, don't know what great thing loomed on the horizon, so Vanji found a reason to raise a toast, to which Mihály gave him a friendly warning that he was lucky that we were alone. Because... in Hungary, you don't raise a toast with beer. When the Great rebellion of 1848 was crushed, the austrian military governor toasted with beer at the celebration of his victory, and the Hungarians don't do that anymore. Ow fuckit, it's been what, 150 years, and it still ails them? Or only the elderly... Who can know, and who can know whether that's so at all, and what size sample is needed to establish or refute that as a fact? It could just as well be an urban legend (which may also mean something, urban legends are a matter of belief).
Someone held a lecture on Visual fox 3.0, and that's when I first met object oriented programming. The big shuffle around [one's] head, which will in the end be simplest to call replacement of head. Whatever you knew about programming is... just an introduction, now we're getting to the real thing, on one side life becomes easier a lot. On the other it becomes complicated. But also very interesting. And we got a disc with demo version of it, and a of a Windows 95.
I managed to clinch the lecturer during a break, and asked a few things, and then he asked what was my project. I told him how I did PolC in fox 2.6 Dos, how many patients it handles, what features it has, 40 workstations... to which he calculated a bit in his head and said „two years, five or six people“. „Nope, me alone, four months“. He picked his jaw off the floor and gave me a weird look.
It was interesting last fall, when we were at some fair in Pest and Joška and I dropped by m$'s stall just to take a peek, and just for kicks asked the folks there whether fox 3 is already out. How the microsofties exchanged bewildered looks, as in „how did this leak so far into this motherfuckland“... The bluff was an enormous success.
They also demoed enty 3.5 or so, on an alfa processor. It sounded mighty and looked FTL fast, at least on that machine. Word did everything in a microjiffy, never saw it blink or wait for disk. Wow, indeed.
Of course, as soon as we came back, we installed that on our machines and played with it for a few evenings. The deeper, the murkier... nothing cleared, but some ideas started spinning around [my] head and finding their places.
On seventh mucked some with validations in PolC, as few more things need to happen when entering materials (in the lab), or for hospitalized patients, and for reception desk, when catal6 needs to be called to pick from a list of known senders, i.e. doctors' practices. Then more work with the history view form. And more fucking with postal codes, because they were four digit, and now the state started adding the fifth digit to deal with ambiguity (two or three places had the same 4 digit code), so now I'd have to check it as 4 first and then as 5... And they still had fuckups both ways - there were codes belonging to two or more places but now also places with two codes. Complete chaos.
On 8th extracted alert.prg as a standalone prg from something that started as Sale's code, and started pushing it to work the way we wanted. Logic was mine, Joška did the design - the buttons should have proper shade, the default button should have a different color, the window should have a shadow etc etc - all of it with no pixels at all, just characters. The point of the exercise was to have arbitrary buttons, which will be automatically lined up, horizontally if up to four, vertically for more. And the text of the prompt should be wrapped up nicely in multiple lines and look nice.
Much much later we understood how wise we were here, when we saw what kind of bullshit is the m$'s messagebox(), which has about eight possible combinations of predefined buttons, and returns some incomprehensible magic numbers. alert.prg returns the ordinal number of the pressed button, or zero for none. As simple as that.