If you have previously subscribed for NetMind to watch this page, you will have to unsubscribe (this applies to your NetMind watch on this page only, not necessarily other peoples' pages), and resubscribe directly with me.
If you want to be notified of changes to this page, go to CoolList and ask to be subscribed to the 'meypoker' list. Note that I fully agree with anti-spam policies including CoolList's, and will not be harvesting addresses for any purpose.
20/Mar/2001 For a long time now, I’ve had the message, hand-type-and-value and current-bet lines pick up their font styles from the main window itself (which means that, if the user should change these, the program only need make one change rather than three); I just realised today that I can do the same kind of thing with the players’ names , the pot’s name, and the associated amounts of money! What I've done is to mount these, not directly in the window as they were originally, but in a green and non-3D (and hence invisible) panel in the top of the window. They now pick up their font styles from the panel — requiring only one change instead of twelve! Neat, eh?
I’ve also worked out a way to simulate network play, for debugging purposes — but as I don’t yet know how to implement network play for real, this knowledge is a tad premature...
16/Mar/2001 Started on writing the help file. I’ve always known that Microsoft Word 97 produces even worse-quality RTF code than it does HTML (in particular, it tends to load the Font Table with every variant of every font installed at the time the document was saved, not just the few actually used); but it’s managed to produce a file which makes no sense at all to Microsoft’s Help Compiler! I had to load the file as text, and manually remove all the fluff which is irrelevant to help-file creation (such as paper width and height, the entire Style Sheet, and so on) — and having done that, Word no longer recognises the document, though Help Compiler now does. You can’t win! :-)
Hence, what I now do is to use Word to edit the “fluffy” version of the source file, then load it and the “clean” version into my text editor of choice (currently Programmers’ File Editor, by Alan Phillips of the University of Lancaster), and copy ’n’ paste the topics that have changed; removing the “fluff” in the process, of course.
The one problem I have encountered with this is that I couldn’t get superscript or subscript to work with numbers for some reason (the code {\super 52}C{\sub 5} should, by my understanding of RTF, produce the desired result 52C5, but instead it it comes out as 52C5 — but {\super *} comes out as the desired *; weird...); to get around this problem I had to modify the special font I've designed, but attempting this crashed Corel Draw 4 (which is what I use to design fonts).
On investigation, it turned out that the program I use to add version and copyright information to fonts (which I won’t name, because I don’t recommend it) was corrupting the font in doing so. I had to redo the whole font from scratch, this time keeping a separate copy without the added information; this, however, proved to be a good thing after all, as it enabled me to redo the glyph assignments in a much more logical manner...
12/Mar/2001 Extended the evaluation routine so that it can now handle jokers. The extension makes the routine about five times the size, and ten times the complexity; but should be worth it...
(Thanks be to the people on the Urban Legends Message Board, for advice and encouragement. No thanks at all to the Ask Jeeves AnswerPoint community, who proved useless...)
18/Feb/2001 Finally got back to work on the Poker project, because I got fed up with fruitlessly searching for a TGF utility. Got the new, improved, much-more-efficient evaluation routine working at last...
26/Jun/2000 The Poker project is still shelved, but not abandoned...
28/Feb/2000 The Poker project is still shelved for the time being, but over the past weekend (because I couldn’t think how to get started on the Project Manager project — that’s always one of the more difficult bits), I’ve been fiddling around with the “Pyramid Numbers” which I’ll need for the evaluation routine. So why don’t you click that link? Go on, you know you want to...
24/Feb/2000 Keeping track of the generations manually (which is the current generation number, which set of floppies needs to be backed-up onto next, which files have changed within the last three generations and thus need to be included in the current back-up, etc.) is proving so much hassle that I’ve temporarily suspended work on the Poker project until I write (or get hold of) a project manager and change tracker tool.
22/Feb/2000 The evaluation problem turned out to be that I was passing the routine the wrong information! For evaluation purposes, the card ranks have values of Two=0, Three=1 etc. up to Ace=12. Thus I was using the code:
z:=Test.rank[x]-2; if z<0 then z:=12;
z
Test.rank[]
The other problem proved to be the usual cause of Access Violations or of GPFs — I was trying to use a pointer before initialising it, that is, the silly prog was trying to write to the Debug window before it opened! D’oh again! Rearranging the procedure calls in the master unit (MEYPOKER.DPR), so that the Debug window opens before any other, has provided a quick and (very) dirty fix; but Delphi now thinks that Debug is the main window, so I’ll have to think of something else, perhaps queuing pending debug messages...
(MEYPOKER.DPR)
On loading the project into Delphi 3, I found that all labels which I’d carefully sized to fit across their background boxes had been collapsed to the exact width of their actual text, making their alCenter attributes meaningless and spoiling their appearance. This was because I didn’t realise that they also had an AutoSize property, which needed to be set to False. D’oh yet again! I should quit this programming lark and get a new job as Homer Simpson — I have the beer-gut for it... :-)
From now on, I’m going to use three-generation filing for this project, so that occurences such as this aren’t completely disastrous...
20/Feb/2000 Disaster! My initial tests of the hand-evaluation routine had it returning a stupidly large (>1,000,000,000 instead of <1,500) value for any No-Pair hand with an Ace in it. I added some debugging statements to try to find out why this was (one of the beauties of Windows programming is that debugging info, instead of having to be written to a file, can be written to a window — so one can see, live as it happens, what's going wrong), and the result is that the program now gives an Access Violation (the 32-bit equivalent of a General Protection Fault) and won't run!
Will I have to scrap the bl**dy thing and start again?!?! I hope not...
18/Feb/2000 After years of wanting to do a Poker game, and months of thinking of converting M. E. Yates’ version to Windows, today I finally got cracking! Looking good so far — see the work-in-progress screen grab below (you may need a PiNG plug-in to see it in your browser window)...
By the way, I’m using QCARD32.DLL to draw the cards.