Collection of drawfiles approached 1600 items. Idea occurred to me that some form of automatic thumbnail-sketch printed catalogue production was necessary.
Idea immediately shelved as being wildly ambitious for the likes of me. Tried to sell the concept to real programmers. Politely but firmly fobbed off by some, and disdainfully ignored by others.
August 1993
-----------
Drawfile collection now 1800 items. Acquired a new hard disc drive, and dedicated the old fiftymeg to the collection of drawfiles.
October 1993
------------
Drawfile collection now 2050 items. Space on fiftymeg getting tight. Thought about producing cataloguer again, and emerged from epileptic catatonia three days later in local institution.
November 1993
-------------
What a find! Picked up another 250 drawfiles, all high-quality.
Compressed drawfile collection, using ArcFS. Not thinking about cata... cata
February 1994
-------------
Received six 160K stiffies of drawfiles, all compressed. Had to unArcFS collection, and redid it with SparkFS's Zip deflate mode to squeeze a bit more out of the disc. Situation getting a tiny bit desperate. Am I a fontophile? Justin sniggers whenever he sees me. My kids press themselves against the passage wall as I pass. Receiving anonymous phone calls offering 420Meg hard drives at inflated prices, with what sounds like smothered giggles in the background. Linda said "Hello" to me today. I wonder what she meant by that? Seems the whole world is going mad. Ah well -- at least I'm still sane.
27 April 1994 (version 0.01)
----------------------------
Could not stomach the 10Km long voter's queues at all the polling stations, omnly to find that they had run out of ballot papers / ballot boxes / invisible spray ink / ultraviolet lights / IFP stickers / pencils, or that a strike amongst the polling station staff / IEC monitors / international peace monitors had broken out over a wage dispute. Sticking needles into my eyes sounded like more fun, but doing this program was less painful.
Started on beta-test version 0.01.
04 May 1994 (version 0.10)
--------------------------
It is finished.
Version 0.10 sees the light of day. It works, 100% correctly, but needs a bit of tidying up. Will buy RiscPC as personal reward. I deserve it.
06 May 1994 (version 1.00)
--------------------------
Began tidying up.
Anti-aliased the fonts (they weren't before).
Added user-selectable tiling option (2 to 10 rows of columns).
Made user interface a little more standard.
Added user-selectable fonts.
Prettied up the print dialogue.
Couple of other minor things.
08 May 1994 (version 1.01)
--------------------------
Did a bit more tidying up. Improved user interface a bit more, by supplying a better view control window, and the ability tp go back a page, or to the start, and to not need to quit when accessing a new catalogue.
10 May 1994
-----------
Added a few comfort windows. Nothing major. Not worth calling it a new subminor version.
11 May 1994 (version 1.02)
--------------------------
Added ability to render untransformed text objects.
12 May 1994 (version 1.03)
--------------------------
Added speed optimisation for window updates, using Jon Ribbens' WimpExtension call to check if all or part of a tile was within the rectangle to be redrawn. Excellent module.
15 May 1995
-----------
Added ability to display the DrawCat textfile, using Message_DataOpen. Don't know if I've done this right. Calling Wimp_SendMessage (18) as a broadcast does cause !Edit to load it, *if* !Edit is running. If !Edit is not running, then nothing happens. I assume therefore that User_Message_Acknowledge (19) must be returned to my task, and I handle it (i.e., reason code 19) by !Run-ning DrawCat via Wimp_StartTask. It works, but as I say, I'm not sure that it is right. I got the algorithn from !MailMan, and !MailMan does *not* have code to manage reason code 19. I cannot run !MailMan, as I have no network, so I don't know that it actually *does* run !Edit not already running. As far as I can see, I've done it is exactly the same way that !MailMan does it as far as User_Message_Recorded is concerned. I have that queasy feeling that I'm something wrong...
20 May 1994 (version 1.04)
--------------------------
Tidied up a bit more. Caught a couple of non-fatal but scruffy bugs. Got that problem above sorted out -- Filer_Run does the trick in one. Silly me. Speeded up the rendition by checking each path object in turn and only rendering it if is in the visible area -- the previous effort only checked whether it was within the bounds of the surrounding box. Also sorted out the failure to reproduce colours in screen modes with more or less than 16 colours (now that I have a low-serial-number RiscPC, heh heh!).
24 Jun 1994
-----------
Gave a fatal error if !Printers had not been seen yet, on RiscPC. Added an XOS_NumberFromString and read flags to see if it knew about the PDriver module, and if not, to give a non-fatal error (asking you to run !Printers first). This was not a problem on the Archimedes, as PDriver was in ROM.
Corrected a minor DIM problem -- insufficient indices to tile a page 10×10.
26 Jun 1994 (version 1.10)
--------------------------
Style’s OLE is so cute! I jus’ *hadda* do something like it. Besides, as mentioned, some vector programs don’t store the bounding boxes correctly, and one then has to load the offending file into !DrawPlus and resave it to correct the error. Now you can! See the !Help file.
