Re: [ql-users] WMAN progress
On 30 Nov 2002, at 0:18, P Witte wrote: This is not documented anywhere in my Qptr manual!! (3rd revision) so its the first I hear of it. It looks like its been in the code all along. Ive needed such a call many a time! Apart from the timeout, are there any other parameters? Enclosed a small text file with some explanations. Some of the comments in this are mine. Wolfgang - www.wlenerz.com QPTR Updates The following features are available on the following systems: QDOS, Minerva, SMS2, SMSQ and SMSQ/E. Additional information on WM.ERSTR == This manual did not mention that there is a limit on the length of own error messages. An own error messages is easy to create: LEA own_msg,A0 ; get address MOVE.L A0,D0 ; into our error register BSET#31,D0 ; an error is negative Now the limit: the length of the string is limited to 40 ($28) characters. If it is longer, unknown error is returned instead! Additional information on WM.LDRAW == WM.LDRAW clears the change bit in the status are of every item which is selectively redrawn. Additions to the CONFIG standard The attributes for strings have been extended, to allow menu-driven CONFIG programs better options for a selection, depending on the type. There are two additional bits used in the string attributes: 8 and 9. These define the type of string, so that the CONFIG program can treat these strings in a special way. The possible combinations are: cfs.sspcequ %0001 ; string strip spaces cfs.fileequ %0001 ; string is filename cfs.dir equ %0010 ; string is directory cfs.ext equ %0011 ; string is extension At present, these features are supported by the new MenuConfig, and ignored by the standard config. Undocumented SuperBASIC Procedure = SPTR has never been documented. Easy to guess, it does the same as IOP.SPTR, i.e. moves the pointer to a given position. The syntax is: SPTR [#channel], xpos, ypos [,key] Option Default Meaning xpos, ypos nonenew pointer position key 1 origin key The origin key should be zero if the pointer coordinates are absolute. A key of -1 will set the position relative to the current window definition. A key of 1 will set it relative to the hit area. Undocumented selection keystroke for SuperBASIC === It is possible to put an underscore under a selection key for text loose menu items and text info items. To do this, specify the type to be text minus twice the underscore position. This means, to underscore the first character, give 0-2 (=-2), to underscore the fifth position give -10 etc. The following features are supported on the following systems: SMSQ and SMSQ/E from Version 2.71 onwards only. Additions to RPTR = The QPTR RPTR call has been modified to accept job events in the most significant byte of the termination parameter. The job event values are, therefore, multiplied by 256. Note that while all pointer events that have occurred since the call are returned in term%, only those job events (including pending events) which caused the return are returned in term%. term% = $2001 Return on button / keystroke or job event $20 RPTR #ch, xabs%,yabs%, term%, swnum%, xrel%, yrel%, bt$ Event 32 is notified by another job so the wait is terminated and term% is set. PRINT term% DIV 256 Prints 32 Additions to IOP.RPTR and Pointer Record Bits 23 to 8 of the event vector in the pointer record are already used by the Window Manager. The 8 job events are, therefore, mapped into the most significant 8 bits (pp_jevnt) of the event vector within the pointer record and for the IOP.RPTR operating system call. Note that while all pointer events that have occurred since the call are filled into pt_pevnt in the pointer record, only those job events (including pending events) which caused the return are filled into pt_jevnt. New Pointer Event = Pointer event bit 6 (number 64) is now used to indicate that the pointer sprite has hit the edge of the screen. e New Operating System Calls == -- sms.sevt Send Event Trap #1, D0=$3A callreturn A0 destroyed D1 destination job ID destination job ID D2.bevents to notifypreserved Common error returns err.ijob The events in D2 are sent to the destination job. If the job is waiting for one of these events, the job is released, otherwise all the events are pended.
[ql-users] WMAN progress(wm_rptrt)
In-Reply-To: 002201c29806$d8b40f60$0100a8c0@gamma This is a very useful call and is documented by JHudson in his WINEXPLO program which explores some errors (in the C68 wrapper ?) for which he provided corrections. The call is int wm_rptrt(WM_wwork *wwk,timeout t,unsigned short event); I use it for transient pop up windows and providing cross-hairs in my drawing program. Christopher Cave
Re: [ql-users] WMAN progress
On 29 Nov 2002, at 22:42, Phoebus Dokos wrote: Other tricks include: fiddle with the TIMEOUT values, change the scroll value to greater than 200, reduce the shadow to 1 pixel or none -never tried the latter really but I will now that I thought about it-). Use only two colours for everything (I have a nice set of blue). I have to dig up my configuration and send it to you. Why not send it here, so that we can all profit from it? It is. The anti-aliased version comes as a separate download... from where? What would make ProWesS really fly would be a complete rewrite in assembly :-) Any takers? Phoebus - www.wlenerz.com
Re: [ql-users] WMAN progress
??? 30/11/2002 3:15:57 ??, ?/? Wolfgang Lenerz [EMAIL PROTECTED] ??: On 29 Nov 2002, at 22:42, Phoebus Dokos wrote: Other tricks include: fiddle with the TIMEOUT values, change the scroll value to greater than 200, reduce the shadow to 1 pixel or none -never tried the latter really but I will now that I thought about it-). Use only two colours for everything (I have a nice set of blue). I have to dig up my configuration and send it to you. Why not send it here, so that we can all profit from it? Sure why not it's a text file only... Gotta find it first though :-) It is. The anti-aliased version comes as a separate download... from where? URL: www.dokos-gr.net/~dj/indexsms.html :-) What would make ProWesS really fly would be a complete rewrite in assembly :-) I see you smiling Wolfgang hm :-) Phoebus (Off to bed... finally... Oh God! I HATE VB.NET!)
Re: [ql-users] WMAN progress
Phoebus Dokos wrote: On SMSQ/E v.2y99 on the QXL with Mode 4 it's as fast as WMAN (using the trick with the fonts, ProWesS reader renders almost as fast as any Windows program) and if you use the gcc compiled version (available from www.dokos-gr.net/ ~dj/indexsms.html ) That and the anti-aliasing obviously was it. I thought I'm already using the GCC compiled version, but I did play around with it a lot while developing the mode32 driver, so I might have changed it again or whatever. Anyway, it's now really a lot faster than before, very good. Unfortunately the pre-compiled GCC version on your site is an older one which still inhibits a bug in the clipping algorithm (can be seen in Paragraph). Can anybody compile a new version? Yes the Guardian (Times, Guardian... once you see one newspaper you've seen 'em all ;-) font by me, freely available via the usual sources :-) (Shameless plug I know... but in any case its frei as you Germans say ;-) Yes, already discovered that one :) In 12pt size it looks really good. What I like most about ProWesS is the type concept. I actually started writing replacement 3D types some years ago but never had the time to continue. I think I really should continue doing that. It is now reasonably fast here and with some modified types it could look really cool. Did anybody else experience a problem that sometimes PW applications just won't start? The job is created but then nothing more happens. It happens often enough here to be considered reproducible I also succeeded in crashing QPC completely a few times (e.g. once by hitting Quit in the calculator). Can't really reproduce this one, though. Marcel
Re: [ql-users] WMAN progress
Wolfgang Lenerz writes: This is not documented anywhere in my Qptr manual!! (3rd revision) so its the first I hear of it. It looks like its been in the code all along. Ive needed such a call many a time! Apart from the timeout, are there any other parameters? Enclosed a small text file with some explanations. Thank you for that. Some of this information is quite new to me. Reminds me a bit about the Qmake-Qmac/link issue. There must be a better way of getting one's information updated than running a poll loop to aggressively scan for updates on every possible utility one uses! Does the concept of an interrupt ring bells anywhere, I wonder? Per
