home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Atari Mega Archive 1
/
Atari Mega Archive - Volume 1.iso
/
lists
/
gem
/
l_1199
/
1105
< prev
next >
Wrap
Internet Message Format
|
1994-08-27
|
7KB
From: Mark.Baker@mettav.exnet.com (Mark Baker)
Date: 27 Jul 94 19:52:22
Message-Id: <UUCP.775429226@mettav>
Subject: Re: GEM-List
To: gem-list@world.std.com (gem-list@world.std.com)
Precedence: bulk
Warwick:
> Now he's arguing with the author of the library about its size, when
> the archive is sitting there for all to see, at atari.archive. The
> archive is dated 23/11/93, and contains files with the sizes I quoted
> above. As for `they bombed more often than not', I can say that the
> example.prg of GEM++ 1.99 only runs on 68030 machines, due to GCC 2.5.7
> being used to compile it.
It crashes on my Falcon as well, as I've already emailed you about. The
editable fields in Nethack crash as well. Could this be a conflict with LTMF?
> This is a different issue. Graphics cards can perform screen-to-screen
> blit operations much faster because they use on-chip acceleration.
> Blitting mainmemory-to-mainmemory is usually pretty fast too. But blitting
> between the RAM on a graphics card and mainmemory is the bottleneck I was
> talking about.
>
> I know it might be hard to resist thinking of one's own machine as the
> target for one's software and optimizing for that, but this really must
> be avoided, especially by anyone want to follow a standard.
But surely optimising for STs, TTs and Falcons, with or without Overscan,
Blowup etc, is better than optimising for graphics cards which a tiny minority
have.
> This opens up a large relm of possibilities if this can become a common
> feature of GEM applications:
>
> 1. Point-to-type
> 2. Exit highlighting
> 3. Pointer-form change
> 4. Balloon help (ala Mac)
>
> Now, at first glance, many of us would say "NO!" to at least 3 of
> those, but our main reason would be "It's not GEM". Yet, it *is* GEM.
> I'm using GEM to do it, and do it efficiently. Most applications don't
> make any use of the event rectangles... because they didn't implement
> any of these `hot' features. (I take the term hot from Tim Oren).
Pointer form change I would definitely support and it _has_ been around since
day one. But not the others. Point-to-type is confusing IMO, and would be even
were it not totally different to GEM conventions. As for exit highlighting I
can't see any point in it, although it is harmless. It doesn't look very nice
with 3d buttons though. Balloon help... why not, it's a pity there isn't an
efficient way to track menus (it can be done using timer events)
Tim:
> Here's an idea for operating back-ground windows... put an event
> rectangle around your window and watch for the mouse to move outside of
> your window. When it does, monitor all mouse activity
> [wind_update(BEG_MCTRL)], and when you get a click, process it. If it's
> an action on one of YOUR windows, then select buttons, whatever; top if
> necessary. If it's someone else's window, send a message to the
> application, telling it to top the window.
What if you get a click on another window but outside the client area? GEM
doesn't guarantee the positions of the gadgets, so what would you do?
> 1. When you find another app's window and find out its handle, how do
> you get the application ID of that app so you can send it a message?
Only possible with aes >= 4 AIUI. You use wind_get().
> 2. How do you detect the presence of MultiTOS?
> 3. How do you detect the presence of Geneva?
Depending on what exactly you want it for, test the aes version in
_AESglobal[0], or the number of concurrent applications in _AESglobal[1] (names
for lattice, the array has a slightly different name under other compilers)
To distinguish between Geneva and MultiTOS
> 5. My compiler (Lattice C 5.6) has full support for MultiTOS... what do
> I do about Geneva?
Geneva supports most of the new aes features of multitos, although you should
check appl_getinfo(). It doesn't support any of the MiNT extensions. As for
Geneva-specific functions I'm not sure how to get these or even if there are
any. They should be in the manual for Geneva.
> Oh, and objc_edit code would be nice too. :) Although, I don't believe
> in sending keypresses to anything other than the top window, so
> replacement objc_edit wouldn't be very useful. <shrug>
Except when you receive a click on a text field in an untopped window, when you
do objc_edit(ED_INIT...) the cursor is not drawn clipped.
???:
> Since it removed the desktop form, why do the icons still show up?
If the background of the desktop installed is transparent, or too small, the
icons of other applications show through. But they don't work.
> Sure, some things would be nice as configurable, but we know more about
> userinterfaces than the average user (for the most part), and we can
:)
> figure out what works well as compared to what is absolutely moronic.
Yes, I don't think we should require user configuration apart from for keyboard
shortcuts. If anyone wants to deviate from the standard by, for example, using
their own fileselector, click to type in background windows or any other thing,
by all means have it as a user option (and certainly don't force it on the
user). But don't say we all have to have it as an option, it's just your
non-standard program.
Michel:
[Atari compendium]
> These are textual errors, but it makes me question what -other- errors
> there might be. Is the code listed reliable? Are the suggested
The bindings are very unreliable. For a high level programmer it's mostly
pretty good although there are a few mistakes.
> because the upgrade price is actually the same price (or more expensive)
> then the price they could purchase it for. The feeling seems to be that
> the upgrade should have been offered as a text file or addendum-booklet
> to registered owners.
There is an addendum with the error fixes in, but not the new material.
Simon:
>> The file should go on the root directory of the boot drive. All this has
>> been discussed but not finalised yet.
>
> I'd like to see something similar to the system folder on the Mac. Yes we
> can make it fall back to the root directory for those who don't know or
> care about neatness, but to be frank I'm growing weary of a cluttered C:\
> directory.
Personally I use the whole of my drive C: as a system folder with everything
else on D: (which is minixfs). You don't really want it to be exactly like the
mac's system folder, which AIUI is not really a folder at all but actually a
way to look inside the system image file.
But I agree with your proposal, it would make sense to do this. How about
putting everything in the multitos folder though?