[ALLY] Re: Freeze Snapshot?
From: agem:crosswinds at: 9-Nov-2000 8:27
[petr--krenzelok--trz--cz] wrote on 9-Nov-2000/19:06:58+1:00
> > Hi Elan,
> >
> > Stable enough, Yes.
> > It is getting very close to release, I think it is only one (or maybe 2)
> > revisions away from being releasable. It is just a matter of refining what
> > is there.
> >
> > ToDo...
> > 1. Changes to ports to allow better backround asychronous
> > events/downloading.
>
> asynchronous, tasking, modular, whatever :-) agreed ...
>
> > 2. Minor cleanup in VID. Expansion of variety of VID gadgets.
> > (some of this can be done by the REBOL community)
> > 3. Consolidation of the various caching functions (
> > read-thru, read-via, load-image, load-thru, read-net)
> > 4. As mentioned by Petr, stay on top windows and real modal popups.
> > 5. Scaling & Font Rotation: Planned features, that could be dropped until
> > v1.1
>
> I am fine with fonts. However - I very rare use default setting for fonts.
> Do you think it is really good to have shadowed fonts by default? I don't
> like them because of worse readability ...
>
> > My Wishlist,
> > 1. A way to create native Menus in the Windows.
>
> Not sure it fits View philosophy - we have very core architecture - faces -
> and as I suggested, we could build menu plug-in, component, module, widget,
> call it whatever. But looking at matlab, their figures have possibility to
> have default os menus .... after all - we use default os window bars, right?
>
> > 2. DB connectivity...
>
> Hmm, if we are talking view - I will not more agree to such arguments :-)
> For me there is only one way for Rebol, or I would probably not use it ....
> e.g. if RT would not show me a signal having components, modularity, etc. is
> the way to go, REBOL would not be technology worth enough for me, as I know
> what Carl gave us by creating AmigaOS and what he has promissed to us - to
> bring us better system ....
>
> So, DB connectivity is case for separate component. Once we have dynamic
> components, you will be probably able to buy just what you need - I e.g.
> don't need /ODBC, /Oracle so why should I pay for it? It's also very
> limiting factor for RT itself. If someone has already Oracle, he/she can
> probably afford to pay some XXX or even XXXX USD for /Oracle. I can very
> easily imagine buying /Shell for 30 e.g. and /Library for 50 - 100 USD ...
>
> current state of things is something I can't understand - do you know
> language where you don't have shell by default, or you have to use special
> version to just call libraries? One and only core.exe and others as
> components (libraries) - it works that way in computing for ages ...
>
> But looking at practical usability, I would better welcome things as
> load/files, load/dirs, load/dir-tree, find/deep, native drawing lines,
> boxes, circles, ovals, filling the areas as effects to faces. They are small
> additions but can help so much to speed certain functionality. Also
> operations upon arrays is so slow. Why don't we have basic matrice
> operations? I've heard many things can be solved by matrices - e.g. own
> effects to faces ...
> Should I add, change of mouse pointer? Mouse is still the same thru all the
> states, it's sometimes confusing ...
>
> btw: why 'load doesn't allow us to load images from strings? Let's say I
> have packed image, will 'decompress it, but can't load it into image
> datatype. The only way seems to be to save it to file first ...
>
maybe you can use [load to binary!] ?
i use enbase/debase without compression for pics.
another: you can fake a ram-floppy by a little protocol.
this gives you binary load from strings :)
could also solve this 'load-thru ..versions, something like
[load thru-net://my.jpg].
Volker