i was able tu run zelda classic v2.10 dos with an aplicattion called dosbox turbo and i configured all the things , and the best i did, the game runs frameskip 18/60 it can be played but it still slow, on the computer it run 60/60.
i saw on the net that is posible to play 90's to 1999 windows games installing windows 98 operating system with dosbox turbo aplication and some games are heavier than zelda classic i guess, i dont understand why zelda classic is so dificult to run because it dont have much graphics like other 90's to 1999 windows games
I'm not one of the developers, so I can;t give you in-depth, technical details. From what I understand, the Allegro (Atari Low-LEvel Game ROutines) libraries (v4, for ZC, IIRC) are the primary problem, and they are a prime dependency for ZC, as it stands today. I don;t know if ZC was ever shifted to Allegro 5,, and Allegro is rather CPU heavy in some aspects, and unfortunately, ZC still has many flaws that make both it, and the Allegro.cc libraries un-optimised. As to the Allegro name, I'm going. to flag that as a 'LISA: Let's Invent Some Acronym'.
Allegro is an additional HAL, running above, and not in direct cooperation with the OS API, which is what makes code so portable. It is less independent than a JAVA RTE, but it is still an interpreter, so it will take CPU cycles, to translate code from the programme->Allegro->local API->Actual functions.
This is worsened on Linux, based on the video drivers in use, or at least, that is the feeling that I get. Gnome runs slowly on my systems in general, so I suspect the integrated Intel chipset drivers, of poor rendering capabilities, and possible, that they are not using all of the facilities available to them. (I haven''t tried to run ZC purely from a shell login.)
As I say, to my knowledge, ZC Uses Allegro v4, which is not available for Android or iOS, although Allegro v5 is. I have no idea how difficult it would be to shift ZC to Allegro v5, or if it is already using the latest stable release. I suspect not, but as I said, I'm not one of the developers. Allegro v4 does have some unofficial ports, making it possible to build ZC on other platforms, mostly those that no-one uses; and it should be possible to do system-specific builds, optimised for specific CPUs, or GPUs (someone may want to clarify this), but without the ZC sources, that is a technological cul-de-sac.
The real question here is, dhow many people would even run ZC on a phone, if it was available? The ZC community has actually decreased in size over time, due to lengthy delays in the release of 2.5; which was originally planned for what? 2005-ish? It's one of those things, where you discover it, and immerse yourself in the culture, and then discuss it with mates, only to find they've never heard of it. Porting it to any other platform is a huge amount of work, and the number of people doing basic maintenance, and responsible for the latest version, are so few, that you can count them on one hand.
Asking for ports to systems, where using ZC requires add-ons, hacks, or other obscure items, is simply too much. We all have a laundry list of 'wish' features for ZC, that have to generally take priority over porting it to a system, that may not even be around, by the time that a port is complete. If you wanted to port ZC to iOS, as an example, and you knew that it would take four years to do it: Why would you bother?
OS will be significantly different by the time that you finish the port, and may no longer even be a mainstream OS by the time that the port is ready. You are better off, as a programmer, to wait for technology to pace up, and catch up; so that mobile devices are as powerful at the expected date of completion, comparable to portable systems, such as laptops, are at present. If you are a small team of developers, working for a non-profit goal, then you have to prioritise features, and feature-requests, that the majority of your users will want.
I just don't see ZC on phone architecture working well, even if it was available. I think it might tax the Pandora, and well, mobile phones have less power than that. They don't have to be as underpowered as they are, but no manufacturer would be willing to produce a mobile device that is comparable to a laptop, given the inherent cost involved. That is why the design that i created in 2000-2002 was ultimately rejected. The manufacturing cost, per unit, at the time, was in excess of £2,500. With a 40% margin, which is the base minimum for a vertical market device, the expected retail was £3500, or $6,000 to $7,000.
That what you would expect to pay today, for a hand-held device with laptop to desktop power, assuming that the components are available, or possible to produce with the present tooling, wafer sizes/pitch, etc..
The coming Linux-based phones, may be able to run ZC but not at much of a speed improvement, as they are simply not designed to run desktop software, will lack critical dependencies that you will need to re-add, and be generally devised to run phone-oriented software, which is in its nature, not graphics intensive; and it is the graphical conversion between Allegro, the local APIs, and the GPU, that slows ZC down on various platforms.
I should really buy a Pandora, just because it is such an interesting device, and to test it with things like ZC, to gauge its usability as a pocket computer. I would love a monocular viewer for something like that. (You can keep the Google Glass: I only desire its display.)