Re: winforms for GUI?
Mark Hahn <[email protected]> Sun, 8 Aug 2004 10:25:28 -0700
| Newsgroups | gmane.comp.lang.prothon.user |
|---|---|
| Organization | Hahn Creative Applications |
| Message-ID | <[email protected]> |
On Sun, 8 Aug 2004 10:09:53 -0600, Samuel M. Smith wrote: >> >> I didn't really realize until Guido's message that it required a 20 >> megabyte >> download to use .Net (he said 200 megs <grin> but he was corrected on >> comp.lang.python). That is a real drag and confirms some of your >> arguments. >> I can see why it is such a big load though. The library IS gigantic, >> which >> can be a good thing. Maybe it can be trimmed. > That would go counter to 20 years of computer science evolution. > Libraries always get bigger with time. > > Your whole motivation for > using CLR is to get the full featured libraries! > > I suppose you could find a minimal subset that was a mandatory install > and have options for other stuff. I like the way you answer your own questions for me. :-) >> I was planning on making the Prothon download/install/run experience as >> transparently like the Python experience as possible. Now I see I've >> already got the long download time as one problem. I don't know yet >> if I >> will run into other gotcha's. If I don't, then I don't think this is >> necessarily a killer problem. > > It is for embedded applications. I can run Linux and python in a small > RAM > footprint like < 8 Meg but I doubt if you could ever get .Net in such a > footprint unless > you use Windows CE which Microsoft will optimize .Net for Once again you are answering your question for me. >> In other words, as the prothon author and person responsible for the >> Prothon >> experience, I will be assuming responsibility for the hiding of the >> imperial >> forces and the dark star hiding within Prothon. If or when I find out >> this >> cannot be done, then I will admit I was wrong and give up. > > Your willingess to chance course even when it means redoing a bunch of > stuff is one of your more admirable qualities. > IMHO this is another opportunity for you to show your admirable quality. One of the advantages of being light on my feet and being able to switch, is that I can choose when to do it. I think doing the .Net VM now and the native Prothon VM later is the right order of things. I think I will learn a lot from .Net. Also, the .Net will force certain small changes to my language. It is good that I do it first to get those changes right. >> Do you buy this argument? If not, is it because you don't think it is >> possible? If not possible, then why? What do you think I will run >> into >> that will stop me? I don't want to start down this road if it is a >> doomed >> venture. > > Machine to Machine computing, embedded network devices etc etc are the > next big transformational advance in computing. > A tightly coded high level language with support for weightless > threads, stackless etc. obviates the need for an RTOS. It makes > cross platform embedded apps possible.This is a huge advantage. > You can > run an asynchrounous multi-process like application(s) in just one > process, that is the one running python/prothon. This does for embedded > system programming what > VB did for DeskTop applications, that is, make non experts capable of > writing and deploying commercial quality apps. We don't know yet that what you ask for cannot be done with .Net. You are making a lot of assumptions. Jim set out to show .Net could not work for Python and was surprised when he found out what it could do when he did IronPython. > I think a Python library interface to Prothon is a better bet for > getting libraries for Prothon quickly. You really only need the network > library ported to Prothon > the other Python libraries could be run like extensions. > > Don't get sucked into the allure of .Net. If there is an allure then there must be a reason for the allure.