[Coco] "Who came up with pyDriveWire ?"

didier at aida.org didier at aida.org
Sun Aug 9 15:34:54 EDT 2020


On 09/08/2020 20:49, RETRO Innovations wrote:
> On 8/9/2020 3:37 AM, didier at aida.org wrote:
>> if running an application means spending 3 days trying to modify the 
>> system to get something working
>> there is a serious problem
>
> It may be a serious problem for the person trying to install, but 
> perhaps not for the author:
>
>  * Am I trying to design a tool for lots of people to use, or is my
>    attempt just another option to consider?
>  * Are there lots of people having the same issue?
>  * If so, do I plan to support it or change to accommodate it?
>
I'm only using a standard windows 10 and a standard Linux buster...
I tried drivewire and pydriverwire on both machine and it failed
java drivewire worked once someone told me that I needed java 7

> I stand by my original point.  You are free to choose to install the 
> app (or not, if you can't or find it too hard to get it to do so in 
> your environment, but casting concerns about the choice of language is 
> overstepping.
>
>>
>> btw I also tried to install on linux and got the same result: java 
>> errors and python sending error messages...
>>
>> I run C programs since 1984 and never had any problems...
> I'm sorry, but I think you've just been lucky.  Trying to get old (or 
> even very new) C apps running on Linux or Windows can be just as 
> frustrating (missing libs, GCC version not new enough, too new, libs 
> get installed in the wrong place, because they were hardcoded to be 
> put special places on a different distro, etc.  I'm glad you have had 
> good luck, but I hardly think that's a resounding success for C in 
> general.
it depends on the programmers, I've ported my code since 1984 without 
any serious problem but it is also clear that I saw some
horror with mix of C and C++ with the C using freed objects...
with C you need to be extremely rigorous or you pay the high price

>>
>> I installed that for liber809 and discovered that it was as 
>> unfinished product totally unsuable...
>
> Boise I think is on here and can comment on that.  I think it is 
> indeed a proof of concept, not a finished product.
it's a nice project, I like the idea, but a bit more documentation and 
links to usable resources would have helped
(initially I bought liber809 to try it on others machine (SYM1,Replica 
1, PET...) it was not compatible so finally I made my own board...
as I now have an Atari I wanted to try liber809 on it...)

if you check on forum you can see that liber809 works, nitros9 works 
but... now what ?
>
> Jim
>
>
>



More information about the Coco mailing list