Buildroot is, like almost everything on Linux world, hard to understand for any *nix newbie.
Hopefully some very useful slides exist on bootlin (ex Free electrons)
So, I now have a very basic distro with ssh enabled.
I’m about to finish ucon64 package so I ‘ll get my own dev kit ready this week.
Everdrive support will follow but propably later, after the test and tweak period.
While I no longer post about it, my dev kit is still in progress
I dropped TinyCore, which required too much work every time I wanted to add something not already avaible through the TC packet manager.
I’m now building my own Linux distro, thanks to build root.
It’s actually on hold since my LCD with SCART died so no way to use my Genny , and so, no need for a devkit.
Hopefully, I plan to finish my GBS8200 combo this year, based on SmokeMonster’s work, and then resume it.
So, nothing new expected before next year!
A thin client is not a mini sized PC.
A thin client comes with a (very) limited OS, like WinCE or a dedicated Linux distro, which main feature is to load a virtual system from a network server.
Let’s see it like a window to another OS installed on a dedicated hardware somewhere on your network.
If this virtual system and this main network server don’t exist, a thin client is NOTHING, just a friend of dust.
So will I show you how to setup a virtual system on a dedicated server, just for genny development ?
Of course no !
A thin client is not a full featured PC, I won’t be able to watch youtube nor play games, not even with Fusion I thing
BUT it comes with the minimal features I need :
Side note : the only thing missing on thin client is GPIO, else it will be easier to learn electronic with a thin client than with an Arduino or a Teensy
How did I select my thin client, heart of my DevKit ?
I just looked for thin clients on eBay or similar, looking for a very cheap one (the expensive ones are overkill for what I need to do).
Everytimes I found one, I opened the website of David Parkinson, a guy who, like me, share as much information as possible on the subject he likes the most : thin clients.
So, if he posted a test on the thin client I found, I checked the available options and how hard it was for him to make a linux to work on it.
Not listed on his website ? not for me !
He needed to make some bios or hardware update ? hidden cost and lost time for sure so not for me !
USB boot, lot of ports, easy to update ? let place a bid !
I was not looking for the “best” but for one which suit my need at a very low price.
In the end, I bought 3 thin clients (total price < 40€), in case of failure, wrong selection or hardware problems.
I selected the Neoware CA21 as a perfect match for my first try, since Neoware was perhaps the best thin client solution some years ago.
It also explain why HP bought it !
See you in some days for more details on my success or failure to boot it with all the software side of the project ready to use.
A lot of people know how much i’m sad so few homebrew games were released these last 15 years.
Of course, we saw some gems (Hello Spain!) but if you count how much (real) games were made, it’s not a lot while SGDK and SpritesMind are alive from some many years !
My goal was to help people to dev for Genny.
Let face it : it was a half failure because “help” included informations, community, tools and code.
Informations and forum are, I think, a success.
Tools, like I said took me a LOT of times and, apart KMod, were a help for perhaps 2 or 3 people.
I (recently) stopped trying to understand why you’re all trying to make your own tools, your own kit, your own ide, your own framework , if not your own compiler or assembler.
It’s not a total failure, because I’m very happy if I was helpful for these 2 or 3 homebrewers but was it the right way to help them ?
My demo/games Code is too old to be used and the few tutorials I made for SGDK were no longer usable 2 months later, after a major SGDK release.
Perhaps it was a mistake from side to say in the first place “He! Guys! make some great games, I’ll give you all you need !”
But now, I’ll stop losing my times.
Tools are still available, still updated and still open source.
But forget about tools made as much as options or liberty so they can be used by anyone.
I’m updated them to suit MY needs.
You use them ? Great ! Continue ! and PM me if you ever find a bug.
You need a new feature ? Fork my tool, update it and ask for merge request….or PM me and if I find it useful, perhaps, I’ll add it
So, what will I do now ?
Well, is it not clear enough ? Do what I was expecting other people to do : RELEASE GAMES !
To do so, I need a valid process to be sure my games work on real hardware and not emulators only.
And so, start my posts serie about making a networked dev kit using
…because help with as much informations as possible is a part of myself, mainly because I’m what I am now thanks to people who share what they know/found all other the net
I didn’t totally finished floppy support but I’m pretty sure of one thing fun : JSI used an already existing lib.
They didn’t fully updated it.
They cleaned some stuff (like the BPB table) but not the format data (6 availables while 4 selectable), because code is based on 6 bytes per data
and, according the format defined, at least one is for 5 1/4 floppy.
It was not coded for SMD because a minimun file size is 512k … won’t fit on a 360k 5 1/4 floppy !
I think it also why the code is a total mess LOL
Anyway, another thing which make it hard to re-code is the fact they use the same function to write and read.
It’s a bit value which define the read or write code part of each functions.
I didn’t notice it on the first pass so I missed a lot of things !
I ‘m finally working on the Floppy support…
It will help me to debug the floppy drive on my broken SMD
It’s really hard to understand all of the features of the floppy controller.
Even more hard since JSI re-use memory a lot !
Floppy dedicated vars are so shared with others… very difficult to handle this with IDA !
I worked hard today… the comIO part of the open source SMD bios is done !
UI is almost finished too
The hardest part is the floppy support…I don’t know if I’ll code it yet since
1/ I’m unable to test it
2/ I plan to add USB support throught the comIO port
I’ll finish UI then I’ll move to my SMD fix since I need to test my bios first .
I’m also waiting for some components for a DIY Floppy Drive PSU 😉
Today, I started to port the bios 3.3c to WLA DX.
So yes, I hope I’ll be able to release an open source bios some day
VDP stuff is done
I’m so happy! ucon64 works on Damn Small Linux !
You’ll need to find the libusb to make it works : using the myDSL browser, you’ll be able to find a package which includes it.
I installed OWFS, because it’s the smallest one 😉
I’m now looking for a thin client cheap but powerfull enougth to get a full SMD dedicated PC
Thanks to Charles, I finished my PSU schematic
(greyed area isn’t available in my own disk drive)
I received the tools to fix it so I’ll be back on development soon … perhaps ?
