It is amazing how, with so limited resources, people were able to write decent UI but today, with GB of RAM and TB of storage they are not. But back then there was a bit of decent competition.
Nice write up, but the author is wrong about the GEOS version. Geodes (GEOS executables, drivers, libraries--pretty much anything that "runs") have their own version information (version, release, protocol) in the file header.
The kernel (GEOS.GEO) in BrotherWorks is version 3.0 4-84, protocol 654.054. The archive.org blurb is correct that BrotherWorks is therefore GEOS 3.x.
The 2.0 Ensemble release kernel was 2.0 24-1, proto 654.001. NDO 3's kernel was 3.1 33-1, proto 654.060.
Oddly, the article looks at the GEOS.STR (a file of strings that the DOS start-up code uses for user messages and concludes that an unchanged strings file "proves" the GEOS version.
(author) No, I'm sticking to this. If it's 3.0, then why are references to 2.0 still present? It reads more like that's what they wanted to call it, not because there's any substantial technical difference.
7 comments
[ 3.2 ms ] story [ 29.9 ms ] threadThe kernel (GEOS.GEO) in BrotherWorks is version 3.0 4-84, protocol 654.054. The archive.org blurb is correct that BrotherWorks is therefore GEOS 3.x.
The 2.0 Ensemble release kernel was 2.0 24-1, proto 654.001. NDO 3's kernel was 3.1 33-1, proto 654.060.
Oddly, the article looks at the GEOS.STR (a file of strings that the DOS start-up code uses for user messages and concludes that an unchanged strings file "proves" the GEOS version.