I’m just curios why “neso” was abandoned?! The code is about 400 lines of code so I’d assume its not too hard to maintain.
Mainly lack of interest, if I remember correctly.
Packaging was lagging behind and support was OS dependent.
Security of an ERP as a desktop application was an open topic too.
Another factor is that nowadays you can run the server on any desktop machine using the docker images for example. So there are no need of maitaining such lines of code.
Well , to add to the arguments above:
IIRC another strong reason was, that neso worked out-of-box with an sqlite database, thus with problematic/no multiuser support. And if someone after some trial with neso wanted to change to a postgresql setup this migration was not supported.
Thanks for the insights. Lack of interest is the most convincing one for me.
I’m still a bit curious about these arguments:
worked … with an sqlite database, thus with problematic/no multiuser support
AFAIK, sqlite is a full-featured DBMS. Why should it not allow multiuser?
docker images
Oh, well, getting docker to run on e.g. a Windows machine is not pracical for non-IT persons.
Sqlite does allow multiuser, but it does file locking for writes. And as the whole database is a file concurrent writes queue up.
I didn’t know this. Does it work ? and is psql → sqlite possible ?
There are tools to help with such migrations, both ways. I have no experience with any of them, I would never try to use sqlite as a Tryton production database.
I took the chance to search for conversion tools. There seem to be some.