yserial development

[venzen]

This is a great module with many uses in different applications. Have been using it for a few years and interested to know if the developer is still around. His profile has gone quiet and email is out of use.

Developer or anyone else interested in pushing this project ahead?

[rsvp]

hi @venzen -- we just converted the yserial GitHub repo from a Mercurial-hg mirror to a purely git development site today. The change is effective as of v0.60.1

How has yserial been useful in your different applications? Look forward to your feedback.

Thanks very much for expressing your interest.

[venzen]

Firstly, sqlite(3) has great applicability to every-day use cases. Python is a great general purpose language, but with some imposition of good practice via PEP8, so the latter-day dumbing-down of programming languages to cater for developers with a) less experience and b) less good practice, is forgivable and remedied somewhat.

y_serial provides Python with a module that eases the "general purpose" and prototyping use case, by getting all the db layer clutter and SQL out the way (altho' available if needed). The y_serial module allows the developer to preserve application state with just a few calls and can extend to maintaining complex information (especially when contained in Python objects - and Objects are a paradigm).

Shifting data can be warehoused and dealt with when developer/application time permits, or development can proceed straight into using y_serial as a persistent stack/heap. The benefits become especially apparent when needing a platform independent database to handle the state transitions associated with blockchains.

Great module and visionary application of computer science. I have some small select() bugs to report, but other than that y-serial gets the db / persistence layer out of the way to facilitate development of the objects themselves and the functions that will utilize y-serial and its warehousing functions.

[rsvp]

Fantastic write-up @venzen -- thanks! In fact, I would like to include a couple of your phrases in the revised HOWTO, if I may.

Your mention of "prototyping use case" is so useful, I will be adding a parameterized memoization DECORATOR which everyone can simply append above their main function.

Please kindly report those "small select() bugs" -- let's fix 'em: https://git.io/yserial-pr

As for "using y_serial as a persistent stack/heap" is the POP=True and fifo method sufficient for your needs?

Your work on the blockchains may be more important for authenticating objects than the underlying currency. Thank you again.