Typing in a blog on the dire state of software only to find out that Niklaus Wirth already wrote most of it in 1995. Some things have changed since then, but mostly his article is even more true today (in new ways) than it was in 1995. https://cr.yp.to/bib/1995/wirth.pdf
Conversation
Notices
-
Embed this notice
bert hubert 🇺🇦🇪🇺 (bert_hubert@fosstodon.org)'s status on Monday, 15-Jan-2024 02:37:33 JST bert hubert 🇺🇦🇪🇺 -
Embed this notice
phiofx (phiofx@hachyderm.io)'s status on Monday, 15-Jan-2024 02:37:32 JST phiofx @bert_hubert software should be more like mathematics. Discrete solutions to well defined problems that once written dont *ever* have to be rewritten, just composed
An occasional refactoring of the edifice is ok, after all we keep pushing the scope of software
The problem is that software conveys power over others (the digitally illiterate) and power thrives in complexity. As every bureaucracy can attest
Would be interesting to calculate the churn or efficiency ratio of the total code corpus
-
Embed this notice
bert hubert 🇺🇦🇪🇺 (bert_hubert@fosstodon.org)'s status on Monday, 15-Jan-2024 02:37:33 JST bert hubert 🇺🇦🇪🇺 One specific thing that changed is that Wirth blamed the software size explosion on "too many features". In his view both developers and users should learn not to do that (still true). These days software has mostly exploded because we ship a whole ass full-fledged browser to create an app that opens a garage door if you press a button.
-
Embed this notice