Problem with software is bitrot. Web app lasts about a year, iOS app 2–4 years
1–10 of 15 posts
Re: Problem with software is bitrot. Web app lasts about a year, iOS app 2–4 years
#2C programs, on the other hand, last for decades. You can take a C program written 20 years ago (using POSIX API) and it will still run just fine today.
Re: Problem with software is bitrot. Web app lasts about a year, iOS app 2–4 years
#3I was thinking that this was not quite correct as I saw the problem. A vanilla html-css-js page from 10 or 15 years ago will probably run just fine in a modern browser (though it will look its age). The loads of external dependancies is what will doom web projects.
A tweet down-thread coined a word for this:
Ry @RyanMorey I think we need a different term for this than bitrot. It’s not data corruption breaking these, but like compatibility corruption? PlatformRot
Re: Problem with software is bitrot. Web app lasts about a year, iOS app 2–4 years
#4I was thinking that this was not quite correct as I saw the problem. A vanilla html-css-js page from 10 or 15 years ago will probably run just fine in a modern browser (though it will look its age). The loads of external dependancies is what will doom web projects. A tweet down-thread coined a word for this: Ry @RyanMorey I think we need a different term for this than bitrot. It’s not data corruption breaking these,…
Platform Drift?
Re: Problem with software is bitrot. Web app lasts about a year, iOS app 2–4 years
#5I've been shocked to discover the number of MVP shops that exist and sell broken shit.
I can't stand the npm ecosystem because every project can't just sit there. When I return to it, something is broken and I can only find out by debugging it via runtime.
Thus is why I'm building my own shit from top to bottom as a vertical investment.
Re: Problem with software is bitrot. Web app lasts about a year, iOS app 2–4 years
#6C programs, on the other hand, last for decades. You can take a C program written 20 years ago (using POSIX API) and it will still run just fine today.
That's not necessarily true. Bitrot and bit flipping is pernicious and largely silent, while it affects source code less than object code, its still an issue regardless of the language medium.
Re: Problem with software is bitrot. Web app lasts about a year, iOS app 2–4 years
#7C programs, on the other hand, last for decades. You can take a C program written 20 years ago (using POSIX API) and it will still run just fine today.
As long as it’s on a VM; and not exposed to Bluetooth, the internet, or any external devices. This may be true.
I miss my VIC-20 too, but those days are over.
Re: Problem with software is bitrot. Web app lasts about a year, iOS app 2–4 years
#8C programs, on the other hand, last for decades. You can take a C program written 20 years ago (using POSIX API) and it will still run just fine today.
That's not necessarily true. Bitrot and bit flipping is pernicious and largely silent, while it affects source code less than object code, its still an issue regardless of the language medium.
I think you're using the strict technical definition of bitrot, but bitrot as used here is the more metaphorical "the OS/libraries/browser has removed some APIs and now the program won't work".
Re: Problem with software is bitrot. Web app lasts about a year, iOS app 2–4 years
#9Earlier quoted context omitted.
That's not necessarily true. Bitrot and bit flipping is pernicious and largely silent, while it affects source code less than object code, its still an issue regardless of the language medium.
I think you're using the strict technical definition of bitrot, but bitrot as used here is the more metaphorical "the OS/libraries/browser has removed some APIs and now the program won't work".
I use the classical and only legitimate definition of bitrot.
Any other instance using that term is merely a corruption of language meant to mislead and confuse, and make communication harder if not impossible.
Its hard enough communicating technical issues without people corrupting the language used to do it. Bitrot has nothing to do with program design and deprecation.
Re: Problem with software is bitrot. Web app lasts about a year, iOS app 2–4 years
#10Earlier quoted context omitted.
I think you're using the strict technical definition of bitrot, but bitrot as used here is the more metaphorical "the OS/libraries/browser has removed some APIs and now the program won't work".
I use the classical and only legitimate definition of bitrot. Any other instance using that term is merely a corruption of language meant to mislead and confuse, and make communication harder if not impossible. Its hard enough communicating technical issues without people corrupting the language used to do it. Bitrot has nothing to do with program design and deprecation.
You should only speak Lojban then. Human language always drifts in terms of word definition. Or do you mean classical in terms of 'part of the axial period of Hellenistic civilization'?