Whole-known-network
"old systems were often designed by people with great taste"<br><br>I think we see eye to eye on that.<br><br>I find I am writing way too much in my replies though, so I'll spare you my thought rambles and keep whatever else I had been scrawling about The Egyptian Lover and original 808s as being a parallel non contemporary/retro split [though in musical composition rather than software dev despite folks "programming" drum machines in 1980s terminology] and such for another time and place. ;)<br><br>Maybe I'll wrap my head around the Zig llvm-mos folks someday, but my personal interests are still with the Hu6280 llvm-mos support and I haven't really done much with that knowledge.<br>
<p><span class="h-card" translate="no"><a href="https://notnow.dev/users/zhuowei" class="u-url mention">@<span>zhuowei</span></a></span> Zhuowei</p>
<p><span class="h-card" translate="no"><a href="https://snac.bsd.cafe/teajaygrey" class="u-url mention">@<span>teajaygrey</span></a></span> I don't really do a contemporary/retro split in software dev. I think (systems in particular) software development could greatly benefit from having people with good taste; and I think old systems were often designed by people with great taste. But this by no means excludes software designed today from it--I would say that smoltcp or prjunnamed are both elegant and accessible to new contributors (for sure, that's the feedback I've heard a number of times)</p>
<p><span class="h-card" translate="no"><a href="https://snac.bsd.cafe/teajaygrey" class="u-url mention">@<span>teajaygrey</span></a></span> I like LLVM but I do think that the development model is ... challenging and unsustainable for many would-be contributors. I mean, this isn't a radical statement, the LLVM folks all know it; it is just the result of the political forces driving its design, as usual.</p>
eSIM is Gen Z's CDMA
<p><span class="h-card" translate="no"><a href="https://zia.io/users/AnachronistJohn" class="u-url mention">@<span>AnachronistJohn</span></a></span> <span class="h-card" translate="no"><a href="https://mastodon.social/@mcc" class="u-url mention">@<span>mcc</span></a></span> <span class="h-card" translate="no"><a href="https://social.treehouse.systems/@dysfun" class="u-url mention">@<span>dysfun</span></a></span> i feel like the whole "yeah we can't really make reentrant C work on the 8051" part of it might be a problem</p>
I can only imagine.<br><br>There's a part of me, a wolf that seems under fed or whatever that parable is, that is screaming at me to find some cheap land in a forest with hot springs or a babbling brook, and go live as a hermit with my Amiga 1200 and some newfangled accelerator cards (they're cramming Raspberry Pis into them now or something?!) and just say "no" to all contemporary software dev and go back to the time and place in computing where I was still happy with MODs bumpin in my eardrums.<br><br><span class="h-card"><a href="https://mas.to/users/h0ffman" class="u-url mention">@h0ffman@mas.to</a></span> seems to be closer to that ideal than anything I have managed, at least with keepin Amigas active and happy.<br><br>Exasperatingly, it's the "cheap land in a forest with hot springs" that is causing me more challenges, at least the "cheap" part, or well, anything I can afford at least.<br><br>But maintaining an LLVM fork just seems like more of what I don't want, which is kind of why I am like, "wow, those Zig llvm-mos folks are on <i>something</i> and I don't get it."<br><br>Reminding me of that PeeWee Herman joke:<br><br>"There's a lotta things about me you don't know anything about, Dottie.<br>Things you wouldn't understand.<br>Things you couldn't understand.<br>Things you <i>shouldn't</i> understand."<br><br>Dottie/(me): "I don't understand" (llvm-mos Zig folks)<br>
@whitequark@mastodon.social @dysfun@social.treehouse.systems @mcc@mastodon.social One might think that LLVM is made better by having more targets that aren't x86 and aren't odd edge cases, like m68k. The 8051, on the other hand...
<p><span class="h-card" translate="no"><a href="https://snac.bsd.cafe/teajaygrey" class="u-url mention">@<span>teajaygrey</span></a></span> the problem is mostly that maintaining an LLVM fork in a way that keeps you up with mainline (required to build modern clang, etc) is absurdly expensive in terms of manpower, speaking partly from experience</p>