Posts by holgerschurig@emacs.ch
(DIR) Post #AZqf9de7NQdwlpdMu0 by holgerschurig@emacs.ch
2023-09-17T05:32:23Z
0 likes, 0 repeats
@louis You might want to swap Lisp with something powerful, like Nim. Not with a language that was designed with a built in handbrake, like Go (Go was intentional designed for beginner or intermediate programmers).In Nim, you even have Macros that can help you define compile-time DSL if you're so inclined.But if you swap anything out for an intentionally too simple thingy, you might end up frustrated with a high probability.
(DIR) Post #AZqh89vSafvSDIATWS by holgerschurig@emacs.ch
2023-09-17T05:54:33Z
0 likes, 0 repeats
@louis Ich frage mich, woher diese ganze Blackout Angst herkommt. Ich komme aus dem "großen Kanton" und unser Stromnetz ist messbar top-notch (gucke Dir mal auf Wikipedua die SAIDI Werte an: Anzahl von Minuten Stromausfall pro Kunde pro Jahr). Über die Jahre wurde das fogar immer besser, von früher ca. 20 Minuten auf nun 10,7 Minuten. Obwohl Atom- und Kohlekraftwerke ausgeschaltet wurden!Im Prinzip ist CH beim SAIDI ähnlich: mit bei Europas Spitzenreiter bei nur 17 Minuten. Verglichen mit DE ist FR oder BE schon siebenmal schlechter.Und dennoch fabulieren hier in DE seit ca. 4 Jahren Journalisten regelmäßig von Blackouts. Es kommen aber keine, der SAIDI hat sich weiterhin verbessert! Eher umgekehrt ... wenn in FR eine große Menge an Atomkraftwerken abschalten muss (wie vor 1 1/2 Jahren), dann versorgen die Anrainer-Staaten die halt einfach mit. Haben aber die Einwohner kaum mitbekommen, eben weil alles tadellos funktioniert.Wo also ist das Problem, faktisch und messbar, nicht als undefinierbare bohrende Angst?* https://de.m.wikipedia.org/wiki/System_Average_Interruption_Duration_Index erklärt SAIDI und zeigt Werte von diversen Staaten* https://www.entsoe.eu/data/map/ zeigt, das wir kein Insellösungen haben
(DIR) Post #AaKCEMXT9xc6TUPEv2 by holgerschurig@emacs.ch
2023-10-01T11:29:39Z
0 likes, 0 repeats
@mitchmarq42xyz IMHO this package has a) a name that transports a bad connotation (I know, this applies to git, too. But Hyperbole's author isn't Linus Torvalds) and b) they put too many different things into one package. So it won't appeal to the "I want my Emacs lean" people.But then again, I never made an Emacs package, so what do I know ...
(DIR) Post #AaKsGoGdEW2N7vNcNU by holgerschurig@emacs.ch
2023-10-01T19:20:43Z
0 likes, 0 repeats
@mitchmarq42xyz Maybe I'm too german for this. If someone (outside of satire or comedy) uses artificial exaggeration a.k.a. Hyperbole, then this is something bad. Such a person isn't trustworthy. After all, (s)he intentionally decided to talk non-factual.
(DIR) Post #AaLjsQIe5TGrMr8tN2 by holgerschurig@emacs.ch
2023-10-02T05:21:22Z
0 likes, 0 repeats
@freemo @kilroy_was_here In most non-english languages there is no "ham", and therefore no bacon either to be associated with our hobby.In german, we call us "Funkamateur" (Funk = sparc, comes from the original spark transmitters). It's basically like the english radio, just that the german. "Radio" normally means an FM or DAB+ radio. It's purely a receiver.And "Amateur" stems from greek (or latin, unsure), it means "lover of".No pigs to be seen here :-)Even more bad: for weird reasons the term "bacon" has a bad connotation is muslim countries. They think that a pig is something bad, not to be eaten. I would therefore never name a supposedly international organization with "beacon" in its name. With "ham" I could at least argue that this is just the english word and doesn't have anything to do with pig meat.
(DIR) Post #AahZ6A82Xf1lc2lx3o by holgerschurig@emacs.ch
2023-10-12T18:04:13Z
0 likes, 0 repeats
@boilingsteam Uaah, .emacs_backups ... rpthus clutters the home directory. Better use one of the XDG directories (eg. .local/).Here's a list of them: https://wiki.archlinux.org/title/XDG_Base_Directory
(DIR) Post #AcBrPPqLXzMEzspvd2 by holgerschurig@emacs.ch
2023-11-26T06:42:52Z
0 likes, 0 repeats
@mms @iak Maybe you make Bezos less rich by buying a Kindle and then not buying the books from their wallet garden.I hot this idea because often the hardware is made really cheap, probably even subsidized, but the wallet garden products are then overpriced. True for some Inkjet printers, maybe the Kindle.Only some (Apple) can have both the entry to the walled garden and the garden itself overpriced.
(DIR) Post #AcBrjybY106QbCic4m by holgerschurig@emacs.ch
2023-11-26T06:46:36Z
0 likes, 0 repeats
@mitchmarq42xyz That single image on the github landing page doesn't IMHO do a good job at promoting it.
(DIR) Post #AcBsGye6slqkHdeY7s by holgerschurig@emacs.ch
2023-11-26T06:52:33Z
0 likes, 0 repeats
@galdor Keep this post on your TODO list, to repost it slightly changed in 10 years. Just replace "stackoverflow" with "CodeLlama" or whatever LLM new programmers will use then :-)
(DIR) Post #AcL99ZR3fBGFi8F26y by holgerschurig@emacs.ch
2023-11-30T18:14:05Z
0 likes, 0 repeats
@louis Oh, why MySQL?Asking for a friend (tm) that is into PostgreSQL :-)
(DIR) Post #AcL9h5w5qcnW48k9z6 by holgerschurig@emacs.ch
2023-11-30T18:20:09Z
0 likes, 0 repeats
@mms I wouldn't be not so sure here.Almost any gitantic organization (like SAP, Siemens, the army, Microsoft etc) has many different departments run by differently qualified (or fvil) people. This incident you mentioned here is quite bad, if you adk me. But it's still anecdotal. And one cannot conclude a trend.M$ 20 years ago misbehaved week after week. Back then on could look at many different incidents. Weekly. A negative one, for sure. That has quite changed since then.
(DIR) Post #Acj0Mg0rB4JYgGRKz2 by holgerschurig@emacs.ch
2023-12-12T06:28:41Z
0 likes, 0 repeats
@mms Computers are (better: can be) creative devices. Mobile devices are mostly consumption devices.People try different things sometimes, e.g. running Emacs on Android. But my prediction is that this will stay niche.
(DIR) Post #AcsTvbMuCG1YzhiNrU by holgerschurig@emacs.ch
2023-12-16T20:12:20Z
0 likes, 0 repeats
@louis Beside -Wall etc look also into clangd (e.g. with eglot or lsp-mode). It has "clazy" built in, and can do some semantic checking. It's nowhere near as good as Coverity, but hey, it's FOSS.
(DIR) Post #AcsUfATIWyhetbQsXA by holgerschurig@emacs.ch
2023-12-16T20:20:35Z
0 likes, 0 repeats
@louis Here is my .clang-tidy file, which is honored by clangd: https://0x0.st/HYuP.txtAnd I meant "static analyzer", not "semantic checking". Basically, clangd has built-in what "scan-build" can do.So, while editing, I get clang's warning. And when I compile, I use GCC. So two different compilers look over my code, and this is quite helpful.Drawback of clangd: you need a compilation database. But both CMAKE and Meson (which I prefer) create them. If you're on pure makefiles, than you'd need something like https://github.com/rizsotto/BearBest of this: this all works not only with C, but also with C++. I program mostly in "mild C++", without STL or Boost, but with Qt, even non-GUI programs. They have all the containers I ever need, and with a much more eye-pleasing C++ code than idiomatic C++ code with tons of < > and : characters sprinkling your code.
(DIR) Post #AcwNcWFx9nd0Xoov4q by holgerschurig@emacs.ch
2023-12-18T16:40:49Z
0 likes, 0 repeats
@mykhaylo @mms I once has a mailbox, self-writtfn jn Turbo-Pascal called INFSYS. First on Apple II zu ndrr CP/M, later on MSDOS.It had a command line interface similar to the commercial mailboxes from GEONET. And later I brought it into Fidonet, with the help of BinkleyTerm. Forgot my node ID, it's been so long. Maybe 2:233/23? Something similar? Something entirely different?
(DIR) Post #Ad0ihk8IpESYKAxNzM by holgerschurig@emacs.ch
2023-12-20T19:35:36Z
0 likes, 0 repeats
@mms Yes and no.We cannot answer, because we don't know what you mean with harmful.We don't even know the scope of your questions. Reddit has thousands pf subreddits... some ate probably certainly harmful. Others? Not so much. Similarly Discord. Many "severs" (I haze the misuse of that word, both here and in "serverless") ate full of info and helpful people. Others waste your time ... is that already harmful?Ultimately, I think that boiling down complex scenarios to a boolean ("yes" or "no") is usually a futile approach. And to be avoided.
(DIR) Post #AdeLo7QXS5lvn3qWUC by holgerschurig@emacs.ch
2024-01-08T22:27:30Z
0 likes, 0 repeats
@mms I do nothing in "eMacs", but many things in Emacs :-)But crazy things? I dunno if anything here is crazy:Writing codemanaging projectswriting 200+ page long manual for a software I wrote at work in LaTeXsometimes reading/writing mail (perhsps crazy: I have code to directly apply a patch from a Linux mailing list to a Linux source)sometimes reading/writing mastodonsometimes doing IRCcreating my blog postsusing magitfixing my bad English with gptel.el+Olama+Mistral and patience (since I'm not into graphics and don't have a proper 3D card)
(DIR) Post #AhwyYenBoavouvHww4 by holgerschurig@emacs.ch
2024-05-15T21:43:14Z
0 likes, 0 repeats
@galdor why would you do that in 2009?One reason could be interop. You could not call many Linux (or BSD) system calls without pointers. And so you still find them in D, Nim, Rust ...If you really don't have them, then you always need some C intermediate later, like Java JNI.
(DIR) Post #AhwyYiCp7kDdVIIxo8 by holgerschurig@emacs.ch
2024-05-16T05:13:26Z
0 likes, 0 repeats
@galdor Using integers for this sounds very lispy: a crude hack that then survives decades. Not because it is good, but only because it is good enough.I mean, which other programming language uses assembler mnemonics from a long died of computer system today? car / cdr ...
(DIR) Post #AhwyYkh1sPZLDYS7AO by holgerschurig@emacs.ch
2024-05-16T10:41:24Z
1 likes, 1 repeats
@galdor You do realize that pointers are memory addresses, which are simply integersSure. But that's a very incomplete picture. Not outright wrong, but incomplete.Today pointers in C are almost never "void *ptr". But instead "struct foo *ptr". So you get some type checking.So you might frown on pointers in your high-level language ... but when you emulate pointers, you fall back to the worst possible implementation. Some integers that cannot be type-checked at all.Linux for example all the time invents new syscalls and ioctl(). Many more than e.g. the GNU C library covers. Many of these use pointers. And if I write a function prototype for them, then I'd of course use typed pointers. It makes little sense to not use this to prevent --- type checking helps oneself from shooting into the foot.So, if you frown about pointers --- which I can fully understand! --- at least provide a better alternative. Not a worse one.But maybe my view is too much a systems programmer view, I value languages mostly from a systems programming point of view (where I e.g. like Nim, D or Rust, but not so much Python, Lua, and anyone of the 100s of Lisps). So I might be preoccupied!