From Steven Guikal to ~sircmpwn/hare-dev
On Sun Feb 20, 2022 at 4:19 AM EST, Alexey Yerin wrote: > Signed-off-by: Alexey Yerin <yyp@disroot.org> > --- > Overall it should work fine, but I have a few doubts about > whether percent-decoding should work on host=(ip::addr) like > http://%31.2.3.4 -> http://1.2.3.4 Quick testing (http://%3127.0.0.1) with Firefox, Chromium, curl, and Python's requests and urllib show they all percent decode the host even if it is an IP address. Thanks, Steven
From Steven Guikal to ~sircmpwn/hare-dev
Hey, I'm the one who originally implemented the base64 encoding and I based my decision of not allowing improper padding (mentioned in the docs) or invalid characters (like newlines) on the Security Considerations described in the RFC: https://datatracker.ietf.org/doc/html/rfc4648#section-12 Essentially, loosely allowing these could covertly send data across applications and as such, it should be handled explicitly. Thanks, Steven
From Steven Guikal to ~sircmpwn/hare-dev
On Wed Dec 1, 2021 at 4:12 AM EST, Drew DeVault wrote: > It works for me. There is no tags.h, do you have a dirty working > directory? Yep, disregard the patch.
From Steven Guikal to ~sircmpwn/hare-dev
On Wed Dec 1, 2021 at 4:09 AM EST, Drew DeVault wrote:
> It should work without the build directory. Does it not?
`./configure` works but `make` fails:
make: *** No rule to make target 'include/tags.h', needed by 'src/check.o'. Stop.
From Steven Guikal to ~sircmpwn/hare-dev
Signed-off-by: Steven Guikal <void@fluix.one> --- content/installation.md | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/content/installation.md b/content/installation.md index dfbf819..03494f2 100644 --- a/content/installation.md +++ b/content/installation.md @@ -29,8 +29,10 @@ Bootstrapping Hare only takes a few minutes. #### Building the bootstrap compiler 1. Obtain [the bootstrap compiler source code](https://git.sr.ht/~sircmpwn/harec) 2. `./configure`[message trimmed]
From Steven Guikal to ~fluix/public-inbox
Oh man, sorry about that. Thanks for the patch. To git.sr.ht:~fluix/fluix.one fa04e05..57c51d2 master -> master
From Steven Guikal to ~arivigo/scalc-devel
On Fri Aug 27, 2021 at 3:38 PM EDT, Ariadna Vigo wrote: > On Fri, Aug 27, 2021 at 03:09:30PM -0400, Steven Guikal wrote: > > Yep, that makes sense and is totally fair. Ever since doing any kind of > > math at school I've had a scientific calculator near me as my "pocket > > calculator" and having that kind of precision is really appreciated. I > > haven't found anything to do that on a computer, but you're probably > > right that this is rather complex to implement and would require > > significant changes (basically a rewrite) to scalc, so a separate > > project is better. > > I know this *will* sound strange coming from me, but for a project like > that Lisp is the way I'd go. Most Lisps compile to native nowadays and > bring this kind of stuff out-of-the-box. In particular, I'd use either > Chicken Scheme or good ol' SBCL Common Lisp. Scheme has the huge
From Steven Guikal to ~arivigo/scalc-devel
On Fri Aug 27, 2021 at 9:45 AM EDT, Ariadna Vigo wrote: > I see where you're coming from, but this runs in direct opposition to > what scalc is meant to work like, i.e. a pocket calculator. > > I mean, we could have some sort of switch that is able to alternate > between representations? But is it really worth? I don't have a > position on this. I may play around with some code to see if it's > practical/worth it... but I seriously need to read code for this and > the broader symbolic algebra features you were proposing (if I > understood them correctly from other resources,) because it would > totally be a disruptive change for the current codebase (for better or > worse... I know the current code is very 0.x-ish.) > > But I definitely don't want any feature creep in scalc. I'm way more
From Steven Guikal to ~poldi1405/comments
On Thu Aug 26, 2021 at 6:50 PM EDT, Steven Guikal wrote: > > I think the correct diagram would have part of Libre Software outside of > > the Freeware bubble. > > Correction: Part of Libre Software should also be outside of Open Source Whoops nevermind. I meant to say that not all Open Souce is Libre Software and you already show that. The rest still stands, I think.
From Steven Guikal to ~poldi1405/comments
> I think the correct diagram would have part of Libre Software outside of > the Freeware bubble. Correction: Part of Libre Software should also be outside of Open Source