Does anybody know of a library that allows you to read from the command line without echo? The obvious example is to read a password.

@lexi.lambda i recommend abstracting their definitions so that they take an identifier for quote-syntax

@lexi.lambda Looking back at your mailing list post, I’m surprised by the '(local module) result. I thought that SRFI 72’s “evaluation of any nested, unquoted syntax or quasisyntax forms counts as part of the evaluation of an enclosing quasisyntax” would imply '(local local).

@mflatt I’ve now seen this error in both the Windows and Mac CI tests: https://ci.appveyor.com/project/plt/racket/build/1.0.3058#L3576

((chaperones) (#f #t (threads-finished #f)))

@jeapostrophe I’d like to remove or fix the ramunk, racket, formica, and rclp packages, none of which have a working package source

@mflatt another odd error: Section(file) ERROR: tcp-connect: connection failed address: ::1 port number: 55598 system error: Operation timed out; errno=60 (from https://travis-ci.org/racket/racket/jobs/353322682)

@jeapostrophe sorry, that wasn’t 100% accurate — they all error when trying to get the checksum, but only racket and rclp are missing the source entirely

formica and ramunk error when getting the checksum in ways that I think the pkg server should handle

@pocmatos: When I’ve needed to do that, I’ve just called out to stty. Not ideal, I agree.

@samth I deleted racket and rclp

We should probably email the authors

Do you want to do that, or should I?

@mflatt I think you are correct; I overlooked that last bit, and my implementation diverges from SRFI 72 there. That said, I’m not immediately sure which behavior I think is better. On the one hand, the SRFI 72 behavior means that program must be rearranged to lift the nested #'x into a separate binding in order to get the behavior I was aiming for, which avoids the confusion of the nested #'x being lexically inside the syntax-quoted let but not being captured by it. On the other hand, as the SRFI itself points out, it yields the following behavior: (bound-identifier=? (syntax x) (syntax x)) ; ==> #f (quasisyntax #,(bound-identifier=? (syntax x) (syntax x))) ; ==> #t …which I can’t help but feel is confusing, since there is no binding of x “capturing” the uses of syntax (unlike in the example I posted to the mailing list). I’ll have to think about the ramifications of including or not including that rule.

I’m looking at the net/imap package and wondering if I’m not understanding something or if there’s some abstraction over the net/* packages I don’t know yet… there seems to be no function to represent the SEARCH command for imap, nor is there a generalized function for sending more raw commands to the imap server. SEARCH seems like a really basic idea, so I’m guessing I’m missing something.

digging in… internally there is imap-send which isn’t exported and I’ve verified that SEARCH isn’t anywhere within.

Is there a way to get my hands on imap-send?

Also, it really should be exported (or the full IMAP protocol should be covered—it isn’t a moving target so that should be easy to do and maintain)

@samth I think I fixed the threads-finished test

@zenspider you might look at the sirmail code for uses of imap

@mflatt great

general question tho: is it possible to dig in and grab a non-exported function?

in this case, it’d make it easier for me to submit patches back

I don’t know that the SirMail code is any use; the suggested functionality is just missing, and adding it would be great. You can get an unexpected function by using module->namespace, but don’t do that. :slightly_smiling_face:

(ie, I won’t have to have a full build to extend this—yet)

hrm… I don’t know how I’m supposed to get ahold of that function. module->namespace is defined with:

> New provide declarations are not allowed.

(require net/imap)
(eval 'imap-send (module->namespace 'net/imap))

derp

@mflatt tangentially related: what’s the underlying reason (module->namespace 'mod) doesn’t automatically require 'mod for you? I don’t have a good enough handle on modules and the runtime to understand that.

There’s probably no good reason

huh

thank you very much for writing the linklet docs by the way, I found them very helpful

heh

@mflatt thank you. that got me back on track. I should be able to use that and wrap up search

@mflatt What UI framework does Racket’s Windows GUI API use?

I know os x uses cocoa and linux is gtk+, but the windows folder just seems to say win32.

Yes, plain native win32.

Ah, okay. thanks.