Re: [racket-users] (namespace-variable-value) not working with a struct definition?

On Saturday, 25 June 2016 08:20:30 UTC+10, Matt A. Peerson wrote: > On Saturday, 25 June 2016 08:13:04 UTC+10, Matthew Flatt wrote: > > If you write > > > > (struct point (x y)) > > > > then `point` is bound as syntax that both expands to the `point` > > constructor and provides static

Re: [racket-users] (namespace-variable-value) not working with a struct definition?

On Saturday, 25 June 2016 08:13:04 UTC+10, Matthew Flatt wrote: > If you write > > (struct point (x y)) > > then `point` is bound as syntax that both expands to the `point` > constructor and provides static information about the point` structure > (as used, for example, by `match`). > > You

Re: [racket-users] (namespace-variable-value) not working with a struct definition?

If you write (struct point (x y)) then `point` is bound as syntax that both expands to the `point` constructor and provides static information about the point` structure (as used, for example, by `match`). You could avoid the indirection through syntax, sacrificing static information, by

[racket-users] (namespace-variable-value) not working with a struct definition?

Hi, When I define a struct (named e.g. "point") and use namespace-variable-value in REPL it gives the error: namespace-variable-value: bound to syntax in: point Is this normal? (although even if it is, it'll force me to use class definitions for my structs in my current project). Thanks! --

Re: [racket-users] interested in TR internal error?

> On Jun 24, 2016, at 1:13 PM, 'John Clements' via Racket Users > wrote: > > I was in the middle of editing a TR file, and I got this error: > > ../../racket/racket/share/pkgs/typed-racket-lib/typed-racket/utils/tc-utils.rkt:270:0: > Internal Typechecker Error:

Re: [racket-users] After font change, bounding box too high in Redex picts

On Fri, Jun 24, 2016 at 12:14:39PM -0500, Robby Findler wrote: > Did you trying using basic-text and the other nearby-documented > parameters to add the inset you want? I think those ones might > actually cover everything. Let me know if you can't get that to work. I did not. I'll take a look,

Re: [racket-users] After font change, bounding box too high in Redex picts

Did you trying using basic-text and the other nearby-documented parameters to add the inset you want? I think those ones might actually cover everything. Let me know if you can't get that to work. The standard TeX fonts (cmr, etc) are strange (and, IMO, ugly, but that's a separate issue ;). So

[racket-users] interested in TR internal error?

I was in the middle of editing a TR file, and I got this error: ../../racket/racket/share/pkgs/typed-racket-lib/typed-racket/utils/tc-utils.rkt:270:0: Internal Typechecker Error: contract-def-property: thunk called too early This should only be called after the type-checking pass has finished.

Re: [racket-users] After font change, bounding box too high in Redex picts

On Fri, Jun 24, 2016 at 11:47:45AM -0500, Robby Findler wrote: > The current-render-pict-adjust functionality isn't really designed to > work around your disagreements with font designers about how much > space they put around letter forms. :) or am I misinterpreting > your comment? If

Re: [racket-users] After font change, bounding box too high in Redex picts

The current-render-pict-adjust functionality isn't really designed to work around your disagreements with font designers about how much space they put around letter forms. :) or am I misinterpreting your comment? If that's really an accurate characterization, then you might actually try

Re: [racket-users] After font change, bounding box too high in Redex picts

On Thu, Jun 02, 2016 at 11:40:14PM -0400, 'William J. Bowman' via Racket Users wrote: > Attached are two screenshots of part of a model I've extracted from Redex. > > The first is after changing the default fonts. As you can see, there is a > *lot* of vertical whitespace > around each judgment

Re: [racket-users] Using Futures with FFI functions

The runtime system currently cannot call a foreign function without suspending a future. Supporting that operation is not out of the question, but I don't think it will be easy. Would using places work in this case --- creating one or more places on start-up to serve data from the C library? At