v0.19 port ssl #1446
No reviewers
Labels
No labels
UX
active development
backlog
blocker
bootstrap
bounty
bug
dependencies
discussion
documentation
duplicate
enhancement
flaky test
help wanted
invalid
javascript
question
release
tendentious
wontfix
No milestone
No project
No assignees
2 participants
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference
mighty-gerbils/gerbil!1446
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "v0.19-ssl"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
brings the ssl lib to v19... and this pulls in a whole lot of things just to run the tests
@ -130,2 +132,4 @@"format/api"(gxc: "crypto/libcrypto""-cc-options" ,(append-options (cppflags "libcrypto" "") "-Wno-discarded-qualifiers" "-Wno-deprecated-declarations" "-Wno-implicit-function-declaration")see #1449 for the ticking rotbombs
Minor comments
@ -0,0 +5,4 @@./libcrypto)(export #t)(deferror-class LibCryptoError () libcrypto-error?Frankly, I don't like these predicate name overrides.
If there's a systematic change from CamelCase to kebab-case, then it should be automated.
Also in this case, why LibCrypto but libcrypto instead of lib-crypto? Or is this part of a special case in handling the libfoos? Now we have an entire parser for some weird underspecified language, just brainside rather than computer-side.
I really don't like it.
eh, it is such a minor thing. We could drop it, i just like to have the option to only export the predicate and hide the class.
For LibFoo errors we should normalize somewhere, hence the attempt.
In my variant, instead of a fixed flag, an optional argument wtih sensible defaults, as in mkostemps. Maybe the mode should be overridable, too.
e.g. O_APPEND, O_ASYNC, O_CLOEXEC, O_DIRECT, O_DIRECTORY, O_DSYNC, O_LARGEFILE, O_NOATIME, O_NOCTTY, O_NOFOLLOW, O_NONBLOCK, O_PATH, O_SYNC, O_TMPFILE, O_TRUNC.
Also, I believe we should include O_EXCL by default.
If using O_TMPFILE (file exists for sharing contents, but pre-deleted so no name sharing) then we don't need the randomness or suffix, only a directory name as the "prefix".
@ -84,0 +60,4 @@(encoded (base64-encode bytespadding: #furlsafe: #t)))(string-append name "." encoded)))For various reasons, we may want a suffix and a prefix, and not just a name. Suffix, because some software wants the file to have a prefix attached (e.g.
.ss.js.xls.doc.exeetc.) If we have a prefix, we don't need the"."just(string-append name encoded suffix)@ -0,0 +25,4 @@(def test-server-address"127.0.0.1:39999")(def test-csr-template #<<ENDNo autodetection that this is inet4 ?
no, we should make the complicate
atonparser that auto detects ip4/6 and dns.I'll make an issue
#1450