| Date | Commit message (Collapse) |
|---|
|
A public-inbox is NOT necessarily a mailing list, but it
could serve as an input point for zero, one, or infinite
mailing lists :D
|
|
This reduces the cognitive overhead for mapping names of
configuration values to internal field names of our classes.
Further changes along these lines coming...
|
|
Unfortunately, most users still prefer their mail delivered
over SMTP; so we'll at least document mlmmj integration for now
until we can popularize pull-based reading over POP3/NNTP/ssoma.
|
|
We'll need to test non-UTF-8 messages at some point, too.
There are lots of legacy-encoded messages in old archives
and I would not bet we behave sanely w.r.t. those.
|
|
The Xapian search index is required for the NNTP server, so
there's no point in calling system() for it like we do in
other tests. This should speed up the test a small amount.
|
|
This should help poor developers who still use rotating disks on
cheap netbooks.
|
|
We need Perl to believe everything we send is UTF-8,
make it so, even if it may not be.
Fixes: 265e79ff82ce 'Revert "nntp: proper UTF-8 support (hopefully?)"'
|
|
This reverts commit f81ad477cb013d05b9b11fa051a9ebc5983a5be6.
The raw, undecoded body is probably what should be sent over the
wire anyways for clients to deal with. We'll need this to avoid
deprecation warnings with Perl 5.24+ since we use
send()/recv()/sysread().
|
|
We can maintain the client HTTP connection if the process exited
with failure as long as we terminated our own response properly.
|
|
git doesn't handle '<' and '>' characters in the author
name at all regardless of quoting, not just matched pairs.
So fall back to using the email as the author name since
the commit info isn't critical, anyways (shallow clones
are fine).
|
|
Mbox formatters may add extra newlines at the end of the
message, and that's not relevant for comparing messages
for deletion.
|
|
This allows messages to be read in chronological order when
read without a mail client (e.g. with "zcat t.mbox.gz | less")
|
|
Ugh, I really need to get off my ass to write automated tests for
an Apache2 + mod_perl config.
|
|
When serving large static files or large packs, we may call
Danga::Socket::write directly to queue up callbacks to resume
reading and defer firing them until the socket is writable.
This prevents us from scheduling writes or buffering until we
know the socket is writable and prevents needless buffering by
Danga::Socket when faced with slow clients.
For smart clones, this comes at the cost of throttling the
output of "git pack-objects" to the speed of the client
connection. This is probably not ideal, but is the behavior of
the standard git-daemon, too; and is preferable to running the
httpd out-of-memory. Buffering to the filesystem may be an
option in the future...
|
|
This empty string check is for middlewares such as Deflater
which may write empty strings, not for direct real callers of
Danga::Socket who (presumably) know what they're doing.
|
|
We only need to use env(1) under mod_perl; since mod_perl
is uncommon nowadays, support native %ENV for a teeny
speedup for folks uncomfortable with running vfork via
Inline::C snippet.
|
|
Process startup times are atrocious for fast tests and there's far
too much setup involved. Rely on git-fast-import instead; but
more work is needed in this area.
|
|
It limits flexibility and makes it harder to switch
to use PublicImport::Import.
|
|
For readers using NNTP, we should do our best to advertise the
clonable HTTP/HTTPS URLs and the message permalink URL for
ease-of-referencing messages, since we don't want the NNTP server
and it's sequential article numbers to be relied on.
|
|
Broken threads should be exposed to hopefully encourage people to
use proper mail clients which set In-Reply-To headers.
|
|
Unnecessary on *nix, and we won't support systems
which do insane things.
|
|
We cannot afford to fire Perl-level signal handlers in the
vforked child process since they're not designed to run in
the child like that.
Thus we need to block all signals before calling vfork, reset
signal dispositions in the child, and restore the signal mask in
the parent.
ref: https://ewontfix.com/7
|
|
Since we use sysread, we must use sysseek for symmetry although
PerlIO may be doing a real lseek with "seek", anyways.
Fixes: 310819ea86ac ("git-http-backend: favor sysread for regular files")
|
|
We can reduce the allocation and overhead needed for
Danga::Socket timers for immediately-executed responses by
combining identical timers and reducing anonymous sub creation.
|
|
This should allow users to change and add headers as needed.
While we're at it, add the X-Original-To header Postfix likes
to add; it seems like pointless bloat with the existence of
(important) Received: headers.
|
|
Adding ':' (colon), ',' (comma), '$' (dollar sign) and
supporting TLS-enabled schemes: ftps, nntps variants as
well as gopher :D
|
|
Tilde is common for some homepages: http://example.org/~user/
There's probably some other acceptable characters I'm missing.
|
|
git clones may take longer than 30s, much longer... So prepare
to wait almost indefinitely for sockets to timeout and document
the second signal behavior for immediate shutdown.
While we're at it, move parent death handling to a separate
class to avoid Danga::Socket->AddOtherFds, since that does not
allow proper handling the parent pipe being closed and would
actually misterminate a worker prematurely. t/nntpd.t is update
to illustrate the failure with workers enabled.
We will work to keep memory usage low and let clients take their
time without interrupting them.
|
|
git clones may take a long time and it's wrong to
drop connections in the middle of a transaction.
|
|
Noticed when using a long URL in the subject.
|
|
This saves us a system call for common GET/HEAD requests
with no upload body.
|
|
Email addresses get out-of-date, so make sure they're mapped
properly for future readers. git and linux-kernel already have
an established convention for this, so we will follow it.
|
|
We need to abort connections properly if a response is prematurely
truncated. This includes problems with serving static files, since
a clumsy admin or broken FS could return truncated responses and
inadvertently leave a client waiting (since the client saw
"Content-Length" in the header and expected a certain length).
|
|
The blocking PSGI server may cause EINTR to be hit, here.
|
|
We must use a normal write instead of send(.., MSG_MORE)
when writing responses of "Content-Length: 0" to avoid
the corking effect MSG_MORE provides. We only want to
cork headers if we will send a non-empty body.
Fixes: c3eeaf664cf0 ("http: clarify intent for persistence")
This needs a proper test.
|
|
Server admins may not be able to afford to have too many
git-pack-objects processes running at once. Since PSGI
HTTP servers should already be configured to use multiple
processes for other requests; limit concurrency of smart
backends to one; and fall back to dumb responses if we're
already generating a pack.
|
|
Using http.getanyfile still keeps the http-backend process
alive, so it's better to break out of that process and
handle serving entirely within the HTTP server.
|
|
We should update $GIT_DIR/info/refs for dumb HTTP clients
whenever we make changes to the repository. The best place
to update is immediately after making commits.
This fixes a bug where public-inbox-learn did not properly
update $GIT_DIR/info/refs after inserting or removing
messages.
|
|
This is probably trivial enough to be final?
|
|
While we're at it, update some references to ssoma in the
Makefile.PL comment.
|
|
|
|
This is used all over the place, but may not be in the future,
so ensure we explicitly load it ourselves.
|
|
No need to maintain per-block environment state when we can
localize it to per-command. We've had --git-dir= in git
since 1.4.2 (2006-08-12) and already use it all over the
place.
|
|
By converting to using ourt git-fast-import-based Import
module. This should allow us to be more easily installed.
|
|
The read could fail entirely and leave $lf undefined.
|
|
Danga::Socket timers are not cheap, so avoid creating up
to 3 timers per-newsgroup by batching resource weakening.
This lets us reduce resource consumption for scheduing
additional resource consumption reduction :)
|
|
hdr_val has not been used since commit 1d236e649df1
("nntp: implement OVER/XOVER summary in search document")
|
|
Hopefully this modularizes things a little and allows us
to work on a combined super server to save RAM.
|
|
We may be importing mail from other lists, so do not
clobber the existing List-Id header.
|
|
ref: https://www.w3.org/TR/html/links.html#sequential-link-types
Followup-to: c4183f56aab6 ("www: add rel=next and rel=prev navigation hints")
|