5 \___|\___/|_| \_\_____|
7 Things that could be nice to do in the future
9 Things to do in project cURL. Please tell us what you think, contribute and
10 send us patches that improve things!
12 All bugs documented in the KNOWN_BUGS document are subject for fixing!
17 1.4 signal-based resolver timeouts
18 1.5 get rid of PATH_MAX
19 1.6 Modified buffer size approach
20 1.7 Detect when called from witin callbacks
21 1.8 Allow SSL (HTTPS) to proxy
23 2. libcurl - multi interface
25 2.2 Fix HTTP Pipelining for PUT
32 4.2 Alter passive/active on failure and retry
33 4.3 Earlier bad letter detection
34 4.4 REST for large files
38 5.1 Better persistency for HTTP 1.0
39 5.2 support FF3 sqlite cookie files
40 5.3 Rearrange request header order
46 6.2 ditch telnet-specific select
47 6.3 feature negotiation debug data
48 6.4 send data in chunks
52 7.2 Enhanced capability support
56 8.2 Enhanced capability support
59 9.1 Enhanced capability support
62 10.1 SASL based authentication mechanisms
68 12.1 Disable specific versions
69 12.2 Provide mutex locking API
70 12.3 Evaluate SSL patches
71 12.4 Cache OpenSSL contexts
72 12.5 Export session ids
73 12.6 Provide callback for cert verification
74 12.7 improve configure --with-ssl
82 14.1 Other authentication mechanisms
87 15.3 prevent file overwriting
88 15.4 simultaneous parallel transfers
89 15.5 provide formpost headers
90 15.6 url-specific options
91 15.7 warning when setting an option
92 15.8 IPv6 addresses with globbing
99 17.2 nicer lacking perl message
100 17.3 more protocols supported
101 17.4 more platforms supported
104 18.1 http-style HEAD output for ftp
105 18.2 combine error codes
106 18.3 extend CURLOPT_SOCKOPTFUNCTION prototype
108 19. Next major release
109 19.1 cleanup return codes
110 19.2 remove obsolete defines
112 19.4 remove several functions
113 19.5 remove CURLOPT_FAILONERROR
114 19.6 remove CURLOPT_DNS_USE_GLOBAL_CACHE
115 19.7 remove progress meter from libcurl
116 19.8 remove 'curl_httppost' from public
117 19.9 have form functions use CURL handle argument
118 19.10 Add CURLOPT_MAIL_CLIENT option
120 ==============================================================================
124 1.2 More data sharing
126 curl_share_* functions already exist and work, and they can be extended to
127 share more. For example, enable sharing of the ares channel and the
132 Use 'struct lifreq' and SIOCGLIFADDR instead of 'struct ifreq' and
133 SIOCGIFADDR on newer Solaris versions as they claim the latter is obsolete.
134 To support ipv6 interface addresses for network interfaces properly.
136 1.4 signal-based resolver timeouts
138 libcurl built without an asynchronous resolver library uses alarm() to time
139 out DNS lookups. When a timeout occurs, this causes libcurl to jump from the
140 signal handler back into the library with a sigsetjmp, which effectively
141 causes libcurl to continue running within the signal handler. This is
142 non-portable and could cause problems on some platforms. A discussion on the
143 problem is available at http://curl.haxx.se/mail/lib-2008-09/0197.html
145 Also, alarm() provides timeout resolution only to the nearest second. alarm
146 ought to be replaced by setitimer on systems that support it.
148 1.5 get rid of PATH_MAX
150 Having code use and rely on PATH_MAX is not nice:
151 http://insanecoding.blogspot.com/2007/11/pathmax-simply-isnt.html
153 Currently the SSH based code uses it a bit, but to remove PATH_MAX from there
154 we need libssh2 to properly tell us when we pass in a too small buffer and
155 its current API (as of libssh2 1.2.7) doesn't.
157 1.6 Modified buffer size approach
159 Current libcurl allocates a fixed 16K size buffer for download and an
160 additional 16K for upload. They are always unconditionally part of the easy
161 handle. If CRLF translations are requested, an additional 32K "scratch
162 buffer" is allocated. A total of 64K transfer buffers in the worst case.
164 First, while the handles are not actually in use these buffers could be freed
165 so that lingering handles just kept in queues or whatever waste less memory.
167 Secondly, SFTP is a protocol that needs to handle many ~30K blocks at once
168 since each need to be individually acked and therefore libssh2 must be
169 allowed to send (or receive) many separate ones in parallel to achieve high
170 transfer speeds. A current libcurl build with a 16K buffer makes that
171 impossible, but one with a 512K buffer will reach MUCH faster transfers. But
172 allocating 512K unconditionally for all buffers just in case they would like
173 to do fast SFTP transfers at some point is not a good solution either.
175 Dynamically allocate buffer size depending on protocol in use in combination
176 with freeing it after each individual transfer? Other suggestions?
178 1.7 Detect when called from witin callbacks
180 We should set a state variable before calling callbacks, so that we
181 subsequently can add code within libcurl that returns error if called within
182 callbacks for when that's not supported.
184 1.8 Allow SSL (HTTPS) to proxy
186 To prevent local users from snooping on your traffic to the proxy. Supported
188 http://www.chromium.org/developers/design-documents/secure-web-proxy
191 2. libcurl - multi interface
193 2.1 More non-blocking
195 Make sure we don't ever loop because of non-blocking sockets returning
196 EWOULDBLOCK or similar. Blocking cases include:
198 - Name resolves on non-windows unless c-ares is used
199 - NSS SSL connections
200 - HTTP proxy CONNECT operations
201 - SOCKS proxy handshakes
204 - The "DONE" operation (post transfer protocol-specific actions) for the
205 protocols SFTP, SMTP, FTP. Fixing Curl_done() for this is a worthy task.
207 2.2 Fix HTTP Pipelining for PUT
209 HTTP Pipelining can be a way to greatly enhance performance for multiple
210 serial requests and currently libcurl only supports that for HEAD and GET
211 requests but it should also be possible for PUT.
223 HOST is a suggested command in the works for a client to tell which host name
224 to use, to offer FTP servers named-based virtual hosting:
226 http://tools.ietf.org/html/draft-hethmon-mcmurray-ftp-hosts-11
228 4.2 Alter passive/active on failure and retry
230 When trying to connect passively to a server which only supports active
231 connections, libcurl returns CURLE_FTP_WEIRD_PASV_REPLY and closes the
232 connection. There could be a way to fallback to an active connection (and
233 vice versa). http://curl.haxx.se/bug/feature.cgi?id=1754793
235 4.3 Earlier bad letter detection
237 Make the detection of (bad) %0d and %0a codes in FTP url parts earlier in the
238 process to avoid doing a resolve and connect in vain.
240 4.4 REST for large files
242 REST fix for servers not behaving well on >2GB requests. This should fail if
243 the server doesn't set the pointer to the requested index. The tricky
244 (impossible?) part is to figure out if the server did the right thing or not.
248 FTP ASCII transfers do not follow RFC959. They don't convert the data
253 5.1 Better persistency for HTTP 1.0
255 "Better" support for persistent connections over HTTP 1.0
256 http://curl.haxx.se/bug/feature.cgi?id=1089001
258 5.2 support FF3 sqlite cookie files
260 Firefox 3 is changing from its former format to a a sqlite database instead.
261 We should consider how (lib)curl can/should support this.
262 http://curl.haxx.se/bug/feature.cgi?id=1871388
264 5.3 Rearrange request header order
266 Server implementors often make an effort to detect browser and to reject
267 clients it can detect to not match. One of the last details we cannot yet
268 control in libcurl's HTTP requests, which also can be exploited to detect
269 that libcurl is in fact used even when it tries to impersonate a browser, is
270 the order of the request headers. I propose that we introduce a new option in
271 which you give headers a value, and then when the HTTP request is built it
272 sorts the headers based on that number. We could then have internally created
273 headers use a default value so only headers that need to be moved have to be
278 The first drafts for HTTP2 have been published
279 (http://tools.ietf.org/html/draft-ietf-httpbis-http2-03) and is so far based
280 on SPDY (http://www.chromium.org/spdy) designs and experiences. Chances are
281 it will end up in that style. Chrome and Firefox already support SPDY and
282 lots of web services do.
284 It would make sense to implement SPDY support now and later transition into
285 or add HTTP2 support as well.
287 We should base or HTTP2/SPDY work on a 3rd party library for the protocol
288 fiddling. The Spindy library (http://spindly.haxx.se/) was an attempt to make
289 such a library with an API suitable for use by libcurl but that effort has
290 more or less stalled. spdylay (https://github.com/tatsuhiro-t/spdylay) may
291 be a better option, either used directly or wrapped with a more spindly-like
296 Add the ability to specify the preferred authentication mechanism to use by
297 using ;auth=<mech> in the login part of the URL.
301 http://test:pass;auth=NTLM@example.com would be equivalent to specifing --user
302 test:pass;auth=NTLM or --user test:pass --ntlm from the command line.
304 Additionally this should be implemented for proxy base URLs as well.
310 Reading input (to send to the remote server) on stdin is a crappy solution for
311 library purposes. We need to invent a good way for the application to be able
312 to provide the data to send.
314 6.2 ditch telnet-specific select
316 Move the telnet support's network select() loop go away and merge the code
317 into the main transfer loop. Until this is done, the multi interface won't
320 6.3 feature negotiation debug data
322 Add telnet feature negotiation data to the debug callback as header data.
324 6.4 send data in chunks
326 Currently, telnet sends data one byte at a time. This is fine for interactive
327 use, but inefficient for any other. Sent data should be sent in larger
334 Add support for pipelining emails.
336 7.2 Enhanced capability support
338 Add the ability, for an application that uses libcurl, to obtain the list of
339 capabilities returned from the EHLO command.
345 Add support for pipelining commands.
347 8.2 Enhanced capability support
349 Add the ability, for an application that uses libcurl, to obtain the list of
350 capabilities returned from the CAPA command.
354 9.1 Enhanced capability support
356 Add the ability, for an application that uses libcurl, to obtain the list of
357 capabilities returned from the CAPABILITY command.
361 10.1 SASL based authentication mechanisms
363 Currently the LDAP module only supports ldap_simple_bind_s() in order to bind
364 to an LDAP server. However, this function sends username and password details
365 using the simple authentication mechanism (as clear text). However, it should
366 be possible to use ldap_bind_s() instead specifing the security context
367 information ourselves.
373 There's no RFC for the protocol or an URI/URL format. An implementation
374 should most probably use an existing rsync library, such as librsync.
378 12.1 Disable specific versions
380 Provide an option that allows for disabling specific SSL versions, such as
381 SSLv2 http://curl.haxx.se/bug/feature.cgi?id=1767276
383 12.2 Provide mutex locking API
385 Provide a libcurl API for setting mutex callbacks in the underlying SSL
386 library, so that the same application code can use mutex-locking
387 independently of OpenSSL or GnutTLS being used.
389 12.3 Evaluate SSL patches
391 Evaluate/apply Gertjan van Wingerde's SSL patches:
392 http://curl.haxx.se/mail/lib-2004-03/0087.html
394 12.4 Cache OpenSSL contexts
396 "Look at SSL cafile - quick traces look to me like these are done on every
397 request as well, when they should only be necessary once per ssl context (or
398 once per handle)". The major improvement we can rather easily do is to make
399 sure we don't create and kill a new SSL "context" for every request, but
400 instead make one for every connection and re-use that SSL context in the same
401 style connections are re-used. It will make us use slightly more memory but
402 it will libcurl do less creations and deletions of SSL contexts.
404 12.5 Export session ids
406 Add an interface to libcurl that enables "session IDs" to get
407 exported/imported. Cris Bailiff said: "OpenSSL has functions which can
408 serialise the current SSL state to a buffer of your choice, and recover/reset
409 the state from such a buffer at a later date - this is used by mod_ssl for
410 apache to implement and SSL session ID cache".
412 12.6 Provide callback for cert verification
414 OpenSSL supports a callback for customised verification of the peer
415 certificate, but this doesn't seem to be exposed in the libcurl APIs. Could
416 it be? There's so much that could be done if it were!
418 12.7 improve configure --with-ssl
420 make the configure --with-ssl option first check for OpenSSL, then GnuTLS,
425 DNS-Based Authentication of Named Entities (DANE) is a way to provide SSL
426 keys and certs over DNS using DNSSEC as an alternative to the CA model.
427 http://www.rfc-editor.org/rfc/rfc6698.txt
429 An initial patch was posted by Suresh Krishnaswamy on March 7th 2013
430 (http://curl.haxx.se/mail/lib-2013-03/0075.html) but it was a too simple
431 approach. See Daniel's comments:
432 http://curl.haxx.se/mail/lib-2013-03/0103.html . libunbound may be the
433 correct library to base this development on.
437 13.1 SSL engine stuff
439 Is this even possible?
441 13.2 check connection
443 Add a way to check if the connection seems to be alive, to correspond to the
444 SSL_peak() way we use with OpenSSL.
448 14.1 Other authentication mechanisms
450 Add support for GSSAPI to SMTP, POP3 and IMAP.
456 "curl --sync http://example.com/feed[1-100].rss" or
457 "curl --sync http://example.net/{index,calendar,history}.html"
459 Downloads a range or set of URLs using the remote name, but only if the
460 remote file is newer than the local file. A Last-Modified HTTP date header
461 should also be used to set the mod date on the downloaded file.
465 Globbing support for -d and -F, as in 'curl -d "name=foo[0-9]" URL'.
466 This is easily scripted though.
468 15.3 prevent file overwriting
470 Add an option that prevents cURL from overwriting existing local files. When
471 used, and there already is an existing file with the target file name
472 (either -O or -o), a number should be appended (and increased if already
473 existing). So that index.html becomes first index.html.1 and then
476 15.4 simultaneous parallel transfers
478 The client could be told to use maximum N simultaneous parallel transfers and
479 then just make sure that happens. It should of course not make more than one
480 connection to the same remote host. This would require the client to use the
481 multi interface. http://curl.haxx.se/bug/feature.cgi?id=1558595
483 15.5 provide formpost headers
485 Extending the capabilities of the multipart formposting. How about leaving
486 the ';type=foo' syntax as it is and adding an extra tag (headers) which
487 works like this: curl -F "coolfiles=@fil1.txt;headers=@fil1.hdr" where
488 fil1.hdr contains extra headers like
490 Content-Type: text/plain; charset=KOI8-R"
491 Content-Transfer-Encoding: base64
492 X-User-Comment: Please don't use browser specific HTML code
494 which should overwrite the program reasonable defaults (plain/text,
497 15.6 url-specific options
499 Provide a way to make options bound to a specific URL among several on the
500 command line. Possibly by letting ':' separate options between URLs,
503 curl --data foo --url url.com : \
505 --url url3.com --data foo3
507 (More details: http://curl.haxx.se/mail/archive-2004-07/0133.html)
509 The example would do a POST-GET-POST combination on a single command line.
511 15.7 warning when setting an option
513 Display a warning when libcurl returns an error when setting an option.
514 This can be useful to tell when support for a particular feature hasn't been
515 compiled into the library.
517 15.8 IPv6 addresses with globbing
519 Currently the command line client needs to get url globbing disabled (with
520 -g) for it to support IPv6 numerical addresses. This is a rather silly flaw
521 that should be corrected. It probably involves a smarter detection of the
528 Consider extending 'roffit' to produce decent ASCII output, and use that
529 instead of (g)nroff when building src/tool_hugehelp.c
535 Make our own version of stunnel for simple port forwarding to enable HTTPS
536 and FTP-SSL tests without the stunnel dependency, and it could allow us to
537 provide test tools built with either OpenSSL or GnuTLS
539 17.2 nicer lacking perl message
541 If perl wasn't found by the configure script, don't attempt to run the tests
542 but explain something nice why it doesn't.
544 17.3 more protocols supported
546 Extend the test suite to include more protocols. The telnet could just do ftp
547 or http operations (for which we have test servers).
549 17.4 more platforms supported
551 Make the test suite work on more platforms. OpenBSD and Mac OS. Remove
552 fork()s and it should become even more portable.
556 18.1 http-style HEAD output for ftp
558 #undef CURL_FTP_HTTPSTYLE_HEAD in lib/ftp.c to remove the HTTP-style headers
559 from being output in NOBODY requests over ftp
561 18.2 combine error codes
563 Combine some of the error codes to remove duplicates. The original
564 numbering should not be changed, and the old identifiers would be
565 macroed to the new ones in an CURL_NO_OLDIES section to help with
566 backward compatibility.
568 Candidates for removal and their replacements:
570 CURLE_FILE_COULDNT_READ_FILE => CURLE_REMOTE_FILE_NOT_FOUND
572 CURLE_FTP_COULDNT_RETR_FILE => CURLE_REMOTE_FILE_NOT_FOUND
574 CURLE_FTP_COULDNT_USE_REST => CURLE_RANGE_ERROR
576 CURLE_FUNCTION_NOT_FOUND => CURLE_FAILED_INIT
578 CURLE_LDAP_INVALID_URL => CURLE_URL_MALFORMAT
580 CURLE_TFTP_NOSUCHUSER => CURLE_TFTP_ILLEGAL
582 CURLE_TFTP_NOTFOUND => CURLE_REMOTE_FILE_NOT_FOUND
584 CURLE_TFTP_PERM => CURLE_REMOTE_ACCESS_DENIED
586 18.3 extend CURLOPT_SOCKOPTFUNCTION prototype
588 The current prototype only provides 'purpose' that tells what the
589 connection/socket is for, but not any protocol or similar. It makes it hard
590 for applications to differentiate on TCP vs UDP and even HTTP vs FTP and
593 10. Next major release
595 19.1 cleanup return codes
597 curl_easy_cleanup() returns void, but curl_multi_cleanup() returns a
598 CURLMcode. These should be changed to be the same.
600 19.2 remove obsolete defines
602 remove obsolete defines from curl/curl.h
606 make several functions use size_t instead of int in their APIs
608 19.4 remove several functions
610 remove the following functions from the public API:
614 curl_mprintf (and variations)
620 They will instead become curlx_ - alternatives. That makes the curl app
621 still capable of using them, by building with them from source.
623 These functions have no purpose anymore:
627 curl_multi_socket_all
629 19.5 remove CURLOPT_FAILONERROR
631 Remove support for CURLOPT_FAILONERROR, it has gotten too kludgy and weird
632 internally. Let the app judge success or not for itself.
634 19.6 remove CURLOPT_DNS_USE_GLOBAL_CACHE
636 Remove support for a global DNS cache. Anything global is silly, and we
637 already offer the share interface for the same functionality but done
640 19.7 remove progress meter from libcurl
642 The internally provided progress meter output doesn't belong in the library.
643 Basically no application wants it (apart from curl) but instead applications
644 can and should do their own progress meters using the progress callback.
646 The progress callback should then be bumped as well to get proper 64bit
647 variable types passed to it instead of doubles so that big files work
650 19.8 remove 'curl_httppost' from public
652 curl_formadd() was made to fill in a public struct, but the fact that the
653 struct is public is never really used by application for their own advantage
654 but instead often restricts how the form functions can or can't be modified.
656 Changing them to return a private handle will benefit the implementation and
657 allow us much greater freedoms while still maintining a solid API and ABI.
659 19.9 have form functions use CURL handle argument
661 curl_formadd() and curl_formget() both currently have no CURL handle
662 argument, but both can use a callback that is set in the easy handle, and
663 thus curl_formget() with callback cannot function without first having
664 curl_easy_perform() (or similar) called - which is hard to grasp and a design
667 19.10 Add CURLOPT_MAIL_CLIENT option
669 Rather than use the URL to specify the mail client string to present in the
670 HELO and EHLO commands, libcurl should support a new CURLOPT specifically for
671 specifing this data as the URL is non-standard and to be honest a bit of a
674 Please see the following thread for more information:
675 http://curl.haxx.se/mail/lib-2012-05/0178.html