Open tcllib tickets: Not reproduced

Triage of all 509 open and Pending tickets of tcllib (bugs, patches and feature requests) on 2026-11-15. Each ticket was checked against tcllib trunk on Linux with Tcl 9.1, 9.0, 8.6 and 8.5, using local servers for the network modules and a Windows 11 VM for Windows-only tickets. Outdated, superseded, duplicate and misfiled tickets are listed separately. Rows marked patch link to a prototype fix with a regression test (147 in all, in tcllib-triage/prototypes). Updated 2026-11-16.

Not reproduced 23

⬜f4efd0f5912023websocket testsuite is brokenNOT-REPRODUCED (all 5 tests pass on 9.1 and 8.6 against a local websocket server); the real remaining problem is that the testsuite depends on the …

Open ticket f4efd0f591 · Bug · websocket · filed 2023-01-30

Full verdict

NOT-REPRODUCED (all 5 tests pass on 9.1 and 8.6 against a local websocket server); the real remaining problem is that the testsuite depends on the external server ws.ifelse.io

  • Module websocket 1.6
  • Tested copied websocket.test with wsUrl pointed at a local tcllib websocket::server that answers "Request served by local". 9.1.1 (http 2.10.3) and 8.6.18 (http 2.9.9) both gave 5/5. Socket reuse (tests 1.3/1.4) no longer happens: http ≥ 2.9.8 opens a fresh socket for upgrade requests (websocket.tcl ~1203).
  • Cause the failures came from http < 2.9.8 reusing the kept-alive socket. Today the tests fail only when ws.ifelse.io is unreachable or changes its reply.
  • Prototype none (better fix: make the tests use a local server, or add a network constraint; bigger than a quick fix)
  • Suggested action close the original issue as Fixed (http 2.9.8+). Consider converting websocket.test to a local in-process server so sak test run does not contact external hosts.
  • Suggested ticket comment With http 2.9.8+ the upgrade uses a fresh socket, and all five tests pass on 8.6 and 9.1 against a local server. Remaining failures depend only on reaching ws.ifelse.io; the tests should use a local server.
⏳9e2bb6d60f2020zipfile::mkzipNOT-REPRODUCED (round trip of PDF-like binary data, with CR/LF, ^Z and high bytes, small (deflate in one chunk) and >2 MB (stream), is byte-identical)

Open ticket 9e2bb6d60f · Bug · zip · filed 2020-05-11 · Pending (resolved, waiting for confirmation)

⏳ Since triage (updated 2026-11-16)

Status now: Pending

    Full verdict

    NOT-REPRODUCED (round trip of PDF-like binary data, with CR/LF, ^Z and high bytes, small (deflate in one chunk) and >2 MB (stream), is byte-identical)

    • Module zipfile::mkzip 1.2.4
    • Tested mkzip then zipfile::decode::unzipfile and cmp, plus unzip -t, on 9.1 and 8.6. Both files were identical and the archive was OK.
    • Cause unknown. The report only shows that a PDF 1.7 file has no trailer keyword. That is normal for PDF 1.5+ with cross-reference streams, so the "trailer not found" message comes from the PDF reader, not from tcllib. aku asked for a reproducer in 2021 and got no reply.
    • Prototype none
    • Suggested action close as Works For Me / insufficient information.
    • Suggested ticket comment mkzip round-trips binary PDF data exactly. PDF 1.5+ files with cross-reference streams have no "trailer" keyword, so the message comes from the PDF tool. No reproducer since 2021, so closing.
    ⬜c27da654f82018websocket 1.4.1 closed sockets PingNOT-REPRODUCED. I tried a server-side connection with -keepalive 1 and the peer either sending a close frame or closing abruptly, waiting 3 s. Th…

    Open ticket c27da654f8 · Bug · websocket · filed 2018-05-25

    Full verdict

    NOT-REPRODUCED. I tried a server-side connection with -keepalive 1 and the peer either sending a close frame or closing abruptly, waiting 3 s. There was no bgerror and no pending after afterwards, on 9.1 and 8.6. Disconnect cancels Connection(liveness).

    • Module websocket 1.6
    • Tested local 127.0.0.1 socket pair, as above.
    • Cause unknown. One plausible path: ::websocket::New resets Connection(liveness) to "" without cancelling an already scheduled ping, e.g. when takeover runs twice on the same socket. That would leave an orphan timer, and since Windows reuses socket names, it could fire later. Unconfirmed.
    • Prototype none (no reproducer)
    • Suggested action close as Works for me unless the reporter can give steps (Windows, 1.6)
    • Suggested ticket comment I could not reproduce this with 1.6: closing the connection cancels the keepalive timer. If it still happens, please attach a small script.
    ⬜f084c0dc032015memory leak in tcllib autoproxyNOT-REPRODUCED (9.1, 8.6; plain-http proxy path)

    Open ticket f084c0dc03 · Bug · http :: autoproxy · filed 2015-09-29

    Full verdict

    NOT-REPRODUCED (9.1, 8.6; plain-http proxy path)

    • Module autoproxy 1.8.1
    • Tested 6000 geturl+cleanup calls through autoproxy::init 127.0.0.1:port with a local proxy: RSS and the number of ::http::* vars stay flat on 9.1 and 8.6. The HTTPS/CONNECT path could not be tested with a real TLS proxy (no tls package here). aku could not reproduce it in 2019 either.
    • Cause the errors reported in 2016 (state(sock) missing in http::Finish/reset) come from the Tcl 8.5/8.6 http package. That package was rewritten in Tcl 9.
    • Prototype none
    • Suggested action close as Works for me / Outdated
    • Suggested ticket comment Not reproducible with current Tcl and tcllib (no growth over 6000 proxied requests). The 2016 traces point at old Tcl http internals.
    ⬜9a38890c982015cron's self-tests may fail sometimesNOT-REPRODUCED (20 concurrent runs, 10 on 9.1 and 10 on 8.6, all 48/48 passed). The suite was rewritten for cron 2.0 (d9b48225d8, 2016-07) and the …

    Open ticket 9a38890c98 · Bug · cron · filed 2015-05-27

    Full verdict

    NOT-REPRODUCED (20 concurrent runs, 10 on 9.1 and 10 on 8.6, all 48/48 passed). The suite was rewritten for cron 2.0 (d9b48225d8, 2016-07) and the 10/11-second "elapsed" tests the ticket describes are commented out ("Test removed - Was too unstable on a busy computer").

    • Module cron 2.2
    • Tested sak test run modules/cron, 20 runs in parallel
    • Prototype none
    • Suggested action close as Fixed / Out of date
    • Suggested ticket comment The timing-sensitive elapsed-time checks were removed when the cron tests were rewritten for 2.x. 20 parallel runs on 9.1 and 8.6 all pass.
    ⬜299cf8e3672013yaml can't parse some files properlyNOT-REPRODUCED (the attached s21SOIImVi parses correctly on 9.1 and 8.6, same as PyYAML); the other file (s21BwGvY4m) is lost with the paste site

    Open ticket 299cf8e367 · Bug · yaml · filed 2013-11-03

    Full verdict

    NOT-REPRODUCED (the attached s21SOIImVi parses correctly on 9.1 and 8.6, same as PyYAML); the other file (s21BwGvY4m) is lost with the paste site

    • Module yaml 0.4.2
    • Tested parsed the attachment with yaml2dict and compared key sections with PyYAML; they match.
    • Cause n/a
    • Prototype none
    • Suggested action close as Works for me / cannot reproduce
    • Suggested ticket comment The attached file now parses correctly on Tcl 8.6 and 9.1. The other file is no longer available; please reopen with an attachment if it still fails.
    ⬜4f9147ded42013autoproxy HTTPS support should handle authenticationNOT-REPRODUCED (CONNECT already sends basic auth)

    Open ticket 4f9147ded4 · Bug · http :: autoproxy · filed 2013-10-25

    Full verdict

    NOT-REPRODUCED (CONNECT already sends basic auth)

    • Module autoproxy 1.8.1
    • Tested 8.6, with a local fake proxy on 127.0.0.1 and a tls stub: after init 127.0.0.1:port + configure -basic -username me -password secret, the CONNECT request carries Proxy-Authorization: Basic bWU6c2VjcmV0. On 9.x it fails earlier because of 29ff5c3abd.
    • Cause tunnel_write has sent options(basic) since 4d9dbdb9fd (2008). Only Basic is supported, no Digest/NTLM, as for plain http.
    • Prototype none
    • Suggested action close as Works for me (or ask the reporter what failed)
    • Suggested ticket comment The CONNECT request sends the configured Basic credentials (since 2008), tested against a local proxy. Digest/NTLM are not supported for http either. Please reopen with details if it still fails.
    ⏳3428262fff2011JSON: Trailing spaces added to number valueNOT-REPRODUCED (Tcl implementation, 9.1 and 8.6: { "a" : 0 } gives a 0). The C lexer was rewritten in 2014 (11390a7baa, bison parser). It now u…

    Open ticket 3428262fff · Bug · json · filed 2011-10-25 · Pending (resolved, waiting for confirmation)

    ⏳ Since triage (updated 2026-11-16)

    Status now: Pending (Invalid)

      Full verdict

      NOT-REPRODUCED (Tcl implementation, 9.1 and 8.6: { "a" : 0 } gives a 0). The C lexer was rewritten in 2014 (11390a7baa, bison parser). It now uses strtod's end pointer, so no whitespace can end up in a number.

      • Module json 1.3.6 / jsonc 1.1.2
      • Tested 9.1, 8.6 (Tcl impl; tcllibc/critcl not available here, C code read)
      • Cause n/a (old C code)
      • Prototype none
      • Suggested action close as Works for me (ticket already Pending/Invalid)
      • Suggested ticket comment Not reproducible; the C parser was rewritten in 2014 and numbers no longer include trailing whitespace.
      ⏳3427030fff2011Memory leak in md5crypt::aprcrypt in tcllib 1.13NOT-REPRODUCED (aku could not confirm in 2011 with a memory-debug build; ticket is Pending/Invalid since then)

      Open ticket 3427030fff · Bug · md5crypt · filed 2011-10-21 · Pending (resolved, waiting for confirmation)

      ⏳ Since triage (updated 2026-11-16)

      Status now: Pending (Invalid)

        Full verdict

        NOT-REPRODUCED (aku could not confirm in 2011 with a memory-debug build; ticket is Pending/Invalid since then)

        • Module md5crypt 1.1.0
        • Tested none beyond code review; the pure-Tcl path has no allocations that persist, and the C path (md5cryptc) returns a static buffer
        • Cause n/a
        • Prototype none
        • Suggested action close as Invalid (already Pending/Invalid for 15 years)
        • Suggested ticket comment Could not be confirmed in 2011, and the reporter never followed up. Closing.
        ⏳3035986fff2010determineSVD gives wrong dimensions in U for m<nNOT-REPRODUCED (Arjen's 2015 analysis holds)

        Open ticket 3035986fff · Bug · math :: linearalgebra · filed 2010-07-28 · Pending (resolved, waiting for confirmation)

        ⏳ Since triage (updated 2026-11-16)

        Status now: Pending

          Full verdict

          NOT-REPRODUCED (Arjen's 2015 analysis holds)

          • Module math, math::linearalgebra 1.1.7
          • Tested 9.1, 8.6. For a 2×4 A, U is 2×2, S has 4 values and V is 4×4. For a 3×2 A, U is 3×2 (thin), S has 2 and V is 2×2. U is never m×n for m<n. The wish for a full U when m>n is an RFE.
          • Prototype none
          • Suggested action close as Works For Me (Pending since 2015, no reply)
          • Suggested ticket comment U is m×m for m<n, as Arjen showed. A full (non-thin) U for m>n would be a separate feature request.
          ⬜2906306fff2009md5::md5 -file option error on closeNOT-REPRODUCED (normal use, and nested md5 -file calls from after handlers during md5's internal vwait, 20+20 runs on 9.1/8.6: correct hashes, no…

          Open ticket 2906306fff · Bug · md5 · filed 2009-11-30

          Full verdict

          NOT-REPRODUCED (normal use, and nested md5 -file calls from after handlers during md5's internal vwait, 20+20 runs on 9.1/8.6: correct hashes, no leaked or missing channels)

          • Module md5 2.0.9 (md5x.tcl)
          • Tested nested-event-loop script on 9.1 and 8.6
          • Cause unknown. md5 -file reads through fileevent + vwait (md5x.tcl ::md5::md5), so it runs a nested event loop. In an event-driven server (Wub) other handlers run inside it and could close the channel, e.g. a handler that closes channels by name. md4, sha1, sha256, ripemd use the same pattern.
          • Prototype none (not reproducible; a robust change would be to read the file synchronously in a loop instead of vwait, but that's a design change in 6 modules)
          • Suggested action close as Works-for-me, or keep as an RFE to drop the vwait in the -file/-channel paths of the hash modules
          • Suggested ticket comment Cannot reproduce with md5 2.0.9 on Tcl 8.6/9.1, even with nested md5 -file calls. md5 -file runs a nested vwait, so in an event-driven server another handler can close the channel meanwhile. Reading synchronously would avoid that.
          ⬜2824771fff2009doctools test case 50.0.1 failsNOT-REPRODUCED (sak test run modules/doctools2idx: 955/955 pass on 9.1 and 8.6; test doctools-idx-50.* exists and passes)

          Open ticket 2824771fff · Bug · doctools · filed 2009-07-21

          Full verdict

          NOT-REPRODUCED (sak test run modules/doctools2idx: 955/955 pass on 9.1 and 8.6; test doctools-idx-50.* exists and passes)

          • Module doctools2idx (doctools::idx 2.x, pluginmgr)
          • Tested full doctools2idx test suite on 9.1 and 8.6 (Linux)
          • Cause 2009 Solaris + Tcl 8.6 CVS head: can not find channel named "stdout" in the plugin's safe interp. This came from early 8.6 std-channel changes and no longer happens.
          • Prototype none
          • Suggested action close as Out of Date / Works for me
          • Suggested ticket comment The doctools2idx suite (including doctools-idx-50.*) passes on Tcl 8.6 and 9.1. The 2009 stdout-in-plugin-interp failure no longer occurs.
          ⬜1779348fff2007profiler test suite peculiar behaviorNOT-REPRODUCED (sak test run modules/profiler on 9.1: all progress output stays on one CR-overwritten line and the output is 7 lines in total; 18…

          Open ticket 1779348fff · Bug · profiler · filed 2007-08-22

          Full verdict

          NOT-REPRODUCED (sak test run modules/profiler on 9.1: all progress output stays on one CR-overwritten line and the output is 7 lines in total; 18 passed, 1 skipped)

          • Module profiler 0.7 / sak test runner
          • Tested sak test run on Linux, Tcl 9.1. The 2007 report came from a Solaris xterm displayed through Exceed; AK suspected the terminal mangling CR.
          • Cause likely the terminal/X setup, not tcllib.
          • Prototype none
          • Suggested action close as Works For Me / Out of Date
          • Suggested ticket comment Cannot reproduce with current sak: the profiler progress lines overwrite each other like those of other modules. Most likely the terminal setup of the time; closing.
          ⬜1709334fff2007Missing trailing newline hangs sendmessageNOT-REPRODUCED (9.1, 8.6; local fake SMTP server, bodies with and without a trailing newline)

          Open ticket 1709334fff · Bug · smtp · filed 2007-04-28

          Full verdict

          NOT-REPRODUCED (9.1, 8.6; local fake SMTP server, bodies with and without a trailing newline)

          • Module mime / smtp 1.5.2
          • Tested logged the DATA section. Without a trailing newline, the body still ends with CRLF and "." is on its own line.
          • Cause The non-Trf path appends CRLF (bug #827436 fix, 2006). The Trf path's flush/write adds it too, and ce8dab3154 (2021) fixed data loss at the end of the Trf path. Trf does not exist for Tcl 9.
          • Prototype none
          • Suggested action close as Works for me / Fixed
          • Suggested ticket comment Cannot reproduce on trunk: a body without a final newline is terminated with CRLF before the closing ".". Closing.
          ⬜1566389fff2006tests fail in fileutil 1.9patchNOT-REPRODUCED on Windows (9.1): fileutil.test install-2.1 and install-2.2 pass (fileutil 345 passed, 107 skipped, 0 failed).

          Open ticket 1566389fff · Bug · fileutil · filed 2006-09-27

          Full verdict

          NOT-REPRODUCED on Windows (9.1): fileutil.test install-2.1 and install-2.2 pass (fileutil 345 passed, 107 skipped, 0 failed).

          • Module fileutil 1.16.4 (tests install-2.1/2.2)
          • Tested fileutil suite on 9.1 and 8.6 (Linux): all pass
          • Cause fileutil.test install-2.1/2.2 call glob -directory D [file join . / *] = glob -directory D /*. On Windows /* is volume-relative, which is likely why it failed there.
          • Prototype prototypes/1566389fff.patch (test-only, +2/-2): the pattern becomes *. It passes on Linux 9.1/8.6. I could not show the failure without Windows.
          • Suggested action run fileutil.test on Windows; apply the prototype if install-2.x fail there, otherwise close as Out of Date
          • Suggested ticket comment Passes on Linux with 8.6/9.1. The tests glob with pattern /* ([file join . / *]), which is suspect on Windows. Proposed changing it to *.
          Windows VM, run 1
          • Verdict NOT-REPRODUCED on Windows (9.1): fileutil.test install-2.1 and install-2.2 pass (fileutil 345 passed, 107 skipped, 0 failed).
          • Tested full sak test run of trunk with tclsh91 on the Windows VM. The odd glob pattern in the tests is harmless now, so the test-only prototype is cosmetic.
          ⬜1566475fff2006pop3-6.1 fails sometimesNOT-REPRODUCED (pop3 suite passes on 9.1 and 8.6, 44/45 with 1 TLS skip. dgp could not reproduce it either in 2006, after aku's dialog/test fixes)

          Open ticket 1566475fff · Bug · pop3 · filed 2006-09-27

          Full verdict

          NOT-REPRODUCED (pop3 suite passes on 9.1 and 8.6, 44/45 with 1 TLS skip. dgp could not reproduce it either in 2006, after aku's dialog/test fixes)

          • Module pop3 1.11
          • Tested sak test run modules/pop3 on 9.1 and 8.6.
          • Prototype none
          • Suggested action close as Works For Me / Fixed
          • Suggested ticket comment Solaris-only timing failure from 2006; not reproducible since aku's test fixes then, and the suite passes on 8.6 and 9.1.
          ⬜1548418fff2006ftp::Put hangs when non-existent directoryNOT-REPRODUCED (Put to nodir/x.txt against ftpd, active and passive, 9.1 and 8.6, returns 0 at once; the dialog server replying 553 after an earl…

          Open ticket 1548418fff · Bug · ftp · filed 2006-08-29

          Full verdict

          NOT-REPRODUCED (Put to nodir/x.txt against ftpd, active and passive, 9.1 and 8.6, returns 0 at once; the dialog server replying 553 after an early data connection also returns 0)

          • Module ftp 2.4.14
          • Tested see above
          • Cause a hang is possible in active mode for larger uploads when the server connects before refusing, via the same un-cleared state.data as 744276. The 744276 prototype also covers put_sent.
          • Prototype none (covered by prototypes/744276ffff.patch, test ftp-active-error-1.1)
          • Suggested action close as Works for me, or close together with 744276 when that fix is applied
          • Suggested ticket comment Not reproducible on current trunk: Put into a missing directory returns 0 at once. The remaining active-mode wait-on-error issue is handled with [744276].
          ⬜1397543fff2006Snit v2 not handling tcl8.5 return -optionsNOT-REPRODUCED (9.1, 8.6)

          Open ticket 1397543fff · Bug · snit · filed 2006-01-05

          Full verdict

          NOT-REPRODUCED (9.1, 8.6)

          • Module snit 2.3.4
          • Tested a mymethod callback to a method that does return -code error -errorcode {MY CODE}, and one that returns break. Error code, info and break code are preserved.
          • Cause RT.CallInstance re-throws with errorInfo/errorCode. Only extra custom return options are lost.
          • Prototype none
          • Suggested action close as Works for me
          • Suggested ticket comment mymethod callbacks keep errorCode/errorInfo and non-error return codes. Closing.
          ⬜1390477fff2005snit can destroy errorInfo in some casesNOT-REPRODUCED (9.1, 8.6, 8.5; the ticket's trace-on-unset scenario)

          Open ticket 1390477fff · Patch · snit · filed 2005-12-26

          Full verdict

          NOT-REPRODUCED (9.1, 8.6, 8.5; the ticket's trace-on-unset scenario)

          • Module snit 2.3.4 / 1.4.3
          • Tested a proc that errors while an unset trace destroys a snit object whose destructor calls cget/configure. Both caught errorInfo and the global errorInfo keep the real error in snit 1 and 2 (the core saves interp state around traces). snit 1 still overwrites ::errorInfo on normal method calls (cache misses). snit 2 does so only for cget/configure. Harmless when no error is in progress.
          • Cause -
          • Prototype none (the attached 2005 patch is against snit 1.0 and is obsolete)
          • Suggested action close as Out of Date
          • Suggested ticket comment The described scenario no longer loses the error on Tcl 8.5+: trace callbacks run with the interp state saved. Closing.
          ⬜1280862fff2005profiler - errorInfo modifiable in enter- & leaveHandlerNOT-REPRODUCED (fixed by Tcl 8.5 semantics). Since 8.5, incr on an unset variable creates it, so the catch {incr ...} in enter/leaveHandler nev…

          Open ticket 1280862fff · Bug · profiler · filed 2005-09-02

          Full verdict

          NOT-REPRODUCED (fixed by Tcl 8.5 semantics). Since 8.5, incr on an unset variable creates it, so the catch {incr ...} in enter/leaveHandler never throws and errorInfo is left alone.

          • Module profiler 0.7
          • Tested catch {error orig}, then the first call of a profiled proc chain, then a check of $::errorInfo. It stays intact on 9.1, 8.6 and 8.5.
          • Cause the catches are now dead code (tcllib requires 8.5+).
          • Prototype none (not needed; the dead catch wrappers could be dropped as cleanup)
          • Suggested action close as Fixed (by Tcl 8.5+ requirement)
          • Suggested ticket comment Not reproducible with Tcl 8.5+: incr auto-creates the array element, so the catch never fires and errorInfo is untouched. tcllib now requires 8.5, so closing.
          ⬜1014795fff2004tcl crashes when I b64 encode filesNOT-REPRODUCED (5 MB random binary encodes and decodes correctly in 7 ms / 80 ms on 9.1 and 8.6, and 1.6 s / 1.1 s with the pure-Tcl path on 8.5). …

          Open ticket 1014795fff · Bug · base64 · filed 2004-08-23

          Full verdict

          NOT-REPRODUCED (5 MB random binary encodes and decodes correctly in 7 ms / 80 ms on 9.1 and 8.6, and 1.6 s / 1.1 s with the pure-Tcl path on 8.5). The original report was the 8.4-era pure-Tcl encoder hitting a memory limit on Windows 98.

          • Module base64 2.6.1
          • Tested base64::encode/decode round trip of 5 MB on 9.1, 8.6, 8.5
          • Cause n/a (on 8.6+ base64 uses binary encode base64)
          • Prototype none
          • Suggested action close as Out of date
          • Suggested ticket comment Cannot reproduce: a 5 MB file encodes in milliseconds on 8.6/9.x (native binary encode base64) and in about 1.5 s with the pure-Tcl code on 8.5. Closing as out of date.
          ⬜883573ffff2004ftp change directory problemsNOT-REPRODUCED (ftpd, 9.1 and 8.6: Cd/Pwd into a b, x{y}, d$[1], żółw all work)

          Open ticket 883573ffff · Bug · ftp · filed 2004-01-24

          Full verdict

          NOT-REPRODUCED (ftpd, 9.1 and 8.6: Cd/Pwd into a b, x{y}, d$[1], żółw all work)

          • Module ftp 2.4.14
          • Tested see above
          • Cause unknown. The report gives no example. Non-ASCII names go through the control socket's default (system) encoding, so they can fail with UTF-8 servers on non-UTF-8 systems.
          • Prototype none
          • Suggested action close as Works for me / insufficient information
          • Suggested ticket comment Directory names with spaces, braces, $, brackets and non-ASCII letters work on current trunk. Without an example of a failing name this can't be investigated further.
          ⬜782367ffff2003ftp.tcl: file list empty after file getNOT-REPRODUCED (List / Get / List sequences against ftpd, active and passive, 9.1 and 8.6)

          Open ticket 782367ffff · Bug · ftp · filed 2003-08-03

          Full verdict

          NOT-REPRODUCED (List / Get / List sequences against ftpd, active and passive, 9.1 and 8.6)

          • Module ftp 2.4.14
          • Tested List sub, Get sub/b.txt, List ., List sub. All correct, no empty or duplicated listing.
          • Cause the symptom (empty list, then double contents) matches data arriving after the control reply. That race was fixed by the WaitDataConn synchronizer in 863ac2e0d8 (2014-02-21, ticket eb0b15d598, ftp 2.4.13, tcllib 1.17).
          • Prototype none
          • Suggested action close as Fixed (probably by eb0b15d598), or works-for-me
          • Suggested ticket comment Not reproducible on current trunk. The late-data-connection fix for eb0b15d598 (ftp 2.4.13) addresses exactly this kind of listing that shows up one command late.