General

Add a note in the documentation about the Net::Citadel repository and home page at GitHub? Also include a reference to where it is available at the "gitpan" CPAN mirror at GitHub?

Add capability to use SSL encrypted communications with the Citadel server, perhaps using IO::Socket::SSL?

Change to using the jame@cpan.org address for Robert James Clay instead of the jame@rocasa.us address?

Other issues: (but uncertain what this is supposed to be about except that it is related to the POD on those lines): $ find -type f \( -iname '*.pl' -o -iname '*.pm' \) -exec podlint {} \; spurious '>' at ./lib/Net/Citadel.pm line 130 spurious '>' at ./lib/Net/Citadel.pm line 285 spurious '>' at ./lib/Net/Citadel.pm line 332 spurious '>' at ./lib/Net/Citadel.pm line 357 spurious '>' at ./lib/Net/Citadel.pm line 383 spurious '>' at ./lib/Net/Citadel.pm line 414 spurious '>' at ./lib/Net/Citadel.pm line 437 spurious '>' at ./lib/Net/Citadel.pm line 468 spurious '>' at ./lib/Net/Citadel.pm line 514 spurious '>' at ./lib/Net/Citadel.pm line 569

$ env PERL5OPT=-m-lib=. perlcritic --noprofile -1 . 2>&1 | grep -vF 'No perl files were found.' ./Makefile.PL: Code is not tidy at line 1, column 1. See page 33 of PBP. (Severity: 1) ./Makefile.PL: No package-scoped "$VERSION" variable found at line 1, column 1. See page 404 of PBP. (Severity: 2) ./Makefile.PL: Code before strictures are enabled at line 5, column 1. See page 429 of PBP. (Severity: 5) ./Makefile.PL: Code before warnings are enabled at line 5, column 1. See page 431 of PBP. (Severity: 4) ./Makefile.PL: Heredoc terminator must be quoted at line 32, column 12. See page 64 of PBP. (Severity: 3)

Constants

There are some numbers being used that are actually constants; change to using Readonly to define them for use in the code, unless there are other standard constants that can be used, like those for the flock operations.

A "000" string constant is used quite often by Citadel but is not defined in the module.

There is a set of follow up reply codes noted in the documentation that are not yet in the module.

Add a link to the online documentation for the status codes: http://www.citadel.org/doku.php?id=documentation:appproto:statuscodes

A "NO_SUCH_USER" return code is mentioned in the command documentation (for the USER command, for instance,) but the module does not currently appear to define it.

Add a paragraph to the documentation after the CONSTANTS header.

Add a paragraph to the documentation after the 'Results Code' header, perhaps including a link to the online status codes page in Citadel Documentation.

Add a paragraph to the documentation after the 'Room Access' header, perhaps including a link to the online page in Citadel Documentation that refers to them.

Functions

There does not seem to an explicit function to close the connection; i.e., close the socket being used to communicate with the server.

Create an internal function to handle LISTING_FOLLOWS return lines from the server? Perhaps called something like "_get_listing"?

"assert_room" function

Is the "assert_room" command coming up with the correct floor information?

Is the "assert_room" function sending the parameters for the CRE8 command properly? (There is a note in the code indicating an uncertainty about what should be sent.)

"retract_room" function

In the existing test code for the "retract_floor" function, there is a comment saying "CITADEL BUG"; investigate if it an issue with the function itself, an actual issue with Citadel, and/or just something to do with the test script. The documentation for the function itself refers back to Citadel v7.20, so there may not be any issue with it now.

function "new"

The line just before the 'return $self' line actually has two commands on it, not just one. The second one is just "<$s>;" and that has the comment "# consume banner". That 'banner' is something like the following:
"200 <systen name> Citadel server ready." Instead of not doing anything with that, save it somewhere? At least the <system name> part of it? (The first command on that line just establishes the $s variable for that read.) At least, put them on separate lines.

Resolve the following from the Perl Critic testing: Always unpack @_ first at line 308, column 1. See page 178 of PBP.(Severity: 4)

"change_user" function

Always unpack @_ first at line 637, column 1. See page 178 of PBP. (Severity: 4)

"citadel_time" function

The current "citadel_time" function only returns the first two parameters from the TIME command: "1347624956|-14400". The Citadel TIME command itself actually returns a line like this: "200 1347625545|-14400|1|1347537300", with the '200' being the OK code and the rest being the four fields that it returns (which also needs to be documented). So as currently written, the function does not return the daylight savings time indication and the actual citadel server start information. Function first needs to be changed to at least return all parameters.

Rewrite the "citadel_time" function to unpack the parameters that the TIME command returns when it is successful, and then return them to the calling program in a hash? Could also then return the hash with a key named 'error' or 'Error' if it is not successful. Update its POD to reflect the changes.

"citadel_echo" function

It appears to test if what was sent, was actually received back but it only returns true if it does not fail because there is no match. Add an explicit return of the value returned from the ECHO command?

Documentation

Add something like Net::Citadel::Tutorial for examples of the use and configuration of the module?

Testing

Change the default name for the Citadel administration user to 'admin' in the t/test.yaml file? It does seem to be the default name in the Citadel setup.

Will the differences with the newer Citadels justify or neccessitate checking the version of the Citadel being tested and doing different tests depending on that version?

Change the names of the test floors and rooms to something like "Test Floor" and "Test Room"?

Update the testing for the "citadel_time" function to at least check the number of parameters returned? Three of the parameters are Unix timestamps; validate those in some way? The other parameter being returned is a Boolean and is used to indicate Daylight Savings Time and should be a '0' or a '1'.

When testing for a floor, it looks for "Main Floor": that exists on a default install but may not be on an working system. Make it another configuration item, which if not configured would default to "Main Floor"?

Separate the testing to: functions that do not require a log in, those that do require a log in, those that are read only, those that write to the server. Or use the separation given in the documentation for the different Sections for the commands?

Since Config::YAML is already a build-depends and is used by the main test script, add a "Testing" section of some sort to the test.yaml file and check that for the various options for testing instead of just attempting to reach a Citadel server?

Instead of having the debugging 'warning' lines commented out, use a DEBUG environment variable and/or a test.yaml configuration item for it. Same for the 'use Data::Dumper' line itself.

Use Config::YAML::Tiny instead of Config::YAML? Does not appear to be in Debian as yet, although YAML::Tiny is. And it has not been updated since 2010.

Use something like Test::MockObject to do at least basic testing of the various functions? And/or Test::TCP? Could use Test::TCP::CheckPort to first check if there is even something up on the standard Citadel port 504.

Testing of the citadel_info function currently just checks the number of information lines returned by the server. That could change over time and different Citadel versions. Make that a testing configuration item?

Do more extensive checking of what is returned by the citadel_info function?

Instead of having the debugging 'warning' lines commented out, use a DEBUG environment variable and/or a test.yaml configuration item for it. Same for the 'use Data::Dumper' line itself. These are where they were found: $ grep -r warn * lib/Net/Citadel.pm:use warnings; lib/Net/Citadel.pm: }; warn $@ if $@; lib/Net/Citadel.pm:#warn "_floors $_"; lib/Net/Citadel.pm:#warn "looking for $name rooms ". Dumper \@floors; lib/Net/Citadel.pm:#warn "found floor: ".Dumper $floor; lib/Net/Citadel.pm:#warn "processing $_"; lib/Net/Citadel/ToDo.pod:Instead of having the debugging 'warning' lines commented out, use a DEBUG t/00pods.t:use warnings; t/02critic.t:use warnings; t/01citadel.t:#warn Dumper \@floors; t/01citadel.t:#warn Dumper \@floors2; t/01citadel.t:#warn "before assert". Dumper \@rooms2; t/01citadel.t:#warn "after assert". Dumper \@rooms3; t/01citadel.t:#warn "after retract". Dumper \@rooms4;

These Perl::Critic result needs further investigation, but possiable good reference is https://rt.cpan.org/Public/Bug/Display.html?id=42486, which notes the policy that is finding these "doesn't take Readonly into account" but it does also note how to exempt 'Readonlhy':
- Subroutine "Readonly" is neither declared nor explicitly imported at line 71, column 1. This might be a major bug. (Severity: 4)
- Subroutine "Readonly" is neither declared nor explicitly imported at line 85, column 1. This might be a major bug. (Severity: 4)
- Subroutine "Readonly" is neither declared nor explicitly imported at line 94, column 1. This might be a major bug. (Severity: 4)
- Subroutine "Readonly" is neither declared nor explicitly imported at line 104, column 1. This might be a major bug. (Severity: 4)
- Subroutine "Readonly" is neither declared nor explicitly imported at line 114, column 1. This might be a major bug. (Severity: 4)
- Subroutine "Readonly" is neither declared nor explicitly imported at line 124, column 1. This might be a major bug. (Severity: 4)
- Subroutine "Readonly" is neither declared nor explicitly imported at line 134, column 1. This might be a major bug. (Severity: 4)
- Subroutine "Readonly" is neither declared nor explicitly imported at line 143, column 1. This might be a major bug. (Severity: 4)
- Subroutine "Readonly" is neither declared nor explicitly imported at line 153, column 1. This might be a major bug. (Severity: 4)
- Subroutine "Readonly" is neither declared nor explicitly imported at line 164, column 1. This might be a major bug. (Severity: 4)
- Subroutine "Readonly" is neither declared nor explicitly imported at line 177, column 1. This might be a major bug. (Severity: 4)
- Subroutine "Readonly" is neither declared nor explicitly imported at line 186, column 1. This might be a major bug. (Severity: 4)
- Subroutine "Readonly" is neither declared nor explicitly imported at line 195, column 1. This might be a major bug. (Severity: 4)
- Subroutine "Readonly" is neither declared nor explicitly imported at line 204, column 1. This might be a major bug. (Severity: 4)
- Subroutine "Readonly" is neither declared nor explicitly imported at line 215, column 1. This might be a major bug. (Severity: 4)
- Subroutine "Readonly" is neither declared nor explicitly imported at line 227, column 1. This might be a major bug. (Severity: 4)
- Subroutine "Readonly" is neither declared nor explicitly imported at line 235, column 1. This might be a major bug. (Severity: 4)
- Subroutine "Readonly" is neither declared nor explicitly imported at line 243, column 1. This might be a major bug. (Severity: 4)
- Subroutine "Readonly" is neither declared nor explicitly imported at line 251, column 1. This might be a major bug. (Severity: 4)
- Subroutine "Readonly" is neither declared nor explicitly imported at line 259, column 1. This might be a major bug. (Severity: 4)
- Subroutine "Readonly" is neither declared nor explicitly imported at line 267, column 1. This might be a major bug. (Severity: 4)
- Subroutine "Readonly" is neither declared nor explicitly imported at line 277, column 1. This might be a major bug. (Severity: 4)

These Perl::Critic result needs further investigation:
- Always unpack @_ first at line 308, column 1. See page 178 of PBP. (Severity: 4)
- Always unpack @_ first at line 637, column 1. See page 178 of PBP. (Severity: 4)

This Perl::Critic result needs further investigation:
- Subroutine "SOCK_STREAM" is neither declared nor explicitly imported at line 317, column 58. This might be a major bug. (Severity: 4)