Clarification of Answer by
bookface-ga
on
30 Dec 2005 02:01 PST
15. Uninstall does not proceed cleanly if TcD Scheduler is open, but
gives no warning. (i.e. it cannot completely remove files, but does
not give a message before starting uninstall to the effect of "Please
close all running instances of TcD and TcD Scheduler.")
16. Issues already discussed:
a) MSVCP60.DLL, a vital library for TcD, is still not included in the install.
b) Scheduler section of help still does not follow naming
convention of remainder of Help ToC.
c) Pressing Control-H on an empty line in Command window still
freezes the window.
d) Still no message about downloads that are not successful (i.e. 404 errors).
e) No message or download about forbidden files (authorization failed), either.
I suggest a log file/error window for d & e, but that can wait
till the next release.
17. Pressing F1 in the Scheduler or Command windows does not work.
(Main window is fine.)
18. After downloads completed, status bar switches between two messages:
Downloaded # files, #.##MB, in #:##. Click magnifying glass to view.
Files: 0 Total size: 0 bytes
I think the second message should begin "Remaining Files:" to
avoid confusion, or simply be hidden if none are left to download.
19. In the help for ls/list/show command, I think it would be nice to
mention that it does not reflect list of URLs loaded in the main
window, but rather a separate list specific to the Command window.
This is documented elsewhere, but the ls help was the first place
I looked when it didn't do what I expected.
I assumed this was broken at first before I read through the help
comprehensively.
20. Missing commas in help:
Open a file: ".flux___,___ it will be treated..."
Begin downloading: "... small files____,____ you may not even notice..."
"...downloaded____,____ your list will be cleared."
Pause downloading: "...progress___,___ this button will be grayed out."
Clear list: "... when you click this button___,___ downloading will
stop immediately..."
View files: "files since opening TcD___,___ this button will..."
Edit current file: "without any .tcd file open___,___ you will see a
blank document"
Commands, Add: "To add URLs to this list___,___ use the add command"
Commands, Get: "If used with an optional size specification___,___ it
will only get those files..."
Resources, Command Window shortcuts: "...use one of these
shortcuts____,____ you may need to click..."
21. After loading a list, View files button still opens download
directory for previous list. (It updates after download is complete.)
22. Non-HTML redirects via 301/302 do not give proper extension.
Example: bookface.frihost.com redirects to a text file, saved file is
"1 -", as opposed to "1 -.txt"
23. All redirects do not pick up proper filename (before extension).
Example: google.com redirects to google.com/index.html, but is saved
as "2 - .html" instead of "2 - index.html"
24. Introduction to the Main window:
"Let's take a look at it the main window from top to bottom."
needs to at least have the word "it" removed, and I recommend:
"This introduction will introduce the components of the Main
window from top to bottom."
Just above the Status bar section:
"When __you___ are download__ing__ files, the URL currently being
downloaded will be shown just below the Progress bars."
25. Reload file button is not grayed out (as claimed in help) when no
downloads in list, or at least after clearing with Clear list button.
26. "lose" should replace "loose" in "Command window" button
description subsection of Main window.
27. In same section, I recommend inserting a paragraph break before
"Exiting the TcD application...", as the two notes currently appear to
be related.
28. set CustomHeaders section: "Would make it appear" doesn't need capitalization.
29. keeph, keept, delh, delt: I assume the h and t in these commands
stands for head and tail, but this is not mentioned.
30. Your regular expressions do not handle * properly:
TcD -> Flux (0)> add images.google.com
TcD -> Flux (1)> make
Searching (1/1) "http://images.google.com"... 9 found.
TcD -> Flux (9)> list
https://www.google.com/accounts/Login?continue=http://images.google.com/&hl=en
http://images.google.com/intl/en/options/
http://images.google.com/advanced_image_search?hl=en
http://images.google.com/preferences?hl=en
http://images.google.com/help/faq_images.html
http://images.google.com/ads/
http://images.google.com/services/
://www.google.com/intl/en/about.html
http://images.google.com/intl/en_ALL/images/images_hp.gif
TcD -> Flux (9)> keep *.html
TcD -> Flux (9)> keep (*.html)
TcD -> Flux (9)> keep (.html)
TcD -> Flux (2)>
...
TcD -> Flux (20)> del *jpg
TcD -> Flux (20)> del (*jpg)
TcD -> Flux (20)> del (.jpg)
TcD -> Flux (0)>
Nothing is done with any regex containing *
31. Framesets are not parsed by make (which shouldn't particularly be
the expected case, according to the documentation; but I would expect
this behavior as part of make's builtin regex.) They are parsed by
spider, but this is undocumented.
32. Neither spider nor make handles for
32. set DirSlashes, set NameTemplate, and set Proxy are not documented.
33. SeqWarningSize seems badly named, as it is more a limit than a warning.
34. Comments in the command window (# blah) are undocumented.
35. "config blah" should give help for config, but gives batchcurrent help instead.
36. Help for fixparents states that it will change dir1/.. into dir1/,
what it should read is that it will change it to
http://www.example.com/pic1.jpg or alternatively that the orignal url
was dir1/dir2/../pic1.jpg.
37. I understood what you meant, but I recommend changing the
description for fusker to read:
* Strings in curly brackets are treated as a comma-separted list of
string options, and will be interpreted to be one URL for each option.
Also consider using the word "converted" or "evaluated" instead of
"interpreted", although any one is fine.
38. I'm pretty sure the stuff about the optional size specification in
the get command is not true. You probably should be calling "size
[getarguments]" before executing the get command, if it has any
arguments.
39. I'm pretty sure this should never happen, but regardless I don't
think the right behavior for
add http://www.example.com/dir1/dir2/dir3/../../../../../../../pic3.jpg
fixparents
should result in /pic3.jpg. Probably the "http://domainname" part
should always be preserved.
40. Help for "header" should read "sent out" instead of "send out"
41. "help help" does not, as it claims in the help for "help", result
in help for "help", but rather gives help for the Command window in
general.
42. cd, get, title, cookies, config, and possibly other items that do
not (or should not) modify the list, directly or indirectly, still
show up in the history (while the history help claims they do not.)
You should probably add more commands to the ignore list, and also
append the help text to clarify that "commands that modify or may have
an effect on the list in the future" will be in the history list. Note
that this is mentioned more again (but more clearly) in the Notes for
the history command.
43. Password command does not work for me on my testbed
http://adam:eve@bookface.frihost.net/testes/
TcD -> Flux (0)> add http://bookface.frihost.net/testes/
TcD -> Flux (1)> spider
Searching "http://bookface.frihost.net/testes/"...0 found.
TcD -> Flux (0)> add http://bookface.frihost.net/testes/
TcD -> Flux (1)> password bookface.frihost.net adam eve
TcD will use adam for bookface.frihost.net now.
TcD -> Flux (1)> spider
Searching "http://bookface.frihost.net/testes/"...0 found.
TcD -> Flux (0)> add http://bookface.frihost.net/testes/
TcD -> Flux (1)> spider
Searching "http://bookface.frihost.net/testes/"...0 found.
TcD -> Flux (0)> add http://bookface.frihost.net/testes/
TcD -> Flux (1)> make
Searching (1/1) "http://bookface.frihost.net/testes/"... 0 found.
TcD -> Flux (0)> add http://bookface.frihost.net/testes/outie.html
TcD -> Flux (1)> password show
bookface.frihost.net - adam:eve
TcD -> Flux (1)> get
TcD -> Flux (1)> get
TcD -> Flux (1)> password http://bookface.frihost.net adam eve
TcD will use adam for http://bookface.frihost.net now.
TcD -> Flux (1)> get
TcD -> Flux (1)> spider
Searching "http://bookface.frihost.net/testes/outie.html"...0 found.
TcD -> Flux (0)> add http://bookface.frihost.net/testes/
TcD -> Flux (1)> spider
Searching "http://bookface.frihost.net/testes/"...0 found.
TcD -> Flux (0)> add http://bookface.frihost.net/testes/
TcD -> Flux (1)> make
Searching (1/1) "http://bookface.frihost.net/testes/"... 0 found.
TcD -> Flux (0)> password show
bookface.frihost.net - adam:eve
http://bookface.frihost.net - adam:eve
TcD -> Flux (0)>
44. Empty list not properly saved.
TcD -> Flux (0)> save blabber
Saved list to "blabber".
TcD -> Flux (0)> load blabber
TcD -> Flux (1)> ls
TcD -> Flux (1)>
45. "saveflux test.flux" crashes TcD if it cannot combine any entries.
Also, as far as I can tell in my "black box" testing, it does not work
to combine {string1,string2} entries.
46. TcD loads with the previously loaded list, I did not notice this
in the documentation if it is indeed there.
That's all I found, and from my minimal experience doing QA for a few
months for two companies, I would deign to say that's definitely not
too bad for freeware. Let me know if you have any further questions
about the issues I've stated, I've continued numbering my points so
you can refer to them easily.
Good luck with your release, and thank you for choosing Google Answers!
- bookface-ga