.net 1999 December/netCD65.iso/pc/Software/Top20/FileHound/filehound.exe/FILEHOUND.TXT

Text File | 1998-05-17 | 27.4 KB | 673 lines

QUICK-START GUIDE
--------------------

To download a file with FileHound, do one of the following:

  * Drag and Drop a web link onto FileHound's Window from
    your web browser.

  * Right Click on a web link and choose "copy to clipboard"
    FileHound will grab the link from the clipboard, you
    only need to tell it to START downloading.

  * Type the URL into FileHound's edit window.


To download multiple files from one page or FTP site,

 1. Drag and Drop the FTP directory or web page onto
    FileHound's window and let FileHound retrieve the
    listing of files.  (Or, type the FTP directory
    name into FileHound's edit window.)

 2. Select the listing URL in FileHound, right click and
    select "Extract URLs".

 3. Choose the items you wish from the extract list,
    starting the download process by double-clicking.



FileHound 1.2b Documentation (Change log at end of file)
----------------------------------------------------------


WHAT DOES FILEHOUND DO?

One thing.  Downloads files from the Internet.  That's it.
It's designed to be extremely foolproof and requires absolutely
no setup.  If you are successfully running a web browser,
FileHound will work out of the box (tech detail: any system
that supports WinSock32 will run FileHound).



HOW DO I MAKE IT WORK?

You have to obtain a file's internet name, commonly referred
to as its URL, and give it to FileHound.  FileHound is not a
plug-in, and it's not a web browser.  You'll use your web
browser as you always have, but when you come across a file you
want to download, instead of clicking on the link in the browser,
you'll instead COPY the link to the clipboard, then PASTE the
link into FileHound's URL field.  Then click on download, and
away you go!  And you'll never have to worry about another
broken download.



WHAT DO I HAVE TO SET UP?

Nothing.  The configuration screens, options, multiple tabs and
so forth are merely auxiliary settings, archive listings, etc.
to extend the basic "download a file" functionality far beyond
what you are used to having in your web browser.

FileHound does not make any changes to your system, and it does
not modify any system settings or move any files.  Feel free,
then, to experiment with FileHound and see what each feature
does.  You can't hurt anything just playing around with it.



HOW DO I UNINSTALL FILEHOUND?

To keep the download size of FileHound smaller, we decided to
omit an uninstall program.  To delete FileHound, do two things:
Delete the \Program Files\FileHound directory.  And, delete the
\Windows\Start Menu\Programs\Hound Agents directory.  Note that
by default, FileHound puts downloaded files in a subdirectory
called download within \Program Files\FileHound, so if you want
to keep any of your downloads, move them before you uninstall
FileHound.



EXPLAIN THE BANDWIDTH SETTINGS TO ME.

Usually when you select a file to be downloaded, your FTP
program or web browser will retrieve the file as fast as
possible.  That's nice if you're in a big hurry, but the cost
is that your further web browsing is slowed down, because your
modem is very busy retrieving all the file data.

FileHound lets you "throttle" its downloading to only use
a small part of your modem's capacity for retrieving files.
By using slow downloads, you can keep surfing the web at near
full speed while your downloads proceed in the background.
They will take longer, yes, but you can keep working online.

FileHound has two speed modes, normal (or slow), and FAST.
FAST mode is typically used when you want FileHound to soak
up all of your bandwidth; for example when you're reading
email or chatting, or away from your desk.  Normal mode is
used when you want to continue surfing the web without
getting lagged by a download.

FAST mode is enabled when you push the GO FAST button.  Other
options, including telling FileHound how much data you want
it to retrieve in each mode are available on the Bandwidth
settings page.



WHAT ABOUT VIRUSES?

FileHound recognizes that files you retrieve from the Internet
are by their nature unsafe.  We have built in external virus-
check support into FileHound.  Although FileHound does not have
a virus check program built in, it can work with any virus
check program you have installed on your system.  This will
require a little bit of research and set-up on your part, but
we've made it as easy as possible.  The Virus Check settings
page contains all the information you'll need to fill in to
enable FileHound to start automatically checking your
downloads for safety.  We tested it with both McAfee VirusScan
and Symantec's Norton Anti-Virus, and it works like a charm.



GOT ANY BELLS AND WHISTLES?

Oh my yes.  FileHound can be very noisy if you wish!
Open the Sound settings page and you can assign your own
favorite windows WAV files to different circumstances in
FileHound.  It's very easy, and adds a lot of fun to your
downloading!



HOW ABOUT A NIGHT-LIGHT?

FileHound has that, too.  If you're like most people, you'll
want to use FileHound to download multiple files when you're
away from your desk or sleeping.  By using the "Good Night"
button, you tell FileHound to go ahead and shut down your Internet
connection when all files have been transferred.  This can save
money if you pay by the hour for your connection.  There's also
an option on the Miscellaneous settings page to tell FileHound
to actually shut down your PC when it's done downloading.
(Both of these options give you a 60 second warning in case
you decided to stay awake, and you can countermand the order
any time.)



ARE THERE ANY MORE FEATURES?

Dozens, but it would be exhausting for both of us if we were
to list them all out.  FileHound has been honed over nearly
half a year of interactive testing by hundreds of beta
participants, and has many time-saving features and lots of
behind-the-scenes magic to try and eliminate as much
complexity as possible from downloading from the Internet.
Odds are any request you might have or feature you'll need
is built into the program already.

If you're into buzzwords, here are some of the things we've
added to FileHound to make your life easier: HTTP Basic
authentication, FTP and HTTP file resume, URL login/password
decoding, filename defaulting, automatic HTTP redirection,
per-server automatic login name and password for FTP
connections, UNIX->MSDOS text file translation, overwrite
protection, power-fail download resumption, full session
logging, file launching, perpetual URL archive with a
re-download feature, multiple simultaneous downloads (no
built-in limit), very accurate estimated download timer,
and on, and on, and on.

But remember, FileHound is first and foremost EASY TO USE.
These features sit quietly in the background until you need
them, and then they're right there by your side.  So all
of these options and settings work for you by making downloads
easier and more reliable, instead of against you by adding
confusion and hard to understand dialogs.  In every case,
FileHound will guess as best as it can in order to provide
you with the perfect download settings.



HOW CAN I MAKE USING FILEHOUND EVEN EASIER?

In version 1.2, FTP and HTML URL extraction was added, and
support for downloading FTP directories was added.  When
you use the Extract URLs command against a web page or FTP
directory, FileHound will create a list of all files in that
page or directory, and add them to the Clipboard/Extract
listing.  From there, you can sort them and schedule
downloads of ones you want.

New to version 1.1a is drag and drop support.  Using Internet
Explorer or Netscape, you can simply drag a highlighted URL
from a browser page to FileHound's window and release it, and
that URL will be automatically queued up.

Another way to easily get URLs into FileHound without typing
is to enable clipboard sniffing.  In the miscellaneous settings
page, there is an option you can turn on that will make
FileHound monitor your clipboard for URLs.

When you copy one, FileHound will ask right away if you'd like
to download it.  There's even an option to just automatically
say "yes" to that question, so starting downloads in FileHound
is actually easier than in your web browser -- you don't even
have to hit OK, just right click on the file link and copy
the link/shortcut to the clipboard.



WHAT IF SOMETHING GOES WRONG?

Even though this release has undergone a very long testing cycle
and there are no remaining outstanding issues, there may yet be
problems.  Your comments and feedback are very much desired,
particularly on these points:

  o  Bugs, crashes, erratic behavior.  To report a bug, please
     list the steps required to reproduce it.  Be specific!

  o  Downloads that break but shouldn't.  The goal is to make
     FileHound be able to download all URLs that Netscape and
     Internet Explorer can using these protocols:  ftp:// http://
     and  file://   If you can't download a URL you think you
     ought to be able to, please forward both the URL and the
     session log from the failure.  I'll fix it.



WHAT HAVE YOU GOT PLANNED FOR FUTURE RELEASES OF FILEHOUND?

If any of these options are something you want or need, please
write in and express your support.  Also, if there are any other
ideas you have for FileHound, let us know!

  o  Proxy support, including "passive" FTP, CERN, SOCKS, TIS and
     other firewall support mechanisms.

  o  HTTPS/SSL secure web URLs.

  o  gopher:// and other URL protocol types.

  o  Plug-In Integration with browsers.



WHAT IS FILEHOUND'S COMMERCIAL STATUS?

FileHound is truly a "shareware" program instead of
crippleware/trialware.  You have full functionality without
any time limit, and it is left to your personal sense of honor
whether or not to remunerate us for our efforts.  If you
honestly cannot afford $15 for FileHound--remembering that
if you use it all the time, you're certainly getting a lot
of value for it--or simply cannot obtain the necessary
currency for payment, please continue to use it in good
health.  Send us a note, let us know you're out there!



HOW MUCH IS IT TO REGISTER?

FileHound is $15US to register, and you can register over the
phone, via email, fax, or secure website.  We take MasterCard,
Visa and Discover as well as their international variants, and
you will receive a registration key either instantly (phone
orders) or within two business days (web or email orders).
The registration key will remove the very occasional reminder
screens and put your name on the title bar of FileHound to give
you a "warm fuzzy" feeling about having registered!



WHAT DOES THE LICENSE GRANT ME?

Your license will entitle you, and only you, to use FileHound
on as many computers as you like.  Once registered, your name
will appear in the title bar of the application.  So, if you
are using FileHound and someone else's name appears in the
title bar, you should register FileHound immediately in your own
name.  We do NOT sell blanket licenses to businesses. Each user
must be registered individually.  Discounts are available for bulk
purchases (inquire).



HOW DO I CONTACT YOU?

Phone:  [phone] - DO NOT CALL FOR TECH SUPPORT OR BUGS
Email:  [e-mail]
  Web:  http://www.nexi.com/albinofrog (secure)

 Mail:  Albino Frog Software, Inc.
        Post Office Box 578
        Port Salerno, FL  34992  USA

 Note:  Do not CALL for support, or to report bugs.  My wife
        takes the orders during the day, and she will NOT be able
        to answer your questions.

        Use EMAIL for all technical communications.



HOW DO I ORDER?

FileHound is $15 US for a single user license.  When you
register, you will NOT receive a disk.  Keep up with the latest
version of FileHound by downloading it from our website.  You
will receive a set of registration keys to unlock your
personal copy of FileHound.  No disk.  No manual.  Just keys!

You may order by Check or Money order through the mail, to the
above address.  Payment must be in US funds drawn on a US bank.
For this reason, we recommend that non US orders use a credit
card.  If you don't have a credit card and are feeling lucky,
place US currency in a sturdy, opaque envelope and mail that
instead of a check.  We're honest, but can't guarantee the same
of the postal service!  Include your email address on all mailed
orders so we can email you registration keys as soon as your
funds clear our bank.

You may order by Credit Card through the mail, over the phone,
via email, and on our secure web site.  Your account will be
charged $15US.  Currency conversion fees, if applicable, are your
responsibility.

You will receive a registration key for FileHound when your order
has been processed.  If you supply an email address, your
registration key will be sent via email with a note confirming
your order.  Keep up with the latest version of FileHound by
downloading from our website. Upgrades are free within the current
major revision (1.x), and can be downloaded at any time.



IS THERE ANYTHING ELSE?

Spread the word about FileHound!  Tell your friends, your
co-workers, your neighbors.  Put a link to it on your
website, and recommend your users try it for downloading your
own files.  Upload it to shareware BBSs.  Mention it in
Usenet posts.  Get the word out: There's no need to ever
suffer broken download again.



WHAT IS FILEHOUND'S DEVELOPMENT HISTORY?  (last question)


Fixed for 1.2d 5/17/98
------------------------------------------------------------
 - FTP Resumes occasionally broken (Oversized) due to FTP
   enhancements.  Whoops.


Fixed for 1.2c 5/15/98
------------------------------------------------------------
 - Cleaned up install script so it doesn't run twice on NT
 - Made upgrade procedure a bit nicer.



Fixed for 1.2b 5/13/98
------------------------------------------------------------
 - We now write out the state file in one controlled, non
   thread conflicting place. And on close, display a small
   dialog box when writing it out to remind you that FH
   isn't quite done closing when the window goes away.
   Also greatly improved speed of saving the state file
   itself.  In effect, this makes FH seem much more
   responsive.
 - Fixed problem with using "Minimize On Startup" in short
   cuts.  FH now can start right to the tray.
 - Flush the file buffer all the time, and close/reopen it
   every thirty seconds.  OS Robustness fix.  (Windows never
   crashes, right?)


Fixed for 1.2a
------------------------------------------------------------
 - Silly bug with URL redirection on WINZIP.COM.  Whoops.
   Guess we'll test that automatic upgrade thingy!


New in 1.2
------------------------------------------------------------
 - Added automatic upgrade feature.  FileHound will check
   once a month to see if a new version is available.
   (This can be turned off for those of you in paranoia
   mode, see Settings/Misc for details.)
 - Added minimize-to-system-tray option.  Defaults on but
   can be switched off in the misc properties tab.
   The "FH" in the icon is green while FileHound is
   downloading, white when idle.  (Is there any point in
   making this an OPTION?  Does anyone NOT want it to
   minimize to the tray?)
 - Can now download FTP directories and extract URLs
   from web pages.  All extracted URLs are added to the
   clipboard list so you can choose among them and start
   from there.
 - Green LEDs now vary in brightness based on how much
   bandwidth is being absorbed by that file.  A user
   wanted bps counters per file, and this is my response.
 - Added dialog for deleting all files rather than
   just one by one.
 - FTP downloads will now try several variations of the
   download filename in order to clean up funky filenames.
   It tries with and without a leading slash, then
   converts % codes into regular characters, with and
   without a leading slash.  Nothin' like being thorough!
 - GoodNight bug fixed for good.  I hope.  (Technical
   explanation at end of file -- I'd love to hear feedback
   on this if anyone knows what the real deal is!)
 - Fixed bug in virus settings - no matter what you
   chose, filehound would always delete an infected file.
   Must be the Surgeon General setting.  Now it listens
   to the options whether or not to delete it.
 - Modified tab names for consistency.
 - Clipboard extractions are now tested against all file
   categories for duplicates, not just active files.
 - All extracted objects now held in a separate internal
   folder, too -- this helps the speed of list refreshing.
 - Increased download buffer size to make filehound faster
   under high speed connections.
 - Slowed down list updating during downloads to stop
   chewing up CPU cycles unnecessarily when list is big.
 - Enhanced generation of default filenames.
 - ENhanced handling of redirects.
 - Many robustness enhancements.  Hopefully no more
   crashes while deleting large numbers of items.
   Also FH should be handling larger lists better too.
 - Added # and & to illegal filename character list.
 - Now handling # suffixed URLs correctly.
 - Added Referer: line to http requests.  We're telling
   the web server that it referred itself.  This gets
   around annoying webmasters like CDROM.COM's who don't
   like anonymous requests.


Tweaks and bug fixes for 1.1a (released only to testers)
------------------------------------------------------------
 - Drag and Drop support for URLs from IE and Netscape
   (This added 22K to the program size.  Ya know, when
   you go to think about program bloat, you can aim your
   spears right at Redmond.  Thanks for OLE2, Bill!)
 - Consistently deleted transfer logs from items going
   into the archive list.
 - Fixed problem related to Good Night crashing Filehound
   on systems with IE4 installed.  The problem was not
   reproducible in IE3 or Netscape installations, and could
   not be reproduced in Debug mode, either.  It cleared up
   when compiler optimizations were changed, leading me
   to believe the problem is caused by some DLL foible not
   handled well by VC++ 4.2 (I don't have VC5).  Anyway,
   it's working again.  I'm not happy with the solution
   but it'll have to do for now.


Bugs Fixed in 1.1
------------------------------------------------------------
 - FTP downloads on multihomed machines (modem & network
   card) now work.
 - FileHound will now run on NT machines without RAS
   installed, but you lose the GoodNight button
   functionality.  Go install RAS!


Goodies and updates to achieve 1.0
------------------------------------------------------------
 - Made authorization-required files start up more quickly.
 - Changed double-click activity on authorized screen to
   "edit properties" rather than "queue."
 - Added "Offline" button in bottom toolbar.  Pushing
   this button will cancel all downloads and prevent any
   further downloads from being started.  When released,
   file transfers resume where they left off.  Useful
   when you want to set up a bunch of downloads in the
   queue but would rather they didn't start right away,
   or when working away from a modem, or other situations.
   Someone requested it.
 - Added "Good Night" button in bottom toolbar.  When
   Filehound is done downloading (queue empty) and this
   button is depressed, FileHound will disconnect your
   modem (and any other RAS connection you have active,
   so be careful if you're multiply connected to the world).
   If you like, Good Night will turn off your computer, too.
   A dialog box is presented at time of shutdown giving
   you 60 seconds to countermand the order, in case you
   decided to stay up late anyway.
 - Added ability to rename if you are downloading over a
   file that already exists.
 - Changed file open flags to permit you to inspect
   downloaded files during the download.  (In other words,
   file sharing is now enabled.)
 - Added Auto-Start for Clipped URLS (Misc Options)
 - Changed connection limit settings to make more sense.
 - Sparked by release of Hexen II demo, added Rude Retry
   feature which can only be enabled on a per-file basis.
   This makes retries retry after only one second rather
   than progressively slower, which is more sane.  Use
   with discretion, please, and pay attention to your error
   codes to ensure it's a busy server and not a bad
   password that's making your download fail.
 - Fixed problem with redirects on services such as ZDnet
   which use case-sensitive filenames (which FH converted
   to all lowercase unnecessarily in internal processing.)
 - Made filehound's datafile 8.3 compliant so it can be
   stored on non long-filename servers.  It is now called
   fhound.dat instead of filehound.dat.  On starting 1.0
   in an upgrade, the previous file will be converted.
 - You can now stop a transfer that is in the retry cycle
   without having to start it first.  (wierd, huh)
 - Added option to show transfer speeds as Bytes/Sec rather
   than K/Minute (user request).
 - Eased waaaay back on the hostility about registering the
   product.  Instead of shutting down after 60 days, the
   program will only gently remind you every tenth startup
   if it is still unregistered.  Also, all functionality is
   enabled for both registered and unregistered versions.

   Note: A request was made to allow URLs with spaces.
   According to RFC1738, the URL standard specification,
   spaces are not permitted in URLs.  Neither are any
   control characters or #, <, >, or ".  FileHound
   follows these rules explictly.  If you encounter a URL
   with an illegal character in it, please alert the owner
   of the file.  Spaces can be encoded in a URL as %20 if
   needed, and any other unsafe character can be represented
   as a similar escape sequence (see RFC1738 for details).


Fix for .95
------------------------------------------------------------
 - Windows NT web server didn't like it when FileHound sent
   an authorization header when it wasn't needed.  Now
   Filehound must first try all http downloads without the
   auth header.  if the auth fails for this anonymous
   connection, it will retry, but next time will send
   the header.  if it fails again, it truly fails.


Fixes and Enhancements for .94
-----------------------------------------------------------
 - Audio Feedback support (new tab in settings)
 - Virus Scanner support (new tab in settings)
 - "Clip-And-Go" automatic downloading when things clipped
 - Very strange bug wrt saving the datafile just after
   downloads.  May not be fixed -- if you get an error
   message mentioning Exceptions and Error Codes, please
   email me!
 - FTP downloading trouble on AOL fixed by not using the
   CWD commands anymore.  We simply request a file by
   its full path.
 - default filenames with characters like ? * and \ are
   truncated to create a valid save file.
 - HTTP redirected URL save filenames are now re-genned
   automatically, fixing problems with sites like
   www.slaughterhouse.com that uses a CGI redirect to
   track hits on their downloads.  FileHound meets world.


Bugs fixed for .93
---------------------------------------------------------
 - Bad parsing on HTTP headers fixed.
 - HTTP resumes were completely broken.


Bugs fixed for .92
---------------------------------------------------------
 - File list can be sorted in different ways by clicking
   on the column headers.
 - shows date file was downloaded when done
 - shows date clipboard item was added
 - made double-click functionality vary depending on
   which tab is active.  the function that will occur is
   shown at the upper right, and seems to be the best for
   each tab page.
 - deletes zero length files.
 - made messages about odd length files more clear.
 - put in code to check for buggy servers that claim to
   handle content ranges but don't -- resumes of these
   files will be restarted from the beginning.  (see
   buggy server list below)
 - made "resume or overwrite" pop-up smarter, won't offer
   to resume files that weren't broken, or that are
   located on buggy servers.
 - fixed window size restoration code.  width was height.
 - arrgh.  fixed bug introduced in .91 wrt "resume all"
   and "overwrite all" thinking "all" meant two files.
 - fixed HTTP and FTP request lines to all have a
   terminating CR / LF.  (STARWAVE server is picky.)
 - made large lists of files draw more quickly, although
   there's still more work to do in this area later.
 - made clipboard code much faster when you copy a
   page with hundreds of links.  it's still pretty
   slow, but at least it's seconds not minutes.


BUGGY SERVER(S):

   Key string: "Server: Microsoft-PWS-95/2.0"

   Microsoft Personal Web Server 'Microsoft-PWS-95/2.0'
   does not resume HTTP transfers properly.  This manifests
   itself as an oversized file.  When you inspect the
   file, you'll find the resumed portion is actually data
   from the beginning of the file, not what was we asked
   for.  Also, the length of the newly downloaded data is
   more than the HTTP header reports; it is always a
   multiple of 4096 bytes.  The end result is a too-long
   file with a bad resume segment.


Bugs fixed for 0.91
--------------------------------------
- Choosing Overwrite will delete the local file first, rather
  than just leaving it intact and copying data over what exists.
- Exiting FileHound in mid-transfer either by design or suddenly
  will be able to resume correctly, assuming the server can too.
- Starting an archived URL from the properties dialog now moves
  the URL from the archive to the pending page (centralized
  control paths), like it's supposed to.
- FTP login failure is now non-fatal.  it will retry these
  forever, since it now assumes the server is full (the same
  failure response is returned for both bad logins and full
  server conditions.  what can I do?).  check your logins!
- ETA for resumed files is now estimated more correctly
- HTTP "Basic" authentication/passwording is now supported.
  Password and user fields are always enabled, and the auth
  header is sent according to the http 1.1 spec encoded as base64
- When a URL is archived, any transfer log associated with it
  is deleted to save space in the history file.


Initial public release 0.9 - "late beta."
---------------------------------------------
- First release upon the world.  It worked, mostly.




About the GOODNIGHT problem with IE4 - A Technical Discussion
------------------------------------------------------------------

 It seems that the problem is that under IE4's
 new RAS, the RASHANGUP function screws up the stack if
 you call it via a loadlibrary/getprocaddr.  So I wrapped
 the call to it in this mess:

                   static DWORD bproof;
                __asm mov bproof,ESP;
                (__RasHangUp)(buff[numras].hrasconn);
                __asm mov ESP,bproof;

 See, when RasHangUp would return, the stack pointer is
 offset four bytes higher than before.  Perhaps the new
 IE4 version is expceting two parameters instead of one,
 but that's certanly not documented anywhere I've seen.
 Anyway, I'm not happy with this solution either because
 I'd really like to know why it's screwing up the stack.
 Although it works, it's ugly.  If anyone knows....