On 11/26/2013 1:27 AM, bill wrote:
>> So you replaced Eudora's original mailboxes...
> A horrible thing to do, but since it doesn't seem to phase Eudora, inconsequential.
No, it could be very consequential, but probably most seriously
because some modifications (e.g. -ft) to original Eudora files
(including "out of sync" TOC files, to be explained below)
might have spoiled the subsequent TB/OSE import procedure,
conducted by an entirely different program than Eudora,
whether or not they seem to have upset Eudora itself.
>> Eudora mailboxes also each consist of _two_ files each (one .mbx + one .toc), and Eudora
>> Rescue appears to use _both_ of those as input, producing only _one_ output file for use by
>> Unix or Thunderbird, etc., which then leaves Eudora's original ".toc" file an "orphan"
> Which Eudora seems to continue to use to make sense of the rescued files.
No, it can't possibly do that, because the TOC file contains absolute offsets
to the beginning of each message within the "messages" file (MBX),
and since ER, by various adjustments which it makes, particularly
to headers, and even to the "separator" between messages in your case,
changes the lengths of messages, hence makes the original offsets invalid.
Now Eudora is not oblivious to the possibility that MBX files might get
modified all by themselves somehow, particularly by users who emulate
"sorcerer's apprentices" and don't know what damage they may cause thereby,
so it compares the "last modified" file timestamps of the
MBX and TOC file pairs, and if the MBX timestamp is significantly later
(where the default for "significantly" is a leeway of 10 seconds),
then Eudora detects the need to rebuild the TOC, which it may do
with or without asking the user. Rebuilding the TOC, in turn,
usually causes some loss of TOC-only information, and sometimes
even resurrects deleted or edited messages, but at least
it ends up correcting the vital exact offsets to each message,
which may lull the user into thinking that mucking around
with MBX files outside of normal Eudora operations
was a brilliant (as well as inconsequential :) thing to do.
Before I go any further, I should back up and amend some things
that I earlier said, because I think I've found more of what you
may have relied on in the following section of ER's on-line manual:
<
http://qwerky.50webs.com/eudorarescue/readme.htm#thunderbird>
In that section, ER's author points out that ER does not handle
attachments, which means that it does not re-package attachments
so that they become part of legitimate MIME-compliant original
messages. This means that to really end up with what TB needs,
unless you don't even want the attachments re-connected with messages
in TB, it will be necessary to eventually use the TB (or OSE) importer.
Normally, one uses _only_ the TB/OSE importer, whose best version
seems to be the one which came with OSE 1.0 -- however,
even that version, when importing only mail,
is somewhat imperfect -- it may crash only part-way through,
it may fail to recurse into all of Eudora's ".fol" subfolders, etc.,
and in such cases, the usual, tedious way of continuing an
interrupted import is to hide what has already been successfully imported,
then resume by launching TB/OSE again and invoking "Import mail" again,
until eventually it finishes importing all mailboxes.
It might even be necessary to move some mailboxes from a Eudora subfolder
up to the top level to continue, which is even more tedious because you
have to then also move imported mailboxes back to subfolders in TB, but eventually,
most people whose adventures I have followed have come to a successful end.
The idea of using ER on Eudora's original mail store is basically
only to fix some things which TB/OSE's importer might fail to do;
I don't know whether the OSE importer already does those small things
anyway, but apparently the TB importer, as of TB version 2,
had not yet thought of everything, so ER was able to do something useful
to prepare original Eudora info for import by TB.
One of the specific things which ER thought it could to was to correct
TB's failure to preserve message status (from Eudora's TOC files),
for which ER defaults to inserting TB-specific header lines into
each message written to its output mailbox files (which is,
as pointed out above, going to make Eudora's original TOC files
out of sync with converted MBX files).
Just to give you an idea of how Eudora's and TB's developers
were significantly changing the importer for TB version 3 and OSE,
here is one single "Bugzilla" case history,
in which Qualcomm's Geoffrey Wenger and Mozilla's David Bienvenu
go back & forth about how each one's code changes affect the other,
sometimes adversely:
Bug 368347 - Importer (Win Eudora, Mail)
does not import message STATUS (read, unread, etc..)
<
https://bugzilla.mozilla.org/show_bug.cgi?id=368347>
If OSE 1.0 had already fixed that, as is claimed,
then there was no need for ER to be used before the OSE importer,
at least not for that particular part of what ER was created for,
but the importer which comes with later versions of TB
may never have included the fix claimed for OSE,
which took TB 3.0.4 as a starting point for producing "OSE 1.0,"
from which Mozilla did not take anything as it kept on modifying
only its "main branch" of TB code for various related Mozilla products.
At any rate, the very detailed instructions by ER's author
for using ER as an intermediate step between Eudora and TB,
said to first "compact" all the Eudora mailboxes (did you?)
It then gives very detailed instructions for using original Eudora input
and a DIFFERENT TEMPORARY PLACE for ER's output, and at this time, the only
suggested ER option (beside location of input or output) is -x=.mbx,
which you said was "recommended" -- yes, it was, after all,
because subsequent use of the TB importer was to follow,
but I would have chosen option -fe (leave separators in Eudora format)
rather than -ft (change separators to TB format, even though output
was to continue to represent legitimate Eudora mailbox format),
and replacement of Eudora's original files with ER output was never suggested,
which might have caused problems for Eudora which you have yet to discover,
for all you know (so I'd tell onlookers "don't do that in your own home" :)
ER's author then gives a tedious procedure for TEMPORARILY swapping
the original Eudora files with the temporary output files from ER,
prior to using TB to import ER's output files,
then to swap those back again, so that Eudora's original files
would finally be left as they were, untouched.
The reason for all that swapping was that TB/OSE's importer
uses a specific Windows Registry key to find the last-used
command line which launched Eudora, from which it locates
the data files to import. ER's author figured to swap
all the data files, once before TB/OSE's importing
and again re-swapping afterward, instead of just
modifying the registry key that the importer uses
(HKEY_CURRENT_USER\Software\Qualcomm\Eudora\CommandLine) or else
(HKEY_CLASSES_ROOT\Software\Qualcomm\Eudora\CommandLine\Current)
which is actually a lot easier, because you don't have to restore
either key afterward, since Eudora is going to merely
set it again, each time it is launched again.
The value to set for that key should have a second path
(using "8.3" file names or else a fully quoted complete path)
to the directory from which TB/OSE
should take its input (files to be imported).
>> And then you wonder why TB/OSE's importer from _Eudora's original mailboxes_ may not have
>> recognized anything in the thus mangled Eudora mailboxes that you left in place of the
>> originals?
> It can recognise the individual mail files as mail files, but won't open them. As you pointed
> out (and I already knew) Eudora OSE is based on an early version of Thunderbird, round about the
> Thunderbird 1.0 that Eudora Rescue was designed to service.
If TB won't open a file as a mailbox, then that's what I would call
"NOT recognizing the file as a mail file" ;-)
BTW, TB also has a kind of "index" for every mailbox file,
having file name extension ".msf" -- this index is very different
from Eudora's, and TB has its own way for re-generating it.
OSE is based on TB 3.0.4 -- one of the reasons why it took
years longer than expected to even arrive as far as OSE went
is that Mozilla's original development of TBv3 was painfully slow,
and Qualcomm had to start with a ready-for-full-release
version of TB before even applying its own changes
to make any of its "beta" versions of what eventually became OSE.
ER seems to have had no idea about TB3's future development,
which puts it in a very peculiar position as to whether
there is any need (or potential harm) in using ER
to modify Eudora's original mailboxes (and drop all TOCs)
before using a much later importer that was already
being designed to handle some of what ER was created for.
> For the life of me, I can't remember whether I tried to get Eudora OSE to import the original
> Eudora files (which I've still got, stashed on a corner of the hard disk, as well as on a DVD.
> My impression is that I tried it, and it didn't work, which is why I dumped the output of Eudora
> Rescue on top of the original Eudora files, telling it over-write anything with the same name.
Attempting to use read-only Eudora files (as files on a DVD tend to appear)
is usually unsuccessful.
Your shortcut to not use an intermediate output area for ER output
left out-of-sync TOC files in the original Eudora area
(at least as of immediately after using ER),
as well as mailbox files already modified as if ready for TB's use,
so with all your personal modifications, who would know what to expect?
> Eudora Rescue did chug through that lot, taking some hours to complete the process. The output -
> as examined by Mozilla Firefox - had been changed, and was a bit more readable, in that the
> actual text component of the e-mail showed up as more or less coherent chunks of text in the
> middle of a load of formatting information.
Since ER does not even claim to do the transformation you observed,
I rather think that what you observed
may be an artifact of using a _web browser_ to display a mailbox file,
rather than using a _plain text viewer_
For more histories of people trying to use ER between Eudora and TB,
Google this: "Eudora rescue" site:
mozillazine.org
This whole affair reminds me of a once famous cartoon:
<
http://www.businessballs.com/images/treeswing/tree-swing-s-hogh.jpg>
<
http://i0.kym-cdn.com/photos/images/newsfeed/000/475/749/fd8.jpg>
<
http://www.britannica.com/blogs/2009/10/the-classic-tree-swing-example-of-production-and-customer-service-gone-awry/>
And here's the Sorcerer's Apprentice, as animated by Disney's studio:
<
http://www.youtube.com/watch?v=cWZJcKM8pO0>
<
http://en.wikipedia.org/wiki/Fantasia_(film)>
<
http://en.wikipedia.org/wiki/The_Sorcerer%27s_Apprentice_(Dukas)>
> I can use my laptop as a sandbox to run the process that you think I should have followed -
> straight from Eudora to Eudora OSE - since it's hard-drive is uncontaminated by Eudora Rescue,
> or any converted files.
>
> I'll post the results here when I've got around to doing it.
> Your style does seem to rely more on being rude about what's been done...
More about how "what's been done" had to be dragged out of you,
instead of your giving enough info to avoid wasting anyone else's time.
My view is that anyone wanting a consultant can hire and pay a consultant
to conduct said kind of interviewing, or can "pay" for the free consulting
by bringing the facts to light up front, which I think it very rude
to have not done in the first place, particularly when done by someone
who is already experienced in the computing field.
> than it does on any kind of demonstrable expertise about what's going on.
That reminds me of a remark attributed to Mark Twain:
"When I was a boy of fourteen, my father was so ignorant
I could hardly stand to have the old man around.
But when I got to be twenty-one,
I was astonished by how much he'd learned in seven years.�
Twain spoke in many places without every word being
written down, but here's a "term paper" sized study about it,
which unfortunately fails to observe how true is the sentiment
anyway, as well as how in character for that author:
<
http://quoteinvestigator.com/2010/10/10/twain-father/>
--