Re: [lmi] 'Οχι day is almost upon us

[Greg Chicares]

On 2015-10-04 23:37, Greg Chicares wrote:
> On 2015-10-04 17:03, Vadim Zeitlin wrote:
>> On Sun, 04 Oct 2015 14:40:12 +0000 Greg Chicares <address@hidden> wrote:
[...]
>> GC> Assuming we'll just say "oxi"...Vadim, is the patch below correct?

The decision is ναι. We'll proceed very much as though it were otherwise.

>>  It is correct in the sense that I don't think it does anything wrong, but
>> I think more code could be removed: is there any point in keeping the
>> validate_print_case_pdf_output() function at all? It doesn't seem to be
>> actually testing anything any more,
> 
> so thanks, you're correct: validate_print_case_pdf_output() no
> longer serves any purpose and should be removed altogether.
> 
>>  Should I make (and test) the patch doing either the former or the latter
> 
> The latter ("remove this function entirely") is correct, and I should
> be able to get that right.

20151007T1947Z, revision 6328: "removed altogether".

http://svn.savannah.nongnu.org/viewvc/lmi/trunk/wx_test_validate_output.cpp?root=lmi&r1=6160&r2=6328

> Please let's do everything in this order:
>  - Kim: cast your vote on the original question; presuming "No"...

Kim made a good case for keeping the "values" spreadsheet. The way I'd
express the case is that this TSV file contains every column that can
be used in an illustration, so it is on rare occasions very useful for
prototyping new illustrations that we would like to develop, e.g., when
a prototype must be shared for compliance approval before XSL changes
are begun.

Therefore, of course it must be forbidden in the same circumstances
as illustrations are forbidden:

http://svn.savannah.nongnu.org/viewvc/lmi/trunk/ledger_xml_io.cpp?root=lmi&r1=6327&r2=6326&pathrev=6327

>  - Greg: nuke this whole 'idiosyncrasy_spreadsheet' thing

But I had already promised myself the delight of removing an
"idiosyncrasy":

http://svn.savannah.nongnu.org/viewvc/lmi/trunk/ledger_xml_io.cpp?root=lmi&r1=6328&r2=6327&pathrev=6328

and a "pyx" option is harder to use, which is good. Typical end users
should not use this option at all. A required command-line option will
not block a determined power user, but it does require a deliberate
commitment to override built-in protection--a "mens rea" that cannot
plausibly be denied--which is not necessarily present when an unusual
string is typed in "Comments" ("I must have copied some old input
file that must have had "idiosyncrasy_spreadsheet" in it already").

And "pyx" options aren't pretty, but they're regular in some way
that "idiosyncrasy" options are not, which is why we've striven
to eliminate the latter over time. Here goes another one:

http://svn.savannah.nongnu.org/viewvc?view=rev&root=lmi&revision=6329
Replace an idiosyncratic interface that is no longer useful
^^^^^^^ I should have said "eradicate"

>  - Vadim: review and test my changes, revised as discussed above

You got the easy part this time.