lmi
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [lmi] Group-quote PDF: whitespace changes, and enhancement


From: Vadim Zeitlin
Subject: Re: [lmi] Group-quote PDF: whitespace changes, and enhancement
Date: Fri, 9 Mar 2018 01:17:12 +0100

On Thu, 8 Mar 2018 15:38:46 +0000 Greg Chicares <address@hidden> wrote:

GC> We think Helvetica is an improvement here.

 OK, thanks for the answer.

GC> Can the new PDF similarly apportion the available blank space (on
GC> the right) among all columns?

 Wouldn't this look really ugly for narrow(er) tables that also occur in
illustrations (e.g. supplemental reports with just 3 columns)?

GC> That would incidentally keep the numbers from running into each other
GC> with Helvetica.

 Actually, looking closer at this, the reason for the numbers running into
each other is that their width was too small for the values that could be
shown in them. E.g. the "Premium Outlay" column was made wide enough to
contain "999,999" but in this example it should be wide enough to fit
"999,999,999" instead. Just fixing this in ledger_pdf_generator_wx.cpp
already gives much better results.

 This change, together with the other ones discussed below, is included in
the new PR I just made:

        https://github.com/vadz/lmi/pull/74

Please let me know what do you think of the appearance you get when using
this version!


GC> We think the new PDF should use the same margins as the old ones.

 Changed too now (included in the PR above).

GC> We prefer the old PDF margins even for all-text pages, where the
GC> new margins give a left-of-center feel. I'm not sure it's easy to
GC> quantify that, but different people who've compared old and new
GC> PDFs here uniformly have this impression.

 For the tables, it's more than just an impression, the right hand side
margin was indeed bigger. This happened because the code shrinking the
columns if they didn't all fit, shrunk them by too much as it preferred to
round down rather than up to ensure they do fit between the margins, and so
the total table width became smaller than the available width after
shrinking. I've modified this code to ensure that we end up with exactly
the available width and this fixes the problem. Thanks for noticing it!


 But for the normal text, I'm not sure what could the difference be. As you
say:

GC> Perhaps the left and right margins are equal, and the ragged text on
GC> the right side adds some irregular white space to the right margin in
GC> some way that makes the empty space on the right look larger than on
GC> the left. I'm not sure that's the reason, but I'm sure that's the
GC> impression we get.

which is probably true, but it wasn't any different for the old PDFs (and I
say this after having painstakingly measured with magnifying glass the text
margins in pixels on a couple of pages in both the old and the new
versions), so I don't understand why nobody had noticed this before. Right
now, I do see it in both new and the old illustrations, but I doubt my
ability to formulate subjective impressions correctly after having done all
these objective measurements.

 The only hypothesis I can venture is that the effect is more noticeable
with wider margins, and while this seems counterintuitive to me, as the
difference is proportionally smaller, it might be true. So I hope that the
margin reduction mentioned above might help with this problem too.

 Please let me know what do you think of the PR 74 or, rather, changes to
the PDF brought by it.

 Thanks in advance,
VZ


reply via email to

[Prev in Thread] Current Thread [Next in Thread]