Hi Johannes,
you have been very productive, thank you for the contributions.
I'm glad to know that \pgfplotslegendfromname improves things.
Unfortunately, while I find the results of your analysis reasonable, it
might be much more difficult to translate them into a bugfix: I fear that
pgfplots uses (legal) constructs of pgf/tikz which are not fully supported
by the tex4ht driver (which is part of pgf). Furthermore, any fixes in
pgfplots which make the tex4ht driver happy might break compatibility with
standard pdf output. On the other hand, fixes in the tex4ht driver require a
very deep insight into pgf, SVG, and tex4ht - which I currently do not have
in that combination. And a detection of superfluos text nodes which actually
contain graphics might not be simple as well. That's why it might take some
time...
I would suggest that you stick to the work-arounds. You might try the
pgfplots unstable as last resort (I believe that I have been working on
tex4ht support some time ago...). Besides the pgfplots unstable, I am
unaware of further fixes.
I fear that tex4ht support cannot be considered to be reliable - to me, it
appears to be strongly dependent on the tikz figure. So, you choice is to
provide bugfixes in the resulting SVG, or - if you like hacking - you might
try fixes in the tex4ht driver of pgf (or in the code generated by
pgfplots). The TeX hacks could be placed into your preable. Let me know if
you are interested in assistance of general TeX programming.
Sorry for the bad news.
Best reg
Am 25.07.2011 08:38, schrieb Johannes Wilm:
>
>
> On Sun, Jul 24, 2011 at 11:21 PM, Christian Feuersaenger <
cfeuersaen...@googlemail.com> wrote:
>
> Hi Johannes,
>
> that you for reporting the issue.
>
> I will look into it. However, it may take some time until we can
improve or fix the tex4ht support... please be patient here.
>
>
> hey,
>
> I have gotten a it further. If I use\pgfplotslegendfromname instead of
\ref, it seems to work.
>
>
> Please let me know if the following problem is known already and whether
there are any fixed out there, so that I don't code forever, trying to fix
it.
>
> This next problem I have though is that the svg-code of the legend is
incorrect and won't render correctly. I have manually fixed that in the
attached version. I can write a small python script that should be able to
fix it automatically, but it would probably be better if you fix it in the
original sources. there are basically two issues with it:
>
> 1. g-nodes are placed inside of text-nodes. That is against SVG-spec. Just
removing the outer text-node works fine.
>
> 2. the elements of the legend are not correctly placed. in fact they all
seem to be placed in the exact same place. I have moved the node of the
symbol/line-style associated with a label into the same g-node as the
corresponding label-text node. Additionally I have adjusted the numbers used
for placement of each a bit. You can diff the the svg-files to see exactly
what has been changed.
>
>
> --
> Johannes Wilm
> http://www.johanneswilm.org
> tel: +1 (520) 399 8880
>
------------------------------------------------------------------------------
Storage Efficiency Calculator
This modeling tool is based on patent-pending intellectual property that
has been used successfully in hundreds of IBM storage optimization engage-
ments, worldwide. Store less, Store more with what you own, Move data to
the right place. Try It Now! http://www.accelacomm.com/jaw/sfnl/114/51427378/
_______________________________________________
Pgfplots-features mailing list
Pgfplots-features@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/pgfplots-features
