On 08-Mar-2013 12:28, John Cliff wrote:
(Speaking of bugs, your "reply" in Roundcube mangled the heck out the message you sent, chopping it off about 1/4 of the way down. I had not seen that before).
Right, cos non of us are the ones who admin user forums, have been responding to user emails, and talking to folk in chatrooms for
multiple years... oh wait we are.
Perfect, so you have lots of feedback from end users. In all that time how many nondeveloper end users have requested the proposed change to Save and "Save as...". Any? I maintain that Export and the redefinition of Save and "Save as..." are solutions in search of an actual problem.
I spent a while looking through the google results of: inkscape export "save as" searching for evidence of demand for this feature by the end users (as opposed to the developers). There was this:
https://bugs.launchpad.net/inkscape/+bug/171054
which (to my mind) revolved more around the implicit "and set 'Save' to also write to this file" in the current "Save as" issue than it did the core Export/Save as issue (conservative file formats only in save/save as). Ie, that's the "Save a copy as..." issue, not the "Export" issue.
I also found one example in the first 10 pages of that search of an end user requesting something like this:
https://bugs.launchpad.net/inkscape/+bug/170894
However that bug report to my mind completely skipped over the observation that Inkscape has neither a PS or EPS open function. That is, the issue was that EPS was a write only format.
And there were lots of developer discussions. No din of end users asking for export though.
once in the ~30 years I have been here has an end user complained to me that the "Save as..." in a program was unsafe because it let them choose a nonconservative file format!
So people who one would hope are highly educated and computer literate can use computers? no shit sherlock. Doesnt take a caltech >education to work that one out. The very fact your in an environment that is probably very much non average for inkscape users makes >your observations less useful. so yeah, I'll look where you are, and then take your observations with a pinch of salt.
The point was that these folks are about as technically sophisticated a group of software end users as you are likely to encounter. Yet even they have not been asking for the proposed changes to "Save", "Save As..." and the addition of "Export".
With for TSM until they hit save and get a warning they have no idea if its conservative or not. the other way its defined by the top level choice.
That tells us that the current warning message is poor (uninformative) and/or the function that generates it is primitive. I have not looked at the code, but I suspect the latter. That is, it emits the same message no matter what is in the original drawing. Inkscape can do a perfectly conservative save to other formats, EMF for instance, if the contents of the drawing are compatible with that format. Even the "round trip" requirement may be satisfied. Yet Inkscape still warns in these cases. With the possible exception of SVG filters it does not look difficult to actually make up a list of what has and has not been lost, such that the actual problems, rather than a hypothetical loss, could be accurately shown to the end user. For instance, at present in EMF export Inkscape's hatch patterns are all converted to a single EMF hatch pattern. This is because the SVG pattern method is not very compatible with EMF (the converse is not true, Inkscape imports all EMF patterns). So under "data loss", or whatever this would be called, the string
Inkscape hatch patterns were irreversibly converted to a standard EMF hatch pattern.
Could be added when this occurs. Others that would show up are:
Transparent and partially transparent objects are now opaque.
Gradients were emulated with multiple colored objects.
Regards,
David Mathog mathog@...1176... Manager, Sequence Analysis Facility, Biology Division, Caltech