Pinned ends and line styles

[Paolo F Cantoni]

There are some issues with line styles and pinned ends.  They may or may not be bugs....  I'd welcome your thoughts.

If you pin one or more ends of a line, and then change the line style, the pinned ends are removed - silently.

For some line styles, if you pin one or more ends, the line style changes to custom line.

[[I tried to use the help system but Pin End shows NO matches,
Pin found: 1109 matches in 609 files - most of which were to do with words containing "ping".
End found: 3510 matches in 1417 files

Can we have our old Help file back please?  the OLD CHM file I keep around for just this eventuality found "Pin End" as the first search entry.]]

Anyway, the CHM file said nothing about line style and end-pinning.

Thoughts?
Paolo

[qwerty]

For me this would be a bug. Pinning and line style are two independent properties. But probably not on planet Sparx :-/

Help! Yes, it's no help so often. This needs a full text search which behaves like one. Those systems are expensive. And you need to tune them. Maybe Siri and Cortana can help? Just kidding.

q.

[KP]

Pinning and line style are two independent properties.

Disagree. Some line styles (e.g. tree style) cannot function with pinned ends.

But probably not on planet Sparx :-/

The sunsets here are lovely. 

[Paolo F Cantoni]

Pinning and line style are two independent properties.

Disagree. Some line styles (e.g. tree style) cannot function with pinned ends.

[SNIP]

Yes, I came to the same conclusion, but...  I suspect that we are a bit like Lord Beaverbrook (apparently the usual allusion to George Bernard Shaw is incorrect) "we're just haggling over the price".

I suspect that we (I mean all users and Sparxians) might disagree on which styles allow/constrain the pinning of ends.  Since there appears to be NO formal description, perhaps we should try to reach a consensus here and then get the documentation and code to be consistent.

Since (as qwerty says) - at least notionally - Line style and end-pinning are independent properties
Should the line style "change under you" if you try to pin.   Which property takes precedence?

I propose:

Line Style       Allow Origin Pinning     Allow Destination Pinning

As the table to be created.  Note, how I have used the "allowance" concept.  As KP says, Tree style doesn't allow users to pin the destination, but why shouldn't it allow the user to pin the origin?

Paolo

[qwerty]

I can imagine the two suns raising over the horizon. But maybe it's just the a result of the smoke or those beers the night before ;-)

I launched the Help from the context menu of a connector, but that only brought me to some commonplace about connectors, not a description of the context menu. Searching for "line style" from there sent me to this relevant page. Searching for "pinned end" sent me to this page. Nothing about change of line styles (e.g. when pinning tree styled connectors).

So since it's not documented, it's wrong (which ever way it's implemented).

q.

[skiwi]

Pinning and line style are two independent properties.

Disagree. Some line styles (e.g. tree style) cannot function with pinned ends.

Even so, why does changing the line style initialise the pinned ends setting?
Even so, why does changing the line style cause the pinned ends setting to be initialised?

[Eve]

Even so, why does changing the line style initialise the pinned ends setting?

Because once the decision is made that a line style is incompatible with a pinned end, there are two options.

• Disable those line styles if the line is pinned - The result is users don't know how to enable that option. Instantly users either give up, or report a bug. Very few even look at documentation in this situation. (I've seen this happen so many times)
• Removing the pinning when line style is selected - Users can now do what they are trying to do, the problems with that selection have been automatically fixed. For most users, this is great, but eventually someone who doesn't realise why they are incompatible complains.

[Paolo F Cantoni]

Even so, why does changing the line style initialise the pinned ends setting?

Because once the decision is made that a line style is incompatible with a pinned end, there are two options.

• Disable those line styles if the line is pinned - The result is users don't know how to enable that option. Instantly users either give up, or report a bug. Very few even look at documentation in this situation. (I've seen this happen so many times)
• Removing the pinning when line style is selected - Users can now do what they are trying to do, the problems with that selection have been automatically fixed. For most users, this is great, but eventually someone who doesn't realise why they are incompatible complains.

Couple of points...
Just because most people DON'T read the documentation doesn't mean that there shouldn't be a mention.  As I read the documentation, I couldn't see any reference to pinned ends being optional and dependent on the line style.

There should be an optional indication that EA is intervening.  That's part of the UI(X) pattern language.

Paolo

[Graham_Moir]

I agree with Paolo that it would be handy, indeed user-friendly, if EA told the user that the pinned ends they carefully put in place are going to be, or at minimum have been, removed.

[qwerty]

+1

[skiwi]

I agree with Paolo that it would be handy, indeed user-friendly, if EA told the user that the pinned ends they carefully put in place are going to be, or at minimum have been, removed.

What I was thinking when I asked my question is why the "setting" is removed. I.e. if it is retained then changing back to the previous line style would resume the pinned end settings.