You can save some loop iterations if you use the collection feature of InDesign. Basically this way you can assign the same value to a whole set of cells with one command: myTable.columns.item(0).cells.everyItem().appliedCellStyle = redCellStyle; Use that approach for the majority, and only run a loop* across the few exceptions. *Unfortunately evaluateXPathExpression() yields an array, not a collection. var rootXE = doc.xmlElements.item(0); var houseXEs = rootXE.evaluateXPathExpression('//Cell[@type="house"]'); for( var i=0; i< houseXEs.length; i++ )       houseXEs.xmlContent.appliedCellStyle = null; Dirk ... View more

There are multiple kinds of dialogs. You probably want a modal dialog. If you just want typing by default to hit your window rather than others, see IKeyboard acquireFocus (or similar name). That should be implicitly called when the dialog opens, unless your dialog has no text edit fields, e.g. just custom drawn widgets. If you want something more esoteric such as set the tool to be dimmed in the toolbar so that the user can not click it but still continue with e.g. the path tool, that's more work. If you just want to choose a different tool from your dialog, that's again something different. Dirk ... View more

I don't know - better ask in the respective forums. Dependent on your use case, I'd also have a deeper look at BridgeTalk. Dirk ... View more

How do you get your cell? This works for me: $.writeln(app.activeDocument.textFrames.item(0).tables.item(0).cells.item(0).associatedXMLElement.markupTag.name); Dirk ... View more

I just had a similar problem. I have no idea whether my solution works in CS3. ( function() {      var doc = app.activeDocument;      // the image page item      var r = doc.rectangles.item(0);      // the destination textframe to hold the inline      var tf = doc.textFrames.item(0);      // it should already be tagged.      // create a nested, blank tag      var xe = tf.associatedXMLElement.xmlElements.add("bla");      // copy the rect as inline      xe.placeIntoInlineCopy(r,true); } )(); Dirk ... View more

This is one of the questions I had submitted for the developer summit, and the only one answered: in short - no. All those notifications are only meant that you can schedule or perform followup activities. Setting the global error code is not expected in most of the cases, and as soon the next command is executed without that check and runs into a stuck error code, the command processor will perform its protective shutdown. It would be an enormous effort to clean up everything, and that is very unlikely. There are plenty gestures / activities that can all lead to a new page item. You can try to intercept them with many approaches, and keep the UI consistent this way (disabled cursor rather than drag-create, and so forth). For the remainder, you'd implement an ICommandInterceptor (a potential heavyweight performance drain, so be careful!) and knock out the command there. Even then, some commands are processed as part of a larger sequence (place a text frame is maybe 10 commands) and suppressing such a command may also have implications. Some other bottlenecks exist to support assignments, or you could eventually lock layers. It all depends on your specific intentions whether that is sufficient. Be prepared for much try and error. Dirk ... View more

Thanks for the reply. I hope that someone can bump the priority. Dirk ... View more

With the events showing up as duplicates (probably per supporting class) one idea is to step back a version and hope the matching TLB does not have the new "feature". Is the quoted TLH file available from some temporary build folder? You could then extract all const BSTR and sort them to eliminate duplicates. Another approach: some C# users have managed to decode the TLB and butcher them to their needs. Probably you mean that with "good look at the TLB" ... Finally, along your current hack: you could continue to use your prototype, advance it towards regular VB and link to that from your VC++. Dirk ... View more

I just reminded myself this is the server forum. Some additions: To create a new session e.g. for InDesign desktop / ESTK to try things out, use a startup script. You should also thoroughly test your scripts against leaks - some constructions are not caught by the garbage collector. As preparation install your script and do a test run. Then invoke the garbage collection yourself. Do it twice, it really makes a difference. my_testscript(); $.gc(); $.gc(); $.writeln($.summary()); Repeat the exercise a couple of times, and see whether there are changes in the summary. Dependent on the amount, you better track them down or plan to restart the server every now and then. Dirk ... View more

See JavaScript Tools Guide CS4.pdf, page 228 #targetengine enginename Defines the target JavaScript engine for this JSX file, within the designated target application. Supported by Adobe Illustrator CS4 and Adobe InDesign CS4; other applications ignore the directive. ➤ For Adobe Illustrator CS4 and Adobe InDesign CS4, if the named engine does not exist, and if the script originates within the application (rather than being executed in the ExtendScript Toolkit or received in an interapplication message), the application creates a new JavaScript engine with this name, which persists for the lifetime of the application session. ➤ If the script originates outside the application, and the named engine does not exist, the directive is ignored. Dirk ... View more

If your intention is to create a prototype anyway, continue to use VBScript rather than the more modern incarnations of VB. InDesign scripting was intentionally designed ignoring strong (object) typing. You might also consider ExtendScript, just because the community is much larger - next door in the scripting forum. That scripting forum would also be a better place to sort out your CS5 TLB problems - a minority over there prefers C#, the bigger VB and other languages - years ago I've even seen reports on Delphi. There are also ways to produce "managed" / .Net code using C++, again working against the TLB. Apparently you try that with the mentioned errors as result. With that approach you'll be a very small minority, though. Some C# folks reported they had success by manually disassembling the TLB, clean up and rebuild it. Maybe an approach for you? This forum is about developing plugins with C++ against the SDK - a separate, much richer set of C++ headers and binary libraries than exposed to scripting (what you see in the TLB). Again, you will not need the TLB. Getting started here will be a completely different experience, and you will most likely not be able to reuse most parts of your prototype code's architecture - anything beyond feeding the document object model. Dirk ... View more

Sounds more like a problem of your list<>::remove implementation that is only showing with gcc. You're wrongly assuming that after erase your iterator "next" is still valid. What happens if you do the same with fKey as int32? Besides there is good old UIDList ... Dirk ... View more

Steven, targetserver got roughly the same answer as the selection suite filter idea - the responsible developer has already left. At least there should be a working example in the SDK - I have not yet revisited it though. Dirk ... View more

Harbs, It was phrased sufficiently vague - "put that ... " could mean "put that number" or "put that fact"  Dirk ... View more

John, if these deleted color UIDs are held by your own entities (page items, attributes, your added interfaces etc.), then you have to provide an kRIDXNotifierService, see IRIDXNotifierHandler for details. The colors are lost when a color is deleted, in that case it is supposed to be substituted by a different color. If the color is stored by someone else, please report a bug. IID_IK2SERVICEPROVIDER,kRIDXNotifierServiceProviderImpl, IID_IRIDXNOTIFIERHANDLER,... Checking for IsValidUID is probably ok if your UID is stored with an optional persistent interface (IgnoreTags ...), but in that case you should clean up within your plugin's IPlugIn::FixUpData() method. DeleteUID() would be the opposite of what you want, the color already has been deleted and you should anyway call that method only from within a command. Along the last paragraph of my last question I had hoped to learn some more details on that RIDX topic during the summit, but unfortunately the event had other priorities ... Dirk ... View more

I do not know about any related documents, but that does not mean there are none. The debug build has utilities to inspect existing UI - mainly palettes. I use that information (dimensions, spacing etc.) and similar utilities to derive my own widgets. Same goes for icons, just try to follow theirs. Colors, fonts - the names in InterfaceColorDefines.h and so forth are self-explanatory. On the other hand, Adobe recently has turned to strongly push the new Flex / Flash based CS SDK, as an alternative to resource based UI. In my opinion even with the "official" Flex skin these palettes look still odd / out of place in most regards: size, colors, shapes and overall behaviour. A big step backward, especially comparing to the previous consistency efforts of the suite. Anyway there are other things that will give a user more value than widgets a few pixels off, such as optional behaviours / gestures: drag and drop, quick apply support, fewer but more powerful dialogs and palettes, intelligent features straight within layout (smart guides, donut) a clean scripting interface, responsiveness thru correct notification handling, fully understood selection and so forth. Dirk ... View more

You can see InDesign as an cross platform application framework that also happens to know about typography. Thru that abstraction it is irrelevant most of the time whether you are running on top of Cocoa, Carbon, Classic or Windows. Dirk ... View more

If these tools connect to InDesign, you'll have to rewrite using a cross platform language. One tool to handle that requirement would be ExtendScript Toolkit, which comes free with your InDesign installation. If your self-developed tools are independent from InDesign, have a look at Mono (probably only covers C#) or RealBasic. I doubt (as in "don't know, but probably not") that they support InDesign bindings in any compatible way. Dirk ... View more

Shaolin, at least here your answer is hidden 😉 . Please repost. Dirk ... View more

Actually I got a bit carried away because the time is running for final preparations and nobody else has picked up this thread. Remember that it won't be my private session - please add your own questions at whatever level you feel comfy with. I think the session will be recorded and made available on the web, so everybody should benefit from more general questions. My esoteric problems may also be covered in the coffee break. Dirk ... View more

Next one in UI: Drag and Drop from outside. Specialties in this situation are you can't control the flood of flavors produced by the external application(s) - e.g. the gang of internet browsers. the greatest common denominator may be the text flavor text flavor may be irrelevant but haunt you from now on The introduction would list the cascade of overloading opportunities which internal helpers will compete with you and how to knock out the competition The most relevant decision - to go with kSysFileDataExchHandlerHelperService rather than kPageItemDataExchHandlerHelperService as suggested by the custom DD example, has to be enforced when the data object is not even internalized. Knowledge of flavors is just not enough here. Did I miss a service here that would allow me to preprocess the data object, and strip + add flavors? The decision for the actual kLayoutDDTargetFlavorHelperService is not directly accessible, instead there is an indirection thru the second level (the multiple HandlerHelperServices). Is there a way to enforce the priority of the FlavorHelperService short from covering every competing HandlerHelperService? My approach with a kSysFileDataExchHandlerHelperService is also consequence that the example leaves the visual drag effect as an exercise. As I understand it the drag effects are not separately factorised, but implemented at the level of the respective kLayoutDDTargetFlavorHelperService, so your flavor helper can't just build a package matching the decoded part of the data object, and throw it at a service to make it behave like the place gun. Am I wrong with that? Btw, another reason against kPageItemDataExchHandlerHelperService and thus kLayoutDDTargetPageItemFlavorHelperBoss - it causes a visual smear (ghost of the spline on the first drop candidate location), as you can see from drag&drop of a library item. Dirk ... View more

In the UI department, I'd love to here more about IID_ITEXTTARGETSERVER. Unfortunately the plugin where I will need it has not yet evolved far enough, so I can not yet offer concrete questions. Let's begin with an overview why/when it is important (used all over the place). A list of do's and dont's, common pitfalls and so forth would be interesting. ... View more

The selection architecture is also a welcome area for explanations. Continuing from the previous problem, we could get an idea how to add input parameter validations to selected suites. For example, it should be possible to insert an own level of indirection in between ASB and CSB. Most Suites would just pass on in a similar manner as ASB implementations, so the boss would be derived from kIntegratorSuiteBoss. For the actual validation some Suites would have their Can() or Do() methods overloaded. To get this working I just have not yet found a way to add my boss to the chain - e.g. a kind of suite filter factory service. Less sophisticated - what is the correct way to implement temporary selection state that is bound to the presentation (window). In other words, a concept such as "active layer", which is per layout window. The discussion would explain where in the selection architecture the storing interface should be held (on the window boss, layout widget or within the associated CSB?), what relevant notifications to use (is ILayoutAction really a good idea, or better use SelectionExt startup?). Which observer should maintain the value (e.g. should it be bound to a panel, if the value is also scriptable?). How to access the value from the CSB side of custom suites - especially if the LayoutWidget is chosen to hold it. Is there a better way than to ask users to grab the "selection owner" off the ISelectionManager at the ASB side, and pass it through in the method? In my opinion this additional parameter exposes an implementation detail in the suite methods that does not belong there. Let's continue with such unknown gems like the idea that suite bosses need not be bound to selections. It should be possible to create your little own kLayoutSuiteBoss, let it do some high level work and go away. This should even work in InDesign Server. The use case for kTextScriptingSuiteBoss is a good example. The selection topic would be finished with a description how to implement a DIY concrete selection boss, so that it is associated with your custom presentation window (e.g. what you did to introduce galley view). I think the interface of the relevant service is missing from the SDK, are there alternatives or at least plans to publish it? Dirk ... View more

I take the first half hour 😉 for error handling. It is a frequent request from customers to implement dynamic constraints to arbitrary model changes. SuppressedUI is a distantly related sledge hammer approach, but incomplete and inappropriate as it just pulls the general plug. To simplifly matters, I'd pick two example model changes. The first is to constrain a subset (e.g. by label) of page items to be fully contained within the upper half of their spread. Let me repeat this is just an example, the constraints could be anything else. Functionality to only touch very shortly would therefor be - how to overload existing trackers (donut ...) that they still work, but obeye the constraints on such page items - a selective custom snap / grid behaviour The larger focus of the discussion would be at the command processor. A variety of commands may cause a violation of the constraint, and some of them are observable by IID_ITRANSFORM_DOCUMENT. An observer could wade through all those notifications, watch out for our special page item, and wish to veto some of the changes. The direct approach (in theory) would raise an error condition straight from the notification, and expect the command processor to detect the transaction failed and roll back everything. The UI gesture that caused the command finally picks up the error code and displays the matching error message. The discussion gets into detail why this is a bad idea. It shows how simple functionality is still distributed across nested commands and sequences of subcommands, with unhandled error codes all over the place and boom you find yourself in a protective shutdown. Examples for such nested command situations, beyond IID_ITRANSFORM_DOCUMENT - an inlined page item wich is pushed across the border by typing - place by drag and drop copy - modifying an object style causes a wider stroke - changing the shape of the page item Then - and that is the actual question, so please major focus here - it would show the alternative approaches how to veto such commands. Given enough time, the second constraint would apply the principle to rendering objects, aka swatch. It would constrain use of a set of swatches to specifc page items (for simplification the swatch must not show up in text at all, e.g. as rule above stroke gap color ... ). The focus here would be a discussion of reference indexes - how custom (thus unknown) swatch consumers participate in that mechanism and how to veto their operations if they violate our constraint. Dirk ... View more

It is a bad idea to modify kComposeStyleBoss unless you implement a text attribute. Dirk ... View more

Replying to myself: Export to IDML, have a look there. If the numbers show up (or was that only for text variables?) you might have a chance to ask the responsible script provider. Some things are only visible thru IDML, missing from general scripting e.g. JavaScript! Dirk ... View more