stressfree - dwf

Why Autodesk should 'Open' DWF

David — Wed, 07 Mar 2007 22:24:39 +0000

Beyond the Paper's Scott Sheppard recently pointed to McDwiff as the first partial example of a Mac-based DWF viewer. Unfortunately for the DWF starved Mac community McDwiff is simply a wrapper around a WebKit browser window pointed directly at Autodesk's own Project Freewheel web service. It fails to qualify as a true desktop application for a number of crucial reasons:

It does not (yet) add functionality beyond what is present in the Web-based Freewheel viewer.
DWF files must be first uploaded to the Autodesk web service.
There is no off-line mode or local caching to improve performance.
The lifespan of the software is entirely dependent on the existence of the host service.

Note: These limitations are not the developers fault as they have obviously only just initiated the project. It will be interesting to see where they go from here.

Looking at McDwiff and Project Freewheel got me thinking about Autodesk's direction when it comes to DWF. Unlike many AEC data format categories what constitutes the dominant 2D/3D design information exchange format has yet to be decided. However things are beginning to change with a battle brewing between Autodesk's DWF and Adobe's 3D enhanced PDF. Unlike pure data formats such as DXF, DWG and DGN goal of a design information exchange format is to provide AEC professionals with the ability to deliver digital building design information safe in the knowledge that any recipient will accurately experience the design in the manner intended by the author.

Why is such a format important?

With the rapid adoption of the Internet within business AEC professionals are conducting an ever growing portion of their design collaboration online. Central to this collaboration is the 2D documentation and 3D models that visually describe the outcomes of this design process. The unfortunate dilemma faced by AEC professionals is exactly what format to use when exchanging such information? There are a broad range of potential formats currently available but they all have their limitations as illustrated in the table below.

Format	Benefits	Limitations
Bitmap Image	Broad software support Consistent display across platforms	Limited accuracy No intellectual property protection Only supports 2D imagery
Vector-based Image	Accurate display of data Relatively broad software support Consistent display across platforms	No intellectual property protection Only supports 2D vector-based imagery
Office Productivity Document	Broad software support Relatively consistent display across platforms	Limited/no intellectual property protection Only supports 2D imagery and data
Traditional 2D PDF	Broad software support Consistent display across platforms Intellectual property protection	Only supports 2D imagery and data
Flash Animation	Supports 2D and 3D data Intellectual property protection Broad reader support Consistent display of design elements across platforms	Limited printing functionality No AEC specific creation tools
Proprietary 2D/3D CAD format	Supports 2D and 3D data	Limited/no compatibility across different software platforms Potential for inconsistent display of design data Limited/no intellectual property protection Conveys the entire digital model rather than a subset identified by the author
Standard 2D/3D CAD format (DXF, OpenDWG, OpenDGN)	Broad software support Supports 2D and 3D data	No intellectual property protection Conveys the entire digital model rather than a subset identified by the author
Propriety Building Information (BIM) Model	Supports rich AEC semantics Manages 2D, 3D and text-based data	No compatibility across different software platforms Limited/no intellectual property protection Conveys the entire digital model rather than a subset identified by the author
Standard Industry Foundation Classes (IFC) Model	Supports rich AEC semantics Manages 2D, 3D and text-based data Consistent display of design elements across platforms Industry standard	Limited software support No intellectual property protection Conveys the entire digital model rather than a subset identified by the author

From the perspective of an AEC professional the ideal format is one that can communicate 2D and 3D information that any recipient can consistently view without misappropriating the author's intellectual property.

The limitations of traditional 3D formats

Conventional 2D or 3D models cannot fill this collaboration need because they are focused on recording design data rather than exchanging design information. Information exchange is the communication of context specific data for dissemination in a certain manner, for example an A3 sheet of elevation drawings prepared to a scale of 1:100. In contrast data exchange is the communication of an entire data-set for interpretation in the manner of the recipient's choosing. There already exist a range of common AEC data formats such as DXF and the semantically rich Industry Foundation Classes. Whilst these will always be of tremendous value within the industry, when it comes to exchanging design information the malleable properties of these pure data formats limit guarantees which can be placed around communication consistency and protection of intellectual property.

The importance of consistency & intellectual property protection

Consistency is crucial in AEC information exchange because the intention of most design documentation is to act as blueprints for a physical creation. If the exchanged digital documentation is not consistently presented to all participants the design team will loose confidence in the medium and be forced to use conventional (i.e. paper-based) information exchange methods. Traditional data orientated 2D/3D formats provide no guarantee to participants that the various CAD applications available will display the parsed model consistently, if at all. To resolve this issue design information formats and their associated reader applications employ measures to ensure that information exchanged with team members is displayed in a consistent manner across all platforms.

Increasingly for many AEC professionals the digital model is the lifeblood of the design process and the source of most chargeable activities. Literally giving this model to untrusted members of the design team could compromise future income streams, devalue their role in the design process, or raise legal liability issues if the model was used inappropriately in such activities as simulation and engineering studies. Formats for design information exchange protect intellectual property by enabling the author to communicate a limited subset of the entire digital model in a manner that cannot be reutilised for anything other than its intended purpose.

The combination of these unique requirements has resulted in the evolution of two similar but competing formats from Autodesk and Adobe: DWF and PDF with 3D extensions. Whilst each have their own strengths and weaknesses they ultimately have their eyes set on the same goal: to become the dominant design information exchange format within the AEC industry. Central to success is the ability for both formats to become ubiquitous and operate above traditional concerns such as operating system, hardware device and CAD vendor.

Format ubiquity and the success of PDF in the world of 2D

Adobe's PDF format has achieved a level of platform ubiquity in the 2D documentation space that DWF and 3D PDF are striving for in the 2D/3D design information exchange arena. This ubiquity has been achieved through a couple of related factors. Firstly PDF's creator Adobe supports native PDF readers on a range of hardware and software platforms. Whilst it is unsurprising to find the Windows version receives the most attention, users of other platforms are not left feeling isolated when it comes to inclusion of new features and ongoing support. Secondly a rich developer ecosystem has increased the acceptance of PDF from merely a proprietary (yet open) format, to a vendor neutral and soon to be internationally standardised platform for business information. There are literally dozens of non-Adobe sponsored, open and closed source implementations of PDF libraries, readers and writers written for a variety of different computer languages and applications. This diverse ecology ensures that even if Adobe were to pull support for some, or all aspects of PDF the format itself would remain viable for years to come.

Time will tell whether Adobe's 3D extensions to PDF will receive the same platform and developer support as its older 2D equivalent. Considering the 3D extensions are being standardised during the forthcoming ISO certification process there stands a very good chance of the extended format gaining broader platform and developer support in the near future. Considering PDF's current momentum DWF faces a difficult challenge if it wishes to become the dominant design information exchange format. Fortunately for DWF's chances its creator Autodesk is not idle when it comes to this challenge.

Autodesk's DWF initiatives

Unfortunately for Autodesk DWF has yet to reach the same level of platform nirvana held by 2D PDF. However they do have five initiatives underway to encourage adoption of the format within the AEC industry:

Extensive DWF support is built into many existing and new Autodesk products.
A revised DWFx format has been released which is built on top of Microsoft XPS technologies. This benefit of this format is that is can be read by Windows Vista without the need to install a dedicated viewing tool.
Recently Autodesk Design Review was released for free alongside existing DWF Viewer and DWF Writer applications for Windows.
A DWF toolkit has been made available for software developers on Windows, Mac and Linux to write for and integrate with DWF applications.
Within Autodesk's labs they are working on Project Freewheel, a Web-based, hosted DWF viewer.

Gauging success through market significance

Unfortunately for Autodesk the ultimate success or failure of DWF will not hinge on how well it is implemented within their own product line. Instead the deciding factor of its success will ride on its level of uptake within the AEC industry when compared to the use of PDF and the myriad of other traditional format options. Market significance will depend on three equally important factors:

The ability of Autodesk to deliver compelling reasons why AEC professionals should use DWF instead of PDF or traditional design data files.
The capability to read DWF in all design environments on any potential platform.
The proliferation of DWF beyond an Autodesk-only software mindset. In essence the concept of DWF must become 'bigger than Autodesk'.

If Adobe can leverage PDF's existing user and developer base to become the mainstream 2D/3D design documentation format Autodesk risks losing one advantage it has held within the AEC industry; dominance in the file format arena. In the past Autodesk has used its control of the legacy DWG format to assist software sales and deter competition in the 2D/3D CAD space. Two such examples of this have been the encryption applied to DWG by AutoCAD 2004 and more recently the copyright lawsuit filed against the Open Design Alliance for their support of TrustDWG. Undoubtedly whichever vendor can achieve a similar level of format dominance in the design information exchange field will hold greater leverage when it comes to future sales, marketing and development activities.

Initiatives that appear to be working

Two of the five Autodesk DWF initiatives appear from the outside to be working very well. Integration of extended DWF capabilities within a range of Autodesk applications is signaling to the industry that not only are Autodesk serious about DWF but it is also a very capable format in its own right. The availability of free reader and writer applications ranging from the complex but very capable Autodesk Design Review down to simple DWF reader/writer tools are also encouraging users who do not use or have access to the latest Autodesk applications to experiment with DWF on a limited yet worthwhile basis.

DWFx: clouding the water or lighting the way?

Of the five initiatives I am least certain about the establishment of DWFx, the derivative of DWF based on Microsoft's XPS (a.k.a. 'PDF killer') technology. The problem DWFx poses is that it makes the process of explaining to AEC professionals what DWF is and why it is useful more complicated. Rather than having one format to choose from there is now two that do exactly the same thing but have varied levels of support in different software versions and operating systems. Whilst there are some fairly valid technical decisions behind the establishment of DWFx its timing seems geared more toward product release cycles (i.e. Windows Vista) than any industry call for a DWF successor at this point in time. The telling question which remains to be answered is how the consumers of this technology react (if at all) to these diverging DWF formats, especially considering the imminent entrance of 3D PDF into this field.

Freewheeling Freewheel and its effect on the developer community

After looking at Project Freewheel for a while I have come to the (maybe wrong) conclusion that it is an excellent concept that in the long run is actually doing more harm than good to the DWF developer ecology. What Autodesk should be focusing on is providing application and Web developers with a tool-set to create their own Project Freewheel rather than building functionality on top of it. McDwiff is an obvious example that the current DWF developer toolkit provided by Autodesk is not what high-level application developers want. What is needed is an easily deployed DWF engine available for any platform that those interested in working with DWF can quickly utilise in a few lines of code. Ironically Project Freewheel provides this functionality but its ties to a central website limits developer potential.

How can Autodesk provide these development tools when they do not have the resources or experience to support multiple platforms? The answer is they cannot, but what they can do is encourage other developers to help them out in achieving this task. Although the DWF developer tools are available free of charge they are not easily accessible under a common license. Instead developers are forced to complete a rather large form and agree to a license which essentially states you can do anything with the code as long as you acknowledge the changes made when distributing the software. Forms and unusual licenses are barriers to developers, especially those used to working in an open source world of anonymous CVS and free (as in freedom) license models.

With the kit already available without cost under a relatively open license agreement why not take the next step and establish a SourceForge project for the software and release it under a commonly understood (LGPL or Apache) free license? After all it is not uncommon for large proprietary software companies to do such a thing. The benefit of the move would be to send a clear sign that Autodesk is serious about fostering a developer community. This extended developer community would be free to create DWF-centric applications and fork or port the existing DWF code-base in a variety of ways. The short term benefit of this action is that the DWF toolkit would be turned from an isolated download into the basis for an entire online development community similar to that of the Open Design Alliance's OpenDWG and OpenDGN initiatives. In the long term the opening of the DWF toolkit could be used to signal the tip of a very large, open source DWF iceberg.

Autodesk should not simply stop at freeing up distribution of their DWF toolkit. Instead this action should be used as the prelude for the open sourcing of the DWF Viewer and Project Freewheel code bases. With Autodesk Design Review now available free of charge Autodesk's DWF Viewer is of no real value from a commercial product perspective. If opened to third parties its code-base could be of tremendous value as a starting point for DWF readers on other platforms such as OSX and Linux. Whilst it maybe difficult to believe there are communities of developers that take pleasure in getting Windows centric applications like DWF Viewer running on multiple platforms. Likewise the goal for Project Freewheel should be to act as an open technical demonstration of what can be achieved with DWF on the Web. Whilst it is easy for Freewheel to be viewed as a potential money maker for Autodesk in reality the concept's true value will only be realised when third-party developers can embed similar functionality within their commercial or in-house developed Web applications operating behind corporate firewalls. Consequently instead of productising Freewheel, Autodesk's goal should be to inspire Web developers to write similar DWF manipulation functionality in PHP, Java, Ruby and other Web languages which may evolve. It is only through actions such as this that Autodesk will be able to proliferate DWF technologies rather than simply monopolising their use in the interests of short term profit.

Conclusion

The looming DWF vs PDF format debate will be beneficial to the AEC industry as the heated competition will no doubt enable new levels of digital collaboration. Currently it is difficult to say which format will become dominant but undoubtedly whichever vendor fosters the strongest developer ecosystem will hold a significant advantage. With this factor in mind Autodesk should be actively nurturing the growth of its external DWF developer community by removing of all barriers to entry. This could be achieved through the release under a common open source license of a select few DWF-centric technologies within a community friendly environment such as SourceForge. Through this process Autodesk would stand to grow an even stronger DWF developer community and in the process improve DWF's chances of becoming the ubiquitous design information exchange format.

Another DWG/DWF search engine

David — Tue, 13 Feb 2007 11:20:49 +0000

An the subject of search engines for DWF and DWG files there is cadbot.com. Unlike DocuPoint Discovery, cadbot.com is not intended for internal sites and instead just allows you to search for CAD files available on the public Internet. Whilst it is not as useful as searching your own resources it is still nice to keep in mind if looking for a CAD model of some description.

Reviewing Autodesk Design Review

David — Sun, 04 Feb 2007 08:16:08 +0000

Recently Autodesk dropped the price tag off their Autodesk Design Review package making it far more accessible to the Architecture, Engineering and Construction (AEC) industry. Design Review is a DWF-centric tool for viewing and reviewing 2D and 3D design documentation. Prior to becoming freely available Design Review appeared to the casual observer as a useful tool for those heavily into Autodesk products but not the mainstream audience. This pricing shift changes the game and enables the software to compete against Adobe Acrobat as the primary, general purpose viewing tool for design documentation. With this in mind I took a fresh look at the software and what follows is what I found.

Obtaining and Installation

Design Review can be downloaded from the Autodesk website but before you can begin you must provide Autodesk with a few personal and professional details. For what is now free software this seems counter productive as it acts as a barrier to distribution. If you are a software vendor and you want your software tool to become ubiquitous so that your format becomes dominant in the industry (i.e. like Flash and PDF) your software needs to be easily accessible and preferably available from a range of download sites (i.e. download.com).

The Design Review installation process

Just to make things even more convoluted you must take note of a serial number discretely printed on the download page, then browse to another web page altogether to obtain the highly recommended Service Pack 1 installer. Maybe it is just me but if you are going to the effort of heavily publicising your now free product wouldn't it also be a good idea to help this flood of new users out by slipstreaming the service pack into the installer and removing the need for a serial number? Hopefully with the release of Autodesk Design Review 2008 in April these things get cleared up, because if Design Review is to really take off (which I think it could) it has to be painless to attain and install. And please Autodesk if you are listening drop the idea of a 'Service Pack' for a product with a yearly update cycle, either get it right the first time or integrate an automated update mechanism within the program like most software vendors are doing these days.

Finally getting into it

Things get a lot better once you move past the overly painful installation process. On opening Autodesk Design Review you are presented with a clear set of options that help you get working with documents or learn the program.

The easy on the eye and brain opening screen (click to enlarge)

Interface wise Design Review feels like Adobe Reader crossed with an slim version of AutoCAD. To avoid cutting into the functionality of other products Design Review is not intended as a content creation tool and instead focuses on the viewing and annotating of documents primary created by CAD or BIM software such as Autodesk Revit. Those familiar with Autodesk products will feel right at home with the interface as it borrows significantly from AutoCAD right down to the appearance of many icons. For existing Autodesk users this is great for productivity but for those used to a competitor's product or who have not used CAD coming to grips with some Autodesk axioms can take a lot of time and patience.

Exploring the components of a DWF model (click to enlarge)

Design Review supports 2D/3D files in DWF format and a wide variety of 2D raster file types such as TIFF and JPEG. Design Review can also import DWG and DXF data into a new or existing DWF file. This is a useful capability to have during collaboration as it means the DWF becomes a bi-directional platform for information, quite unlike PDF which are fairly static without the purchase of a costly editing tool like Adobe Acrobat Professional.

The File -> Open dialog illustrating most of the supported formats

When a DWF file is opened you can browse through the various layers, objects and views. This allows the user to quickly locate or hide aspects of the document when taking measurements or annotating. Like most Autodesk products Design Review utilises side-panel to display the properties of a selected object and a run-down of a document's annotations. Again this is all very familiar territory if you are an Autodesk user but somewhat overwhelming if new to the product line or CAD in general.

Collaborating with Design Review

As the name of the product suggests its core functionality lies in reviewing designs. The review process takes place through a very capable callout system which enables users to highlight aspects of the design for commenting on or questioning. The DWF can be then saved and sent back to its original source for these questions to be answered. Design Review does a fairly good job of tracking all of this activity chronologically so it is easy to see who said what and when. However from a usability perspective it is a little confusing that when creating a comment it is referred to as a 'Callout' but on the side-panel these comments are listed as 'Markups'. This is a small thing by why the terminology change when a single term could be consistently applied?

Creating a callout, note the properties in the bottom left (click to enlarge)

Once you get the hang of creating callouts it is a fairly painless process to visually identify a design issue, comment upon it and email the modified file to the intended recipient. There is an email option direct from Design Review but it would be nice to see a toolbar button for this (as in Adobe Reader) so people do not have to look for it halfway down the File menu. Unsurprisingly Design Review has built in Buzzsaw support for opening from and saving to Autodesk's on-line service. What is a little surprising is that there is no direct promotion of Buzzsaw within Design Review. Considering the two go hand in hand together it would make sense to include an easy sign up function built into the software in the same manner Apple promotes the .Mac service within their own product lines.

A promising beginning but there is room for improvement

As a commercially priced piece of software Design Review only made sense for large or long term design teams with an established suite of Autodesk applications. However with this significant shift to a free model Design Review suddenly becomes a viable tool for nearly all AEC professionals seeking to digitally collaborate in a medium that has the accessibility of paper plans and the clarity of CAD. With this in mind I think there exists a number of areas where Autodesk Design Review needs to improve or change to fully realise its potential.

The installation

For clients and other non-technically inclined users to use from the comfort of their own desktop it must be as simple to attain as downloading a single, easily located file and mindlessly running an installer that asks nothing more than do you want to accept the default installation directory? No serial code, no service pack and certainly no system restarts.

The interface

Autodesk Design Review inherits its interface from a long lineage of Autodesk products. If you are a die hard Autodesk user is not such a bad thing, but if Autodesk is seeking to extend the use of Design Review beyond this tightly controlled customer base they need to give a second thought to the interface. To reach its full potential Design Review should be accessible to users of all technical abilities and software backgrounds. It is counter-productive to adopt an ugly and often confusing historical interface when in fact Design Review is not an extension of the AutoCAD franchise but the creation of an entirely new product line.

Office 2007's new interface may still be ugly but at least it is efficient

How could things be improved? For starters they could take a look at the way Google Earth handles 3D navigation and take a leaf from that. Setting up a view should not involve clicking on nested icons to achieve tasks that should be accessible in a far more intuitive manner. This would result in a new interface paradigm but as Microsoft has illustrated with Office 2007 such events can be beneficial. Through providing a dynamic toolbar the little utilised menu strip could be disabled in a similar manner to Internet Explorer 7, providing even more screen space for document viewing (the purpose of the application).

The clean interface of Google Earth coupled with intuitive mouse controls makes it easy to pickup even for inexperienced users (click to enlarge)

By default there is also far too much screen clutter for an application that is meant to be all about reviewing designs. There are too many toolbar icons visible especially considering half of them are generally disabled. The interface would be so much clearer if the interface was dynamic to suit your current task (i.e. like Office 2007 and many OSX applications). The sidebar works okay but this would also benefit from a degree of customisation depending on the user's current activity and what is present within the opened file. For example when viewing or creating a callout in a 2D view why not hide the Model, Animations, Cross Sections and other non-essential side-panel sections so that users can focus on the Markup and Properties panels? This would ease the need to scroll around the side-panel and remove distractions to navigation and commenting by an inexperienced user.

A Web-centric approach to markups/callouts

The callout functionality within Design Review is excellent but this discussion is buried within the DWF and cannot be extracted or viewed independently of the file. This results in a lot of data transfer within a design team for what equates to a dynamic text conversation which is referenced to a static document. A more efficient and functional approach would be to provide the option to host the markup conversation on-line, allowing the DWF to be distributed throughout the design team once and from that point on used as a static point of reference (see the below diagram).

Diagram of proposed online callout functionality (click to enlarge)

An approach similar to this is used quite successfully at blogger.com when providing comments on a post. Comments are created and exist in a central location and are referenced to by the respective blog. Externally referenced markup would increase the responsiveness of Autodesk Design Review in team situations as new callouts would be pulled down live from the Internet without having to redistribute the entire DWF file. It would also make searching and monitoring of the design review process easier because existing Web search and RSS technologies could be applied to the on-line data, effectively making the DWF a hub around which many other services could be offered.

A Web-centric approach would also enable a more fluid design review process whereby a handful of reviewers could actively comment on the progress of a design. The current operational model is engineered for a simple one-to-one exchange and would get unwieldy if more than a couple of people want to partake in the conversation at the same time.

Cross-platform compatibility

Autodesk is in many ways a bigger supporter of Windows than Microsoft itself, but in order to gain universal acceptance they really need to drop their silly 'Windows or the highway' approach to software when it comes to DWF. If people are to use Design Review they have to be certain any current or future member of their design team will be using Microsoft Windows. The software becomes of no value to the process if just one person in the team cannot access and add to the discussion. As soon as the Windows mould gets broken DWF goes out the window (pun intended) and is replaced by a format that everyone can effortlessly read, PDF supported by email.

Who could this non-Microsoft person be? The first user group that springs to mind is the client. Whilst not dominant in the general office environment Apple has a strong home, upper management and IT professional base. Also on the other side of the team the many technical specialists (simulators, technicians, engineers) may also use Linux. Therefore if Design Review (and DWF) are to succeed at all levels it needs to support the three primary platforms: Windows, Linux and OSX. This support cannot be a token gesture either, if it is done correctly the same set of functionality must be shared across all three platforms to ensure all members of a design team are equal partners.

PDF support

This may sound a little strange considering Autodesk and Adobe are currently pitching their DWF and PDF formats against each other but in the long run it makes a lot of sense. Even though it probably pains Autodesk to admit it PDF is a widely used format in the AEC industry for everything from product specifications to design plans. Autodesk Design Review needs to support importing this information so that users can build DWF files containing designs, specifications and all the other useful pieces of information that go along with the review process. Autodesk Design Review needs to support the exporting of structured PDF information because of this same fact. Design Review must behave as a first class citizen if it is to be treated like one within the general AEC community.

Conclusion

Autodesk Design Review is a very promising tool for Windows based design teams seeking a simple way of exchanging design information, questions and opinion. The tool is free, powerful and relatively easy to use by those somewhat literate in CAD. The markup tool set makes it easy to hold fairly complicated design discussions but it does incur a heavy cost in terms of the data sizes communicated between team members. There are a number of areas where Design Review could be improved and I hope in the forthcoming releases we see some, if not all, of these issues seriously addressed.