ODA shoots itself in the foot

To justify the display of the now famous (or rather infamous) “TrustedDWG” alert message, Autodesk claims that DWG files created by ODA’s DWGdirect libraries causes AutoCAD to become unstable. This is an excerpt from their initial complaint.

One problem with third parties’ implementations of the DWG format is potential data corruption resulting in instability introduced into the files by competitive products. Autodesk’s customer support personnel have logged numerous instances of Autodesk customers receiving DWG files from outsiders, then attempting to open the files with AutoCAD software, only to encounter serious errors.

In their reply, the ODA refuted by saying: “Autodesk has offerred no empirical evidence that its files are any more stable than others”. Indeed, they didn’t. And I wondered why, especially since any evidence, however small, could make their case rock solid. I investigated a little and came up with something truly interesting.

This is part of the DWGdirect 2.0.3 Release Notes (2.0.3 is the release in which the ODA added support for DWG 2007):

Bugs Fixed in 2.0.2
3565 DD writes DWG file which crashes AutoCAD.

3565 appears to the bug number and DD stands for DWGdirect. After digging more I found these bug fix entries for other DWGDirect releases:

1742 Linetypes with text are incorrectly saved to DWG.
2008 Acad crashes on a file saved by DD, containing an xref.
2211 AutoCad crashes on file saved by DDT – proxy issue.
2780 OdaMfcApp saves an invalid R12 file, which crashes AutoCAD.
2782 File saved to DXF 13 can’t be loaded by AC2004.
2783 File saved to DXF13 can’t be loaded by AC14.
2887 Acad crashes on attached file. DD audit/recover find no errors.
3150 Saved DXF file crashes AutoCAD.

The list goes on and on. Here we have the ODA (very honestly) admitting that their libraries cause AutoCAD to crash. They just shot themselves in the foot. Autodesk does not have to prove anything. The ODA did it for them.

This looks like clinching evidence to me. A smoking gun with lots of smoke. Its baffling why the Autodesk attorney’s didnt use it in their initial complaint. Could it be that they didn’t know about it? Seems hard to believe.

  • Anonymous

    Looks like you should justifiably receive a job offer from Autodesk to join their legal department! Excellent detective work!

  • Anonymous

    Looks like you should justifiably receive a job offer from Autodesk to join their legal department! Excellent detective work!

  • Ron Prepchuk

    As a developer of IntelliCAD I can point to numerous instances of users sending me drawings developed completely in AutoCAD which have errors that are uncovered when they open them in IntelliCAD. Sometimes the drawings will have hundreds of errors that AutoCAD has ignored. I don’t think their ‘TrustedDWG’ is all that trustworthy. In fact, I am kind of glad that a judge has issued a TRO against the ODA, whose libraries I use. I would like Autodesk to be the only company that can use the TrustedDWG watermark. That way, when IntelliCAD opens a file and discovers that AutoCAD has generated many errors, IntelliCAD can issue a warning to the user informing them that the file is from a source that cannot be trusted!

    Ron Prepchuk, President
    Autodsys, Inc.

  • Stefan Boeykens

    I’m not familiar with any software development project that doesn’t have lists of bugs. The fact that the ODA at least documents them and makes the file documentation available is worth much more then Autodesk refusing to adequately make the format available. Moreover, none of the competitors of the AutoCAD product are allowed to use the “RealDWG” libraries from Autodesk.

    I have often needed to open a DWG file into another application to resave it as DWG (obviously using the DWGDirect libraries) to make sure AutoCAD is able to open it.

    And you cannot claim that AutoCAD itself doesn’t corrupt DWG files. In fact, when all seems to go wrong with a DWG file, created and modified in AutoCAD alone (a real Trusted DWG so to say), one can sometimes recover the mess by opening it in MicroStation or a similar non-Autodesk program.

  • Stefan Boeykens

    I’m not familiar with any software development project that doesn’t have lists of bugs. The fact that the ODA at least documents them and makes the file documentation available is worth much more then Autodesk refusing to adequately make the format available. Moreover, none of the competitors of the AutoCAD product are allowed to use the “RealDWG” libraries from Autodesk.I have often needed to open a DWG file into another application to resave it as DWG (obviously using the DWGDirect libraries) to make sure AutoCAD is able to open it.And you cannot claim that AutoCAD itself doesn’t corrupt DWG files. In fact, when all seems to go wrong with a DWG file, created and modified in AutoCAD alone (a real Trusted DWG so to say), one can sometimes recover the mess by opening it in MicroStation or a similar non-Autodesk program.

  • Deelip Menezes

    I have never said that Autodesk software is divine and the DWG files created from their software come error free. I am simply commenting on the legal aspects of the law suit. Otherwise the title of my article should have been “AutoCAD only makes good DWG files”.

  • Deelip Menezes

    I have never said that Autodesk software is divine and the DWG files created from their software come error free. I am simply commenting on the legal aspects of the law suit. Otherwise the title of my article should have been “AutoCAD only makes good DWG files”.

  • Anonymous

    I am not sure I see the revelation in a list of bugs reported by the ODA that cause AutoCAD to crash.

    Files created by AutoCAD cause AutoCAD to crash. This is commonplace. I am sure Autodesk has a much longer list of bugs, by many times, related to AutoCAD-created files that cause AutoCAD to crash.

    You quote that the ODA refuted Autodesk’s complaint with the statement, “Autodesk has offerred no empirical evidence that its files are any more stable than others.” Then this list of bugs documented by the ODA is provided as some sort of proof that the ODA’s statement is not correct.

    The ODA did not claim their files do not ever cause AutoCAD to crash, they simply said Autodesk was disingenuously claiming that files not created by AutoCAD were more likely to cause AutoCAD to crash and that no evidence was provided by Autodesk to prove otherwise.

    Am I missing something?

    I contend if the truth be known it would be immediately evident that by far most crashes documented in Autodesk’s bug database are due to files generated by AutoCAD or other Autodesk products. Maybe this is why they did not rely on this sort of evidence, once it is introduced, it would open them up to revealing the magnitude of their own bug lists and the true cynical and hypocritical nature of this underhanded marketing ploy. Maybe this is yet to come in the jury trial if it happens. Let’s all pray for such a righteous conclusion to this shameful episode.

    This argument really makes no judgment on Autodesk’s software quality, software is complex and crashes are part of the business. I expect all vendors probably have more crashes due to their own files than those generated by others. It just puts the lie to Autodesk’s claim.

  • Anonymous

    I am not sure I see the revelation in a list of bugs reported by the ODA that cause AutoCAD to crash.Files created by AutoCAD cause AutoCAD to crash. This is commonplace. I am sure Autodesk has a much longer list of bugs, by many times, related to AutoCAD-created files that cause AutoCAD to crash.You quote that the ODA refuted Autodesk’s complaint with the statement, “Autodesk has offerred no empirical evidence that its files are any more stable than others.” Then this list of bugs documented by the ODA is provided as some sort of proof that the ODA’s statement is not correct.The ODA did not claim their files do not ever cause AutoCAD to crash, they simply said Autodesk was disingenuously claiming that files not created by AutoCAD were more likely to cause AutoCAD to crash and that no evidence was provided by Autodesk to prove otherwise.Am I missing something?I contend if the truth be known it would be immediately evident that by far most crashes documented in Autodesk’s bug database are due to files generated by AutoCAD or other Autodesk products. Maybe this is why they did not rely on this sort of evidence, once it is introduced, it would open them up to revealing the magnitude of their own bug lists and the true cynical and hypocritical nature of this underhanded marketing ploy. Maybe this is yet to come in the jury trial if it happens. Let’s all pray for such a righteous conclusion to this shameful episode.This argument really makes no judgment on Autodesk’s software quality, software is complex and crashes are part of the business. I expect all vendors probably have more crashes due to their own files than those generated by others. It just puts the lie to Autodesk’s claim.

  • Deelip Menezes

    Anonymous said: “… by far most crashes documented in Autodesk’s bug database are due to files generated by AutoCAD or other Autodesk products. Maybe this is why they did not rely on this sort of evidence, once it is introduced, it would open them up to revealing the magnitude of their own bug lists…”

    Quite an interesting insight. If that is indeed the case then it would be interesting if the ODA dared Autodesk to disclose their bug list in court.

    By the way, I tried to get my hands on some release notes of AutoCAD, but couldn’t find any. If you happen to come across any let me know.

  • Deelip Menezes

    Anonymous said: “… by far most crashes documented in Autodesk’s bug database are due to files generated by AutoCAD or other Autodesk products. Maybe this is why they did not rely on this sort of evidence, once it is introduced, it would open them up to revealing the magnitude of their own bug lists…”Quite an interesting insight. If that is indeed the case then it would be interesting if the ODA dared Autodesk to disclose their bug list in court.By the way, I tried to get my hands on some release notes of AutoCAD, but couldn’t find any. If you happen to come across any let me know.

  • Ron Prepchuk

    As a developer of IntelliCAD I can point to numerous instances of users sending me drawings developed completely in AutoCAD which have errors that are uncovered when they open them in IntelliCAD. Sometimes the drawings will have hundreds of errors that AutoCAD has ignored. I don't think their 'TrustedDWG' is all that trustworthy. In fact, I am kind of glad that a judge has issued a TRO against the ODA, whose libraries I use. I would like Autodesk to be the only company that can use the TrustedDWG watermark. That way, when IntelliCAD opens a file and discovers that AutoCAD has generated many errors, IntelliCAD can issue a warning to the user informing them that the file is from a source that cannot be trusted!

    Ron Prepchuk, President
    Autodsys, Inc.