I originally reported the problem to ESRI in 2011, it was given a NIM and never addressed again. I do not believe they ever plan to address the issue.
Typical ESRI Support
According to my notes on the issue from August 2011. ESRI marked the NIM with DECLINED.
Unfortunately, ESRI's base appears to be dominated by major government departments and large corporations. For small companies they have no incentive to fix issue i.e. accountability to us is absent without leave. Their customer communications (for the same reasons) is also pathetic. We shouldn't have to chase significant problems where a base functionality - tools they included in the software they licensed us to use for many $ - fail to work properly. Beyond an acknowledgement that they reproduced my problem, I heard nothing about actions they were investigating to fix it. I understand it may be complex, but it's a tool, sold with the licence, that randomly doesn't work. According to Naomi (next post) ESRI marked the NIM as DECLINED; does that mean they're not interested? I guess so.
Agreed. Shame about the lack of courtesy in not informing me - as the original poster in this thread - that this may be the case. Why aren't I surprised?
It has been seven years almost to the day since the original post was generated on this forum, and the problem still hasn't been fixed. Since I commented on this forum about six years ago, I have completely revamped my teaching of GIS away from ESRI products for this precise reason. I won't mention the freeware I use by name (because I am pretty sure that ESRI's indifference to user concerns stops when rival products are mentioned in their forums), but it is the leading open-source product out there. The digitizer is a bit wonky, but works fine. Over the years, I have resisted teaching this software to students because I wanted them to graduate with transferrable skills, and the fact of the matter is that ESRI still dominates the GIS market. However, the freeware has a lot more flexibility than ArcGIS in so many ways because of the myriad plugins you can get. There are many online resources and user forums to help you troubleshoot problems--since the software isn't proprietary, there is almost always a fix generated in a matter of days. Given the complete indifference of ESRI to its customers' concerns, I strongly advocate those reading this post to give the other software a try. There is a bit of a learning curve at first, but it is not steep and you will not regret cutting the ESRI cord.
Without healthy competition in the marketplace, this is what you get... corporations that really don't value the user experience!
This has just started happening to me on 10.5 as well. None of the above suggestions or using classic snapping helps. It's just started after using the same laptop since 10.5 came out so it's really confused me as to why its suddenly decided that this bug is going to affect me.
Do you have many "arc" segments in the map? This happens to me when I have complex CAD data converted to feature class with lots of curved segments. Selecting only those I want to trace and checking "Trace selected features" worked, as suggested by others above.
Hi Anna,
Thanks for your reply. To clarify, I use the Bezier tool to draw polylines and polygons and often, as I have mentioned, the area covered by the map is large - country scale or larger, so there are many 'arc'. Continuing to zoom in before the trace tool would work again is of no practical use. ESRI did reproduce the error, which seems to have geographically bound component i.e. affects particular parts of the map *only* once it decides that it doesn't like that area. I can't define anything special that I do to cause the trace tool to fail - it seems to be random and can happen sooner or later in the compilation process.
I use the Bezier tool as it is quicker and more intuitive for drawing geologic unit shapes (for me at least) that will rarely, if ever, have a straight segment. I also use the equivalent tool when I draw in a vector graphics package, so it's always my go-to tool. Maybe if I only used straight line tool with the attendant issue of increasing the number of mouse clicks by orders of magnitude, the problem would go away...
I have tried a couple of suggestions in earlier responses to my original post, without success. I had effectively given up and used my work around - draw a polygon through existing ones and then clip it using those existing polygons. It's a pain, but work and is good for ensuring the topology is ok.
I'm on a much needed vacation at the moment, but over the next 3-4 months will be compiling several more maps and will try all suggestions again and report back.
Thanks for taking the time to post.
Cheers!