Hatching and Poly line joining Still Crap after 20 years.....

[steve.coleman]

I have been using AutoCAd since 1989...  And in all that time hatching and polyline joining is unbelievably crap.  I was originally taught microstation in the late eighties and at that time microstation was unbeatable.  Especially hatching and patterning.  But Autodesk got their game together and overtook microstation to the point that Bentley really gave up on general CADD.

But.....To this day I still can’t hatch without re-creating an uninterrupted polyline around a complex series of entities.  We are Civil engineers in Australia, so we work in meters in an co-ordinate system that is no way close to 0,0. So in Australia we have learnt to manipulate the UCS/WCS and change the SNAPBASE to something that works with the architectural hatching.

But.......Trying to join polyline, even with an fuzz factor. ( Reminded: that for a number of years we all used a  3rd party Lisp routine that did the same thing) to this day it is almost impossible to get to work without editing, manipulation, or just tracing a polyline over.

Even picking up the end grip of an entity and snapping it to the next is not enough to join. 

The” boundary” command still does not work given large areas.  I am talking about 5km long by 11m wide areas that have been created in 2m segments. Even with the Fuzz factor in the Polyline edit command. Really its mathematics its not that hard....

Come on Autodesk this has been going on for decades. 

Civil construction tolerances are to the nearest 10mm. So working in 1 unit = 1m means 0.010 tolerance.  Do we really need Autocad to work out the vertex co-ordinates to 16 decimal places.  This was the excuse I was given by an Autodesk Representative a number of years ago.  And quite frankly given the success rate in trying to join the entities I can believe it.

I have just completely lost it and yanked the mouse out of the machine (damaged the neighbouring  USB port, adjacent HDMI video and scaring the wife and dogs) and thrown it across the room.  It may have also something to do with just paying  $7.23262 AU for a an additional license and subscription for maintenance for another year.

I have given up ringing support given that the support personal have no actual experience and they seem to be following a “script”. They get to the point of saying .”Yes that seems to be known issue and Autodesk are looking it to it”… 

Really after 20 years.

[rkmcswain]

1. I agree, it's silly that we have to fight with hatching and closing polylines in year 2014. It seems to be somewhat better than 10 years ago, but I still find it quicker to draw a closed polyline - even if that take a few minutes.

2. Why not drop Autodesk and go back to Microstation? (serious question, not trying to be "smart")

[hwalker]

Have you tried using the HPGAPTOL variable?

This will hatch even if there is a gap between elements. You have to state the gap in units.

eg. you have to objects which are 5 units apart. If you set HPGAPTOL to 10, it wll hatch between those two objects.

[pkolarik]

I'm not sure why you're having all these troubles after all this time. We do civil projects here all the time and have no trouble joining plines (provided they're all at the same Z elevation) and hatching of complex areas works just fine probably 75% of the time or more (even without drawing a pline around the required area, but with drawing a pline around it it works 100% of the time of course), provided we're not attempting to use xrefs for the hatch boundary.

[steve.coleman]

Hi HWalker "

Yes we have played with HPGAPTol variable.  Its been around for a long time.  I have used it in the early 90's to definen hatching as we plot to printers that supported the HPGL2 plotting launguage.It gave us a better quality of final print than the defaualt Autodesk drivers.  (Seems to be a trend here).

We find that even though the coordinates of end vertices are shown as the same. in x,y & z,  is still will not join.  I am very tired of using variouis work-arounds around "like HPGAPTOL" when you pay nearly $8,000 Au for a license..... 

Thanks  for taking the time to respond.

Regards

Steve Coleman

Senior Design Consultant 

for Coleman Engineering Service Pty Ltd

[steve.coleman]

Hi Pkolarik,

You have have hit the nail on the head.

Is 75% of the time good enough?"??????  In my example the amount of area we have to hatch means that even with 75% we spend considarable amount  time trying to get the remaining 25% to join.  To the point where we just dont bother.  And construct a new polyline boundary from scratch.  At least we know it will work.

Again After 20 years, you would thin Autodeks could do better than 75%.  Especially given the cost of the license and subcription?????

Thanks for giving the time to respond.  Appreciate it.

Regards

Steve Coleman

Senior Design Consultant 

for Coleman Engineering Service Pty Ltd

[nrz13]

I think some of us have just become so accustomed to AutoCAD's failures over all this time.  Let's look at it another way:

The line command is fine.  It works 75% of the time.  The other 25% of the time, you can get it to work by drawing a rectangle, exploding it, deleting three of the sides, and then snapping the ends where you want them.  Like I said, it works fine. 

[Anonymous]

have you tried using osnaps while drawing

[steve.coleman]

Hi nrz13

I had to laugh  Because its true,,,  We start believing that the workarounds we come up with for the failures in the software is acceptable....  

I had a draftsman who had a little blue book that she filled up with all the little work arounds that I would teach her over the years.  I couldnt believe it when she said one day that she was on the second volume.....

The standard we walk past is the standard we are willing to accept.....

Thanks  for taking the time to respond.

Regards

Steve Coleman

Senior Design Consultant 

for Coleman Engineering Service Pty Ltd

[steve.coleman]

Hi Kenttr8

Good Point, however we are not always the orignator of the entites we work with.  Architectural drawings a case in point.....?????!!!!!!!  In most case if the entities are draw correctly in the first place there is no problem.  However this is an unrealisteic expectactation.   I recieve dwg models from Architects, etc.  and quite often we have to try and work with entites that have been draw by left handed larmas that suffer from dixlexia and parkisnons.....  

The Boundary command and HPGAPTOL should get around this but no..  As I have mentioned in previous posts. Even if the x,y,z list the same they do not join.   

Argghhhh...  More Red Wine..

Thanks  for taking the time to respond.

Regards

Steve Coleman

Senior Design Consultant 

for Coleman Engineering Service Pty Ltd

[steve.coleman]

Just read my previous posts.

                                                                              Sorry for the spelling mistakes.  Especially after the dyslexics comment.   I am just trying to say.  This is our software.  We pay for it with our time and passion and $.  We should not be putting up with the fundamental problems that Autodesk seem to be ignoring.  Us as users are their future. I can tolerate minor issues for at least a couple of service packs.  However, this is a fundamental problem going back 2 decades.  I remember a launch presentation in the mid 90's, in Canberra, ACT, Australia. . Where the presenter explained the fuzz factor being incorporated into the standard Poly edit command and huge cheer went up from the audience.

Bugger!!!!!

[3wood]

At least it doesn't make a lot drafters lost their jobs. Tell the boss we have to trace manually 15km polylines at 2m interval.