Here are two files y'all can look at and see the inconsistency with SW.
Loft 3 and 4 originate from SW2001+ and later SW2003sp5, so it seems the loft algorithm is the same and maintained here.
Unfortunately, when a loft is re-created using the same edges in SW2004 sp2.1, they look like crap!! Why!?
Another different inconsistent algorithm which screws around with the users data and time = totally unproductive!!
..
looking at it I think this is actually the same distortion as the snowboot...seems to me its not an inconsistency but rather a flawed algorithm....its just going sideways in between...you cant actually straighten things out by playing with settings..might also say here I had a simple sweep that failed unusually but I haven't had time to look into it to see if that was me or not.
Neil.
Paul's point is that his perfectly good data (which he has spent time/money on) is being screwed up from one SP to the next. The fact that it CAN be fixed is totally irrelevent. This costs him more time/money, and is inexusable. For a independant contractor, or a small consulting firm (like mine) this is a huge deal. We bill by the hour.
SW has never understood this concept. I started keeping track of rework costs directly related to migration failures around early 98. To date, this has cost my company well in excess of $150,000.00. Most of these losses were between 98 and 02. Since then, I've managed to control these losses by only upgrading on an as needed basis. That is, If we get a new client who uses a later version, I will isolate their data and assign specific engineers to work on that project. These guys will have that version on their machines, and I will give them very specific instructions with regards to it's use. Of course accidents happen, but luckily, all of our data is backed up every night. Our current production version is 03 SP5.1. It will probably stay that way untill I have the budget to upgrade most of our workstations.
This means that we have to have multiple versions installed on most machines. As you can imagine, this makes my life VERY complicated.
SW has never understood these problems because they don't do real world production design for a living. Also, large percentage of their customer base doesn't do anything complex enough for most of these serious inconsistencies to be a real problem.
If more users were to do a comprehensive "cost of ownership" audit, they would be horrified.
Regards
Mark
I am sorry I didn't realise I said anything to contradict Paul...I thought I was agreeing with him... in fact these examples seemingly cant be fixed because the algorithms in 2.1 are stuffed?..2.2 coming?...I am very well aware of the lost time etc due to a number of SW shortcomings...I thought I made my views known a few days ago... against the run of play in some peoples eyes. in many cases it would a lot more productive if I just got out a bit of old fashioned paper and ink! cheers (fellow sufferer trying to do everything well on my own)
Mark,
I think Neil understands it and see he just posted. Whether the algorithm is flawed or inconsistent, well, it's consistently messed up that's for sure!
As for loss of time,.. wow... that is something we independent consultants are very aware of!!!
What is true is SW Corp has shown they "CAN NOT" consistently fix the loft problems as well as sweep and fill inconsistencies!!!!
Clearly it has been an issue and continues to be an issue, that is a fact!!
If SW Corp "did" care about the loft, fill and sweep issues, then it they would "not" be issues!?
But, they continue to be issues per release and per sp's!
For you who are reading this and considering using SW or buying SW, "beware", your ROI on your data per release and service pack is a joke!! This is a FACT!!
For you SW investors, your company continues to SCREW THE USERS in this area!!!!
..
MM wrote:
Neil,
S..OK... I probably over reacted myself. Guess I'm too close to the pain.
I don't know why they just don't go out and license a component like GeomWare "Nilb"
These people (and many others like them) have already figured all this stuff out. Rhino is based on such a component.
Trying to re-invent the wheel isn't cost effective or fair to the users. It would be a different story if they were doing a good job.
Regards
Mark
P.S. Thought you went fishin
My only recourse. Stop paying subscription fees and let them know why. Harsh yes. Limits what I can do, yes. Stuck on an earlier version, yes. But since I am only in SW for maybe 30% of my time anymore I can NOT afford to waste any time bothering with stupid crap like this. I put up with it in the early days, but not anymore. I'm not willing to pay them for being a beta tester anymore. They may care about it, but don't understand the math well enough to fix it.
clay
Paul Salvador wrote:
PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.