bug = hidden defect ... Don't exaggerate either!
bug = hidden defect ... Don't exaggerate either!
Thom@s : we are not far from the hidden defect with the bugs that have persisted for years without being repaired. When almost all users advise to wait at least for SP3 before installing the software, it's because there's a big problem, right?
SW is more focused on adding new features than fixing bugs.
They would be better off releasing a version every 2 years but better finished.
Hello
I say that when a car manufacturer notices a defect on its vehicles, they call them back and repair them for free, no need for a maintenance contract.
May the force be with you
Hello
@OBI WAN, yes ok but if you break the gearbox after 2 months and 10,000 Km because you went from 6th gear to reverse on the highway at 130 Km/h, does the car manufacturer repair you for free?
Yes, there are bugs that recur systematically, on any workstation with any user and there we can call it a "hidden defect" because the publisher should have seen it, but I guarantee you that 75% of the bugs are due to the way SOLIDWORKS is used and the IT infrastructure that is used (network, server and workstation) and there it is impossible for the publisher to predict all the scenarios.
@+
Coyote says : "but I guarantee you that 75% of the bugs are due to the way SOLIDWORKS is used and the IT infrastructure that is used (network, server and workstation) and there it is impossible for the publisher to predict all the scenarios"
I want to take your word for it, but can you explain this to me:
In my company we work on identical stations cloned to the install (one is installed and duplicated on the others).
The stations are bought in batches at the same time, so even the serial numbers of the graphics cards sometimes manage to follow each other.
It happened to us that on one station some functions crashed while they passed on the other, and the opposite the next day...
I would like to point out that the user profiles are identical (replication from one workstation to another)
Where is the responsibility for the infrastructure?
The example that struck me the most: opening a file, creating a sketch and a material removal => "Solidworks has stopped working"
It's a far cry from the transition from 6th to reverse gear here!
Hello all, great debate! Thank you!
I answer.
Yes, for me, they are hidden defects. Because these are defects, which are not apparent, nor mentioned, when the license is purchased. It's clear.
Then, yes, Stefbeno, I'm alone, but I've had training, including one recently at Avenao, to try to solve the implausible slowness of SW. I often read "the system is not responding", as it is so slow to solve my assemblies. The trainer did not see any anomaly in my method, he simply advised me to reduce the external references. I spent two days doing it! And it's only slightly better.
As for the bug of "Solidwors has stopped working", when creating a sketch in place (in the assembly), it happens to me all the time. I'm relieved to see that I'm not the only one.
Another detestable flaw (not a bug, I admit): when you create a sketch, it happens that a very small piece of segment is done without your knowledge, depending on how you stop the sketch with the mouse. This kind of thing is perfectly repairable by SW. They don't care. And this generates, for beginners, time wasted and misunderstood.
In my company, we design measurement software ourselves. Such m.. for our customers.
There you have it, to be continued!
Regards to all
Hello
I confirm that PTC's pricing policy is the same.
I went a long time without maintenance and had to pay an arrears when I came back under maintenance.
The worst thing was that we wanted to acquire a new link and for the compatibility of the files between the different workstations, we had to update everything.
On the other hand, as far as bugs are concerned, I don't get too full:
* I have a few crashes (abrupt closing of the software), the last week during a disk access, so possible that it is an internal server problem.
* The most delicate functions are when creating shelves on complex parts with drafts, but knowing this, a regular backup is sufficient.
* There are still some incomprehensible crashes (I would say once every 6 months) during a basic action such as redefining a function, and which recur every time I restart the software and restart the action on this part.
What surprises me about SW is the feedback during SOLIWORKS WORLD 2017, numbers 1 and 2 of the Top Ten List, focus on bugs and software stability...
http://www.lynkoa.com/actualites/solidworks/solidworks-world-2017-jour-3-nouvelles-fonctionnalit%C3%A9s-solidworks-2018-nouveau
S.B
If I continue to play devil's advocate!
For my part, there are no recurring problems or crashes to report in a "sketch" context, except maybe once or twice when importing and converting DXF into a sketch.
For the question of "The system is not responding" you may also have to question the machine, the installation, the antivirus, the network...
Do you have enough RAM? Powerful enough graphics card? Which operating system? Which processor?
Hello
@Camexaspere you say
"Another detestable flaw (not a bug, I admit): when we create a sketch, it happens that a very small piece of segment is done without our knowledge, depending on how we stop the sketch with the mouse. This kind of thing is perfectly repairable by SW. They don't care. And this generates, for beginners, time wasted and misunderstood."
New in SOLIDWORKS 2017, it is no longer possible to create your own detestable little lines (yes it happens to everyone) this has been corrected, which shows that they don't give a damn about their customer's problems at Dassault Systèmes.
@+
External references are absolute evil!
It creates a monstrous load for the machine: when you open a room, it opens all the referenced files in the background (and it recalculates them in case of modification). At worst, you may be able to create circular refs (normally detected by the software).
Either your trainer didn't understand that you were making significant use of it, or he wanted to spare you, or...
When I create them, it's just enough time to recover the position of holes (for example), and I immediately take the sketch concerned to remove the external refs.
@stefbeno +100000.
External refs are kaka, occasionally it's very good but keep them NO!!
may the force be with you.
I don't agree with you about the external refs.
You have to look at the context.
We have been using them for our products for several years without problems or slowdowns.
Sketches in the assembly control the machining or dimensions of several parts and it works very well.
On the other hand, the parts are only dedicated to this assembly and it does not have more than 150 parts, only about ten of which have external refs. We have a model assembly that we use as a base and that we duplicate for each new study.
There are a bunch of functions like this, commercially promoted by SW, which have limits to their use but which no support tells us about.
Summary: the function is great, but you shouldn't use it too much
@Pascal: we totally agree! 10 pieces out of 150 with external refs in a controlled setting, it can work. But, without any pejorative intention, 150 parts in an asm is cat pee for SW, in a special machine, it's not even the minimum for a station of a machine that will have a few dozen (stations).
And I think Daniel is more in the second case...
Other practices that can lead to performance drops:
- sketches that are not completely constrained;
- components (classically screws) that retain a degree of freedom;
- flexible assemblies.
You have to tell yourself that if it can move, SW will try to determine the position, see if it is impacted by another element
In general, I avoid the first 2 and limit the last one as much as possible (but it's very practical).
@stefbeno : It's clear that it's not big blends. It's to show the (ridiculously low) limit for it to work properly.
In the case of special machines, as you don't start from a model, the use of external refs is limited.
The screws, which keep rotation possible, are a pain! And when you use the toolbox it's even worse. (another nice example of a practical function not to use)
For flexible assemblies, some colleagues use them but I don't use them
Another heavy function to manage: mechanically welded constructions with complex profiles (Elcom type: full of radii and internal surfaces).
For flexible assemblies, it's normally reserved for commercial parts like cylinders, robots (the configs are sometimes impossible to manage). The problem is that sometimes, these assemblies end up at the 3rd or 4th level compared to the final asm (which is rarely opened by the way).
Hi all!
Thank you for your answers and the dialogue.
About the external references, yes, the trainer told me to use them to create the piece, and then break them up. So, the assembly is no longer associative... CAD software must be associative!! If not, what is it?
Thanks for the info about the heaviness coming from badly constrained sketches, or pieces that still have freedoms, are you sure? In this case, I am concerned.
As for the Toolbox, I forgot about this implausable, shabby thing that even training pros disdain.
SW's correction concerning the vicious little line that is put at the end of the sketch without our knowledge... corrected in 2017 only?? And you think it's good, to wait so long for such a gross and minimal defect??
I remember how much of a party it was when I received a new TOP SOLID update patch. Lots of fixes, improvements. SW's updates? Fro... always the same defects that remain. No, I can't go back to TS, too expensive, too heavy to reform. Am younger!
My machine is good, to respond also to a suggestion. I did the performance test, we're in the first quarter. QUADRO 4000 graphics card, i7, 16GB of memory... W7 pro 64
Have a good day, thank you again, cordially,
Daniel
The software is associative, but then it's the machine that no longer takes responsibility.
I'll try to draw a parallel that should speak to you:
You buy a milling machine, the salesman told you that it could hold a certain speed.
You install it in a barn (with a good slab anyway) and use ARS cutters. This is a configuration that is suitable for machining S355.
Now if you want to machine 40CMD8, in the same conditions, it will go badly.
It's been 25 years since I discovered associative CAD. It was even a course (at the end of the DUT) on Pro/Engineer (now Creo) that made me want to push my training in this field, I found it magical. My next training was solely focused on CAD/CAM. It shattered the dream a bit: our teachers made us aware of the limits of software AND machines.
Software, like any tool, has its limitations, especially since it depends on other tools, in the same way as if you are looking to manipulate a very large hammer without having the muscles to do so. Unfortunately, I think you'll have to accept some limitations, review your methods of using the software a little (just the leads you are given should be enough). It's clear that it won't be miraculous, there will always be bugs, waiting times.
Currently, with a configuration similar to yours, I am working on a small project (230 components including 2 mechanically welded parts in complex profiles, 2800 bodies, 1 5-axis robot in flexible mode), it takes between 1 and 2 minutes to open (through the network which is not optimized). I reopened the last big file (more than 5000 components, 15000 bodies, mechanically welded parts in complex profiles) in 4-5 minutes.
For updates, publishers play the one who pisses the furthest, announcing annual updates, quantities of functions. It's like car manufacturers, cars are always more powerful and claim to consume less. There is a demand from some (too many) customers...
Thank you Stefbeno, moderate and probably true analysis.
Nevertheless, DS's attitude is, for me, unacceptable.
Nevertheless, this software has been dragging bugs and major flaws, apparently for years, and we are taken hostage.
Good evening everyone, thank you again.
Hello, this is a bit out of the scope of your initial question, but since it could help you, I take this opportunity to share it with you:
"My machine is good, also to respond to a suggestion. I did the performance test, we're in the first quarter. QUADRO 4000 graphics card, i7, 16GB of memory... W7 pro 64"
This is not true, the performance of your machine seems good, but its reliability is not intended for the use you make of it!
I myself have on my personal computer an i7 haswell with 16 GB of ram and a Quadro 4200, it's an acceptable configuration to work from home, but in business, it's exposing yourself to problems ...
Indeed, the i7 is efficient, but their reliability with calculation software such as "mechanical" CAD software leaves something to be desired.
For software that does a lot of calculations like CATIA or SolidWorks, a Xeon processor with a professional motherboard, and especially ECC ram sticks, are the assets of a professional configuration that works properly.
https://fr.wikipedia.org/wiki/M%C3%A9moire_%C3%A0_code_correcteur_d'errors
So yes, it's relatively expensive, but there are second-hand markets for this type of computer equipment, and sometimes it's better to have a computer that's a little slower but reliable, than a beast that ends up in the wall 3 times a day...
To conclude, "There has therefore been a unilateral modification of the current contract. ", I haven't read the software contract, but I guess Dassault is smart enough to have included sufficient closures to protect itself from any recourse.
Good luck, and have a good day
Hi @ Meacnic +1 for your answer
The fact is and tested right here
an I7 is good but not suitable for logs like SW
but a xeon with ecc memory is 10 times better
The test was done via repetitions in xeon it goes in i7 it crashes
and your idea of second-hand that I've been defending for a long time works very well
I gave x times viable addresses of workstation desktop or laptop
tested and approved by forum members there is only good feedback so far
Have a good day
@+ ;-)