Quality Software?... Oxymoron?

In your case 99.9999%

After decades of using software, my only concern is when those bugs are not fixed pronto, in the high end programs, where I rely on the software to make a living.

Bo

hmm,...the measure of quality,.. a relative offset of sustainable values,.. or maybe,.. a submission to ones limited realities within ones perfect balance of attainable social-economic needs, ..?

f&#k that sh*#,.. just give us something with performance and is consistent!?

ah,... never mind... my hope brain cell was acting up again.

..

I had a thought the other day about the recent versions of swx and I've been hashing out what it might be like to go back to using maybe sw2001. I worked on a particularly large assembly back then on a P3-750 with 256RAM. If I were to load up that assembly now in sw2007 with my P4 whatever and 2gb RAM, etc., I'm thinking it would be too slow to realistically think about doing drawings and the other usual work. My theory is that sw2001 should seem lightning fast on a modern PC and the speed would more than compensate for a few "time saving" features that have been added along the way. I do machinery so the curvy stuff features aren't a concern. I think the delays with sluggish software have a strongly negative influence on a person's focus, blood pressure, etc that kill productivity. There's a fair percentage of my work that doesn't involve swx files from customers, although I do work with customer's parasolid files pretty often and occasional step files. The biggest drawback I can think of is being out of the loop on library stuff. With access to a late version, I suppose one could export library models into x_t or whatever for the old version. I'm not sure sw2001 would be able to load x_t files exported from sw2007. Anyone know of this being tried at all?

bill

Bill, you may be onto something which might actually have a chance of working. I know they upgrade the ProE & Unigraphics format over time, but IGES might pull through.

The 2nd thing would be for SolidWorks to implement.

Why is it that every single improvement needs to be in the main program with its silent overhead, when you don't need it? If a user like you, me and others are doing a project with prismatic solids, why couldn't we choose to leave out surfacing and any other modules, which we don't currently need, so our work flow improves. We could simiply uncheck "Addin" modules we don't currently need.

Obviously that is easy for me to say, and not necessarily easy to do. Still, Pro E & Unigraphics have a lot of addins they sell.

Hence, I think there is a reasonable chance that speed and stability could improve if we loaded only what we needed.

Bo

we,.. they,.. us,.. them,... solids,.. surfaces,..

ok,..I'm seeing the pattern.

The undertones of separation and prejudice runs deeeeep...... in the solids only world....

.. 8^)

are you one of those radical curvists?

I'm telling you, there is a reason I still use 2004.

TOP

truth be told,.. the cult I belong too prefers to be called... swooshist.

.. ;^)

you don't need to use swx files from others (pretty much the biggest concern for me) ?

bill

Not usually. If I do, and it is mainly vendor parts, I can always use a newer version and create a dumb solid. You can do quite a bit in an older version and then convert at the end if necessary. So I would say use the older version for productivity and switch to a newer version if needed.

TOP

You're using sw2004, I believe. Are rebuild times noticeably faster on your machine than on new versions? Is your machine kind of older or pretty up to date?

Bill

Let me put it this way. For a lot of conceptual stuff I am using a laptop I got from a VAR. It was fine back in the 2001 era. It runs

2004 just fine for small assemblies and reasonable parts. I tried to put 2006 on it and it choked during install. 2008 is out of the question. For one thing it is running software opengl because is has a puny 8mb ati graphics card. My STAR benchmark was developed on it and is pretty slow.

So since this laptop could do usefull work in 2001 thru 2004 but can't for 2008 my conculsion is that performance has dropped like a rock. To put it another way, if 2004 was run on a machine that was fast under

2008 it would be blazing fast under 2004 (and more stable).

I never thought I would see the day when SW was competing against itself.

TOP

Interesting. Has your swx2004 had any trouble loading a dumb solid export from recent software? I think I'm going to give it a try. As I remember, the final SP's of 2001 were pretty solid. I don't remember being too enthused about 2002 or 2003. I think I still have the SP files and everything for 2001 (I hope). If 2001 is too primitive in some way, I could bump up to 2004 and still be fast, it sounds like.

I've seen it suggested a number of times that swx should offer swLite, which could essentially be an older (faster) version with less capabilities, but modern file compatibility. Maybe one of these days, an open source group will do that :) Thanks, Bill

Actually, dumb solids seem to do better in 2008. So there are reasons to have both around. My conceptual work doesn't have much to do with dumb solids. Just modeling.

TOP