What could Autodesk change that would make you want to install to a new release

[kathy.oconnell]

Hi everyone--

If you are running a version of AutoCAD that is older than 2009, what version are you running and what is the main reason you have not upgraded to a newer version? (for example, format change, cost of update version, performance, installing and migrating settings from my current version to the new version, the time it would take to learn the new features, etc)

And second question--is there something Autodesk could "improve" that would make upgrading something you would want to do more often?

Thanks!

Kathy O'Connell
Sr. Product Manager--AutoCAD

[RobertB]

Answer 1: In cases where we are forced to use older versions of AutoCAD it is because the client's requirements are archaic. However, performance is large part of the resistance to upgrading in the minds of some of our staff.

Answer 2: Put at least twice the resources you currently do into F3 (beta testers know what this means). Just about any great feature introduced in the last 5-7 years has been taken to about 75-85% functionality and the left there to linger. Tool palettes, SSM, FlatShot, MLeaders, Solid history, Tables, Datalinks, STB support (10 years out on at least 1 bug), PMP files, and so on.

The happy part is that I think the tech support feature for Subscription customers is worth the cost of subscription as long as response time improves. I've got quite a list of open cases at the moment. :wink: The effort the teams have made to provide "goodies" for subscription users in the last year have been appreciated, although the verticals often are not supported like vanilla AutoCAD (boo!).

[cadtag]

at work, we're running 06 and 09, although we've been on subscription since at least 2002. As was noted, one of the big reasons for sticking with 2006 is client requirements - (although i refuse to call them archaic :-) ). They have their reasons and requirements -- many of the clients with earlier version needs are more interested in long term archival utility and stability, vs short term eye candy. (Note please, that stability does NOT mean "It doesn't crash that often", stability means it's functional and usable day-in and day-out for a decade or more. C3D's annual object changes make any stability claims in that regard laughable.) But I'm not here to bash anything....

F3! again, as noted, this is missing. It's difficult to become enthusiastic about new features and capabilities, when the last decades worth of new features and capabilities are still not finished. How enthusiastic can anyone be about the wonderful usability of parametric constraints, when MLines have been abandoned for years? If SSM editing is barely usable, and it's still impossible in 2010 to use a multiple line drawing title from SSM in a field in a title block? It's been 2 and 1/2 decades since I started running ACAD, and the basic ability to draw a leader is still not all the way there! Linetypes are vastly less capable aesthetically today, than Zipatone tape was in 1985. I still can't include a table in a block of Mtext, or include arbitrary graphics in either Mtext or a Table. Excel links are not all that useful to me compared to that.

Having a choice of two shades of gray UI (depressing gray and dismal gray) is not and never was acceptable. Please respect the user's window display settings, instead of over-riding them out of some odd sense of whatever -- it seems like nothing more than uncaring arrogance. It was certainly not meant that way (I hope) but the lack of response to the many requests to look 'normal' simply tells people that the 'Desk doesn't care.

Please review and take into consideration the thread "100 Paper Cuts" on this board. I think you'll find a lot of bang for your buck in there.

And really, to make a compelling upgrade, open things up. It's ridiculous that CTB and STB files are not human readable and editable. If anything does, they look like ideal candidates for XML formatting. If the .dst file format was properly documented, then the customizers and third party people could solve many of the deficiencies. Ancillary file types like these are hardly the family jewels! Please use open formats and document them.

And finally, performance. The 2005 version seemed to hit the sweet spot in terms of features and performance. Beginning with R2006, performance has decreased dramatically, and the new features are apparently a direct causative factor. I note that every since 2006 came out, the ongoing mantra here, on the Adesk boards, and from subscription support, has been to turn off the new features, from Dynamic Input to the Layer palette, to regain some of the lost performance. My hardware is much more capable than the P-III 800 I ran in 2005, but the software performs much more slowly.

Best of luck.....

[Mamma Jamma]

Personally, I've been using 2007 until very recently (and 2010 as a guinea pig). As people have gotten new computers, they've been upgraded to 2009. Our plan is to roll out 2010 in the next couple of weeks.
Generally we only upgrade every 3 versions, generally to maintain maximum compatibility with our clients (though IT doesn't understand why we should let our clients' requirements hold us back - uh..because they are our clients...they pay the bills?), but also because it's inefficient for us to have to deal with the bugs, learning curve, etc. every year.
I'm particularly p*ssed at 2009/2010 because of the major format change. As a long time user (since version 2.something) I have lived through many major changes, to say the least, but the ribbon concept seems to be among the worst ideas (feels like going backwards in the number of picks required to get anywhere) and changing what the icons look like is just stupid. While I do type primarily (thank God!), I do use the icons occasionally, and if I'm used to seeing, for example, a stick of dynamite for Explode, and now it looks like a box falling apart, well, you might as well have translated the whole d*mn thing to Greek.
So, what could AutoCAD change to make me want to upgrade more often? How about spending more time to work out all the bugs before releasing a new version (even our reseller tells us not to get the latest version for at least a few months after release), and quit screwing around with stuff that's perfectly fine the way it is. Are a new (depressing and ugly) color scheme and new icon designs really an improvement? Seriously?

[rkmcswain]

Answer 2: Put at least twice the resources you currently do into F3 (beta testers know what this means). Just about any great feature introduced in the last 5-7 years has been taken to about 75-85% functionality and the left there to linger. Tool palettes, SSM, FlatShot, MLeaders, Solid history, Tables, Datalinks, STB support (10 years out on at least 1 bug), PMP files, and so on.



F3! again, as noted, this is missing. It's difficult to become enthusiastic about new features and capabilities, when the last decades worth of new features and capabilities are still not finished.

Having a choice of two shades of gray UI (depressing gray and dismal gray) is not and never was acceptable. Please respect the user's window display settings, instead of over-riding them out of some odd sense of whatever -- it seems like nothing more than uncaring arrogance.

And really, to make a compelling upgrade, open things up. It's ridiculous that CTB and STB files are not human readable and editable. If anything does, they look like ideal candidates for XML formatting. If the .dst file format was properly documented, then the customizers and third party people could solve many of the deficiencies. Ancillary file types like these are hardly the family jewels! Please use open formats and document them.



I have lived through many major changes, to say the least, but the ribbon concept seems to be among the worst ideas (feels like going backwards in the number of picks required to get anywhere) and changing what the icons look like is just stupid. While I do type primarily (thank God!), I do use the icons occasionally, and if I'm used to seeing, for example, a stick of dynamite for Explode, and now it looks like a box falling apart, well, you might as well have translated the whole d*mn thing to Greek.

...and quit screwing around with stuff that's perfectly fine the way it is. Are a new (depressing and ugly) color scheme and new icon designs really an improvement? Seriously?


Well, I wanted to write up a long response, but the people above just about covered it, all of which I agree with.

In summary, in as few words as possible, I would say:
1. Fewer half baked features, and more updates for them.
2. Open file formats for PC3, PMP, CTB, etc.
3. Much less UI tweaking, and quit following Microsoft.

[dgorsman]

We are running 2008. The primary reasons for not upgrading is we are running a number of different third party applications, so if AutoCAD were to be upgraded we would have to install *all* of the other applications as well (provided they are available for the new release). Most of them don't have a silent install/deployment so there is significant time involved. The developers took a bit to catch up with the yearly release cycles. I was primed to migrate us to AutoCAD 2010 - but somebody messed up with the mixed MSSA/network licensing for Raster Design.

If everything worked perfectly, we could upgrade AutoCAD without having to re-install third-party applications, instead running them in "compatibility mode" until they released an update. Of course, those updates would also be a simple plug in, no more "Install the installer, pick the options, point to the correct folder, BUGGER - I need to install .NET 3.5 SP1 first!". Completely unrealistic, but hey - its a nice dream. Updates that you could just "drop in" and "pull out" regardless of user permissions would make pushing updates almost absurdly simple. Anything to make it easier to distribute for CAD admins who aren't IT folk.

Reiterating the open file format mantra, including sheet set data (FYI - being able to use sheet set fields in model space title blocks *would* be a major incentive for us to upgrade).

[kathy.oconnell]

Hi Dgorsman--

Out of curiosity, how many 3rd party applications do you use and if you don't mind listing them would you name some of them? Thanks.

Kathy

[dgorsman]

Hi Dgorsman--

Out of curiosity, how many 3rd party applications do you use and if you don't mind listing them would you name some of them? Thanks.

Kathy

CADWorx Plant Professional
CADWorx Equipment
ProCAD AutoFLOW (used to include both AutoORTHO and AutoISO, as well)
ProSteel
Eagle Point
CloudWorx/Cyclone
Raster Design (not *exactly* third party but it still needs updating)

... as a start. Some people make use of the NavisWorks Navigator plug-in, but most just use the full-blown Review alongside AutoCAD. We've got a mix of PDF drivers on top of those. Pretty certain theres a few user-loaded freebies as well, but officially those fall into the "don't care" category.

*Edit*
That doesn't include our in-house "stuff". Most of it is LISP but I'm getting to the point where .NET is the better solution. That will need rebuilding to new releases periodically as well.

[rkmcswain]

...Eagle Point


You have my sympathies...

[UKITManager]

Simple: Please release Service Patch No.2 for AutoCAD 2010 & LT to fix the slow speed and bugs!

[ccowgill]

Eagle Point

I feel for you, I have the exact same head aches - Eagle Point is the only reason we haven't switched to 64 bit systems.

[cadtag]

... is the main reason you have not upgraded to a newer version? (for example, format change, cost of update version, performance, installing and migrating settings from my current version to the new version, the time it would take to learn the new features, etc)

Thanks!

Kathy O'Connell
Sr. Product Manager--AutoCAD

Kathy, I've thought about this some more, and all of the reasons you mentioned to _not_ upgrade apply, as well as the cost/time of third party software - some of which are no longer available, some are no longer relevant or necessary. But we still have to spend the time to test/evaluate and retrofit.... and non-billable time is scarce. And the cost/time of internal customization -- does it still function as expected under the new environment, is it still needed, or has the core app caught up in that particular functionality?

Perhaps a more useful approach would be to look at the historical list of compelling" upgrades" , and see if that sheds any light. Ignoring subscription customers, which releases had the highest percentages of upgrade sales v.s. new sales? You ought to be able to get that information from your internal beancounters -- I certainly don't have numbers for that. (And ignore subscription seats in the upgrade count -- many subscription upgrades languish on shelves until they are completely coated with dust, leading people to question why they should continue to pay for subscriptions that doesn't get used)

Speaking for myself and the organizations I've worked with, the compelling upgrades were

v2.5 to v2.52, (stupid dongle.....)
v2.6 to R10, (lost the unix boxes post R9)
R10 to R12 (-p network plotting without a license!!!)
R12 to R14 (NT Y2K)
R14 to 2005
R2005 to .....

still waiting on a compelling upgrade post 2005, 2010 is close to threshold, but is only crossing it because of the Map3D features. I'd probably call 2008 - except we have very little need for 3d modeling.

[dgorsman]

You have my sympathies...

We don't use it that much, thankfully - its mostly to grab point lists from surveyors to create a somewhat realistic ground plane for the models and maybe a few cut/fill calculations.
With Eagle Point migrating to Civil3D and away from AutoCAD, I'm putting together a lightweight LISP routine for building terrain from points. If we had a big/overly complex project we could "rent" a Civil3D user and license from our parent company, but for now the in-house development will be cheaper. A lot of our work is in the flatter parts of the world so only involve a couple of hundred data points.

[s_morgan_b]

I'd install an updated Autocad if it was LDT 2011 . . .

[irneb]

2008 ... although we're still on sub. We simply cannot do the upgrade to 2009 or 10 as nearly all our PC's would require a hardware upgrade just to make the new versions usable. We've tried & tested.

So for us: performance is the major reason behind not upgrading. And for me personally, I can't see any benefit of 9 or 10 above 8. Sure there's constraints, but for that type of functionality - Revit is decades ahead!

For all the hard work of pumping AC full of half-done features which don't even look like anything new, I'd have suggested: just fix the long standing issues which have not been touched in 10 years!

And another reason I'm unwilling to even try these so-called "new features", how many forgotten features there? E.g. db connections, this would be a prime suspect for tables (why in hell do you need Excel?) Multi Lines, etc. Annotative Scales seem to already be in this category.

Here's a small example of just how badly ADesk has made ACad a non-performance CAD program: In 1995 I created some very large 3D models in R12 on a 486 with 16MB of RAM. The DWG files went up to 25MB in size. Then in 1998 I did even bigger DWG's in R14 on a Pentium 1 and 64MB RAM. I've recently tried opening these in 2008 on a Core2 with 2GB RAM - it crashed no matter what I tried, the only new ACad I could get them to work with was 2009 on 64bit with 16GB RAM. I could even render these in R12/14, big joke even on the hugely upgraded PC with the newer AC's.