Topic: Using Excel to control AutoCAD

[GegH1]

Thankyou for your active discussion, it has been most helpful.

1. The Client does have AutoCAD LT at the moment but is more comfotable with Excel. I realise LT can't do lsp straight out of the box but there is a product called 'Cadsta max' which may work. If it doesn't...we're stuffed before we start.

2. This product does need to scale. The box represents a shed, the shed will have a pitched roof, and we will have to add a number of windows and doors into each side of the shed. So i think it would be a little too much for .scr.

3. I am a consultant of sorts and have a couple of mill in insurance. At the moment i want to advise on the feasibility of developing the product, whether i could do it or get a third party to do it and either way whether it is worth doing from an economic point of view.

4. I estimate i would have about 140 hours in the budget (my time) to develop such a product. Therefore, if there were some code i could, in the words of TT 'Hack', and adapt into a working model that the client and myself could develop over time, it may be feasible. Otherwise i'm getting the impression it may be a bridge too far and learning to code would have to be something i do in my 'Spare time', not that i have much of that.

I do very much appreciate the discussion though it will probably have saved me getting into something that would take a lot longer than expected to finish, unless i have read it wrong?

[TheMaster]

Thankyou for your active discussion, it has been most helpful.

1. The Client does have AutoCAD LT at the moment but is more comfotable with Excel. I realise LT can't do lsp straight out of the box but there is a product called 'Cadsta max' which may work. If it doesn't...we're stuffed before we start.

2. This product does need to scale. The box represents a shed, the shed will have a pitched roof, and we will have to add a number of windows and doors into each side of the shed. So i think it would be a little too much for .scr.

3. I am a consultant of sorts and have a couple of mill in insurance. At the moment i want to advise on the feasibility of developing the product, whether i could do it or get a third party to do it and either way whether it is worth doing from an economic point of view.

4. I estimate i would have about 140 hours in the budget (my time) to develop such a product. Therefore, if there were some code i could, in the words of TT 'Hack', and adapt into a working model that the client and myself could develop over time, it may be feasible. Otherwise i'm getting the impression it may be a bridge too far and learning to code would have to be something i do in my 'Spare time', not that i have much of that.

I do very much appreciate the discussion though it will probably have saved me getting into something that would take a lot longer than expected to finish, unless i have read it wrong?

FWIF, I "hack" all the time, and the term wasn't being used in a derogatory way.

The difficulty in doing something like this with minimal experience is that while you may be able to hack a working model or prototype together, the problem usually manifests in the form of it not meeting all of the client's expectations that weren't identified before the fact.

If I'm not familiar with the APIs or other components involved in a solution, it's difficult-to-impossible for me to accurately estimate the time required to do something, including the learning curve, assuming it's acceptable for the client to have to bear that additional cost.

I know a few consultants that don't do the work themselves, and subcontract it out to others, and what they usually do, is get an estimate from them, and then double it.

[GegH1]

I like the use of the word Hack it's an honest word

I've fallen into the 'Client Expectation' trap already, this is actually a continuation of a previous post
 
http://www.theswamp.org/index.php?topic=44344.0

The client kind of 'Adjusted' the brief when i went for a progress meeting which is why i'm going to be very carefull about what i promise and a little more black and white about his options.

[hugha]

Cadsta Max may be excellent at what it does but it does not offer any integration with the VBA that you have shown interest in using in that prior post. Nor does it offer any integration with .NET which is, IMHO, the future. Given the complexity, I think you should seek the client's response to aquiring either a full AutoCAD seat or a .NET programmable workalike. Reluctance to invest could be a warning.

140 hours from go to whoa is, again IMHO, very tight. TT's right about client's spec creep biting you hard in the absence of an extremely prescriptive *WRITTEN* specification and I suspect there's some expectation of a fixed price budget which can lead to endless callbacks to fix tiny, tiny detail and is not to be recommended if there's any hint of R&D involved.   If the client is unwilling to be billed on a time & material basis then I'd suggest your offering to quote for a preparatory investigation/report and gauge the reaction to the accompanying Terms & Conditions.

[Micaletti]

I would start with dgorsman's approach. The client doesn't really need to know AutoCAD to open it up, select an excel file, and press OK.  This approach looks expensive and much more time consuming.

[TheMaster]

I would start with dgorsman's approach. The client doesn't really need to know AutoCAD to open it up, select an excel file, and press OK.  This approach looks expensive and much more time consuming.

That approach isn't even an option. It only provides the means to read/write DWG files, it doesn't display them, and can't be used from Excel.

[GegH1]

Hugh,
Thankyou for the info regarding Cadsta, this kinda makes the decision process simple, 'Are you going to buy full CAD?', and i kinda figured the 140 hours was a bit tight especially without knowing the programming language inside out. Looks like the project may be a labour of love for me in my 'spare time'.

I'll throw this one out there. Take AutoCAD out of the equation, we have a form in Excel with the dimensions of the shed, number, dimensions and position of the doors and windows, is there another way of creating a dynamic illustration of the shed based on the figures on the excel form?

[TheMaster]

Hugh,
Thankyou for the info regarding Cadsta, this kinda makes the decision process simple, 'Are you going to buy full CAD?', and i kinda figured the 140 hours was a bit tight especially without knowing the programming language inside out. Looks like the project may be a labour of love for me in my 'spare time'.

I'll throw this one out there. Take AutoCAD out of the equation, we have a form in Excel with the dimensions of the shed, number, dimensions and position of the doors and windows, is there another way of creating a dynamic illustration of the shed based on the figures on the excel form?

Sure is. 

For starters, WPF is a very capable 2D/3D graphics kernel (most think it's about UIs but it can be used to generate any kind of 2D and 3D vector graphics as well) see http://msdn.microsoft.com/en-us/library/ms747393.aspx).

You can also use any number of third-party components, from simple libraries that allow you display simple 2D vector graphics (http://en.wikipedia.org/wiki/Gtk_Sharp), to full-blown embedded 2D and 3D CAD and modeling functionality (http://www.devdept.com).

[hugha]

Much depends on what the client expects to see at the push of a button. 

If the field's wide open on what to work in Sketchup might offer some fun:
http://www.youtube.com/watch?v=7Kcs6Ruspro

Sketchup uses Ruby scripting for automation.
http://www.sketchup.com/intl/en/developer/

If I had the time, I'd be playing with this myself

But it's a crowded market. Make or buy is always a tough decision in the design world but you can get an idea of what's out there by searching "Shed design software" and it may be that you could find a suitable system to set up to work with your client's Excel - for a suitable fee..

[Micaletti]

That approach isn't even an option. It only provides the means to read/write DWG files, it doesn't display them, and can't be used from Excel.

I know it doesn't display them however it most certainly is an option given the client "doesn't have autocad" and is "only familiar with excel". An invisible or visible stand alone app can use both APIs. At first glance devDepot looks like a better option in that situation if it can be used from excel alone, but I would avoid either if at all possible. It would be fairly straight forward to consume excel data from AutoCAD at the press of a ribbon button.

I do like the DWG idea better. It's always nice to have a DWG of a product, especially if you can automate it's production.

[TheMaster]

That approach isn't even an option. It only provides the means to read/write DWG files, it doesn't display them, and can't be used from Excel.

I know it doesn't display them however it most certainly is an option given the client "doesn't have autocad" and is "only familiar with excel". An invisible or visible stand alone app can use both APIs. At first glance devDepot looks like a better option in that situation if it can be used from excel alone, but I would avoid either if at all possible. It would be fairly straight forward to consume excel data from AutoCAD at the press of a ribbon button.

I do like the DWG idea better. It's always nice to have a DWG of a product, especially if you can automate it's production.

It's pretty clear from the OPs description that he wants to see the graphical result as values are being changed in the Excel spreadsheet, as opposed to spitting out a .DWG file that he must then open and view.

For that reason, RealDWG is not an option, because it can't display the graphical output while the user is changing values in the spreadsheet (or any other UI), so I have to beg to differ.

[WILL HATCH]

Regarding Autocad OEM:

The licensing fee structure is fairly simple and straightforward making it easier for you to be flexible and competitive in your market. The fees consist of $5,000 for licensing the OEM (this is per upgrade) and 25% royalty fee calculated from your total OEM based revenue.

[GegH1]

TT, have you had any eperience with (http://www.devdept.com).
I have this right don't i, this is building a stand alone .net CAD app custom made to produce whatever you want, as in one of the videos, a kitchen, in my case a shed. and the components could be developed in AutoCAD and imported into the application.

I'm assuming it is complicated to develop? beyond me anyway? or do the develop it for you?

Sorry for sounding like an idiot, i'm not quite sure what the product is, but i'm very interested in the possibilities

[WILL HATCH]

I have this right don't i, this is building a stand alone .net CAD app custom made to produce whatever you want, as in one of the videos, a kitchen, in my case a shed. and the components could be developed in AutoCAD and imported into the application.

Looks like that will work for you.

I'm assuming it is complicated to develop? beyond me anyway? or do the develop it for you?

You'll be on  your own developing for their app, from a brief look at their forum there isn't a lot of support as it's a pretty new product.  Considering that you'd still need to spend 900 euros to get surface support it's not too cheap

[TheMaster]

TT, have you had any eperience with (http://www.devdept.com).
I have this right don't i, this is building a stand alone .net CAD app custom made to produce whatever you want, as in one of the videos, a kitchen, in my case a shed. and the components could be developed in AutoCAD and imported into the application.

I'm assuming it is complicated to develop? beyond me anyway? or do the develop it for you?

Sorry for sounding like an idiot, i'm not quite sure what the product is, but i'm very interested in the possibilities

In general, because it is a component, it generally is more difficult to use for building what amounts to a very-customized CADD application. So, many of the things that AutoCAD's API makes easy for you, may not be there in the same form, and would most-likely require more work.

In your case, it might be overkill since you're not really trying to enable a full interactive CADD system, you're just trying to automate creation of a drawing from user-supplied data that doesn't require them to do any kind of drawing.