TheSwamp
Code Red => AutoLISP (Vanilla / Visual) => Topic started by: JohnK on April 14, 2011, 09:27:29 PM
-
EDIT Log:
- 04.15.11 - Crossed off `pseudo code' & `svn docs'
- 04.19.11 - Crossed off Soldier's documentation template
TODO:
Lieutenant:
Please create SVN details (brief docs) and recommendations for interfaces. - (dgorsman)
Lieutenant:
Finalize and combine directory structures from soldiers and assassins (add in required folders for final package).
Soldier:
Determine PSEUDO CODE - (pBe)
Soldier:
Devise error procedures (returns for procedures, error procedure call, error messages, etc.)
Soldier:
Create a documentation template file which can be used by other coders and which can be delivered to Assassins.
Devise a folder structure for code. - (Se7en)
Assassin:
Devise a folder structure for docs.
Assassin:
Create a final documentation template for all types of documentation (from
code and project procedures)
-
PSEUDO CODE
This routine will allow the end user to select an item and change the item(s) selected to a layer of the end users choice
Set up Selection Method
Ensure PICKFIRST is set to 1
if not 1
Include to initialize function
Enable ":L" filter for before/after tool is called
if Selection nil ("after" tool is called)
if Select
Toggle button for previous selection
if No previous set
Disable button (mode_tile 1)
else
Default to is_enabled
else ("before" tool is called)
Toggle button for option to Unlock based on two points obtain thru ssget before proceeding to add new objects to selection set
if Enabled
initiate 7WIP:ReselectPreviousLocked subfunction
Initialize Dialog Box Destination layer
Initiate 7WIP:LayerListCleanup
Set Current layer as Default value, include Filter option for selection of Destination Layer to minimize list and exclusion of Xref layers
If listed
Select Destination Layer
if mispelled/Incorrect Naming convention
Prompt to rename
If "accept"
Accept all values and initiate 7WIP:ChangeEntityLayer subfunction
if Rename Flag T
Initiate 7WIP:RenameLayer subfunction
Initiate 7WIP:ResetGlobalVar
Initiate 7WIP:StoreLayerName
else "cancel"
Initiate Error Handler to include 7WIP:ResetGlobalVar
-
Thanx pBe. possible idea: generate and clean the layer list before the dialog is displayed.
-
We could call the project: "Project7"
Sounds cool.
-
I'll see about getting SVN instructions up in the next couple of days.
-
This thread is intended to be a place for the general poulation to offer praise, money, first born, etc., etc. (read: for informational purposes).
Moved discussion to correct thread :-)
see if you cant work up some sort of writeup explaining how your dialog is to be called, what does it return, etc., etc..
This was the header I used for the module:
;; NAME
;; 7WIP:GetLayerSelection
;;
;; DESCRIPTION
;; Provides a dialog interface through which
;; the user can select a layer from a list of supplied
;; layer names displayed in a list_box.
;;
;; Dialog includes an edit_box filter to enable the user to filter
;; the list of layer names using a wildcard string.
;;
;; ARGUMENTS
;; List of Strings to be displayed in the list_box
;;
;; EXAMPLE
;; (7WIP:GetLayerSelection '("Layer1" "Layer2" "Layer3"))
;;
;; RETURNS
;; Name of selected layer if the user presses OK, else
;; returns nil if either the user presses cancel or the dialog
;; could not be loaded.
;;
;; BY
;; Lee Mac
;;
;; DATE
;; 15.04.2011
;;
-
We could call the project: "Project7"
Sounds cool.
Well I like it but I may be a little biased. *lol*
Speaking of titles/names: Who can draw? Can we get a simple logo for the docs (or is that a bit too much)?
-
I would supposed I should probably just sit tight until documentation editing is required, and I'm hoping when we get to that point, I can get a little more clarification on what deliverables are being expected of me. Proof reading/writing I can do, I as for folder structure for documentation, It all depends on the intent of how things are to be packaged. Are we thinking an actual help file/readme type thing, or am I totally missing the mark. It also may be easier to develop a template after the first go around, then we can make slight modifications to it if need be for future projects
-
Chris,
No you are on the mark. I would most certianly count on some proof reading/writing and other editiorial duties.
What about docs in at least HTML and TXT formats? My thoughts are: HTML, a browser window can be launched and/or for TXT, a dialog could be displayed from with-in the program.
-
We could call the project: "Project7"
Sounds cool.
Well I like it but I may be a little biased. *lol*
Speaking of titles/names: Who can draw? Can we get a simple logo for the docs (or is that a bit too much)?
Perhaps?
(http://www.theswamp.org/screens/leemac/Project7.png)
-
Who can draw?
I draw with AutoCAD everyday. :-) but if you want something by hand, sorry I cant help much there. What type of logo are you looking for? I may be able to come up with something
-
<snip> It also may be easier to develop a template after the first go around, then we can make slight modifications to it if need be for future projects
I agree with that, but I was just thinking a simple template with any headers, pharases, logo's, whatever so all the Assassin's are using the same format document for now (we dont really have any of that right now so where do we begin?).
-
Chris,
No you are on the mark. I would most certianly count on some proof reading/writing and other editiorial duties.
What about docs in at least HTML and TXT formats? My thoughts are: HTML, a browser window can be launched and/or for TXT, a dialog could be displayed from with-in the program.
I've still got some old programs, when using html to be able to convert to chm if you wanted. HTML is doable, just need to create a template for the pages, then add content. Our office customizations help file is html based, just having some issues with formatting on Firefox.
-
I like it Lee! Can we get some border around the VII part ...make it more ominous looking.
-
Chris,
No you are on the mark. I would most certianly count on some proof reading/writing and other editiorial duties.
What about docs in at least HTML and TXT formats? My thoughts are: HTML, a browser window can be launched and/or for TXT, a dialog could be displayed from with-in the program.
I've still got some old programs, when using html to be able to convert to chm if you wanted. HTML is doable, just need to create a template for the pages, then add content. Our office customizations help file is html based, just having some issues with formatting on Firefox.
Cool. Let's try the CHM when we get some docs going and see how it works (options are always good).
-
I like it Lee! Can we get some border around the VII part ...make it more ominous looking.
How about a touch of shadow/contour - check the updated above :-)
-
HTML for the documentation is good; as text it can be change-managed through the SVN. Other files like CHM can also be managed but they would have to be checked in/out.
-
Toggle button for option to Unlock based on two points obtain thru ssget before proceeding to add new objects to selection set
if Enabled
initiate 7WIP:ReselectPreviousLocked subfunction
Everyone concur with this?
-
I'm not sure i understand that section pBe. I was re-reading the PSEUDO code and that part confuses me. Does that mean that you are suggesting that we retain the previous selection(s)?
If you are referring to an option to unlock a layer; the locked, frozen, off, and xref layers will not be displayed in the dialog.
-
Here is how I might approach the pseudo code to tie in with my submission here (http://www.theswamp.org/index.php?topic=37930.msg429584#msg429584):
;;--------------------={ Project7 }=--------------------;;
;; Change Item(s) to Layer - PSEUDO CODE ;;
;;------------------------------------------------------;;
+-----> Cycle through Layer Table (tblnext) and construct a
| list of layer names, excluding XRef layers (not (wcmatch "*|*")).
|
|
+-----> Prompt for Selection of Objects (ssget), use "_:L" mode string
| to exclude objects on locked layers.
|
|
+--+--> IF non-nil SelectionSet
| |
| |
| |
| +--+ THEN:
| | |
| | |
| | |
| | +--> Call 7WIP:GetLayerSelection with list of layer names
| | | [sorted in Alphabetical order?]
| | |
| | |
| | +--> IF 7WIP:GetLayerSelection returns valid layer selection
| | |
| | |
| | |
| | +--+ THEN:
| | | |
| | | |
| | | |
| | | +--> Iterate through SelectionSet and call
| | | 7WIP:ChangeEntityLayer to change the layer of each
| | | entity to the chosen layer.
| | |
| | +--+ ELSE:
| | |
| | |
| | |
| | +--> Princ *Cancel* as User pressed Dialog Cancel button
| |
| |
| |
| +--+ ELSE:
| |
| |
| |
| +--> Princ *Cancel* as User hasn't selected anything
|
|
|
+--> Exit Cleanly
Or, alternatively (my preference), use a single COND instead of nested IF statements:
;;--------------------={ Project7 }=--------------------;;
;; Change Item(s) to Layer - PSEUDO CODE ;;
;;------------------------------------------------------;;
+-----> Cycle through Layer Table (tblnext) and construct a
| list of layer names, excluding XRef layers (not (wcmatch "*|*")).
|
|
+-----> Prompt for Selection of Objects (ssget), use "_:L" mode string
| to exclude objects on locked layers.
|
|
+--+--> Start COND
| |
| |
| |
| +--+ Condition 1: (null SelectionSet)
| | |
| | |
| | |
| | +--> Princ *Cancel* as User hasn't selected anything
| |
| |
| |
| +--+ Condition 2: Call 7WIP:GetLayerSelection with list of layer names
| | | [sorted in Alphabetical order?]
| | | (null 7WIP:GetLayerSelection)
| | |
| | +--> Princ *Cancel* as User pressed Dialog Cancel button
| |
| |
| |
| +--+ Default Condition => Valid Selection & Selected Layer
| |
| |
| |
| +--> Iterate through SelectionSet and call
| 7WIP:ChangeEntityLayer to change the layer of each
| entity to the chosen layer.
|
|
|
+--> Exit Cleanly
IMO, the COND structure is a lot cleaner, not to mention more readable since there is no nesting.
Lee
-
I'm not sure i understand that section pBe. I was re-reading the PSEUDO code and that part confuses me. Does that mean that you are suggesting that we retain the previous selection(s)?
If you are referring to an option to unlock a layer; the locked, frozen, off, and xref layers will not be displayed in the dialog.
oh i see, i was not referring to the destination layer but the selected objects layer ,reason why included that option if by somehow the user invoke the tool before selection and realize afterwards the object is locked while in selection mode, a toggle button to unlock would be handy instead of exiting the tool and unlocking the layer outside the tool. or perhaps a better selection method should be use? forcing the objects layer to unlock when selected?
-
Or, alternatively (my preference), use a single COND instead of nested IF statements:
IMO, the COND structure is a lot cleaner, not to mention more readable since there is no nesting.
Lee
I concur COND is indeed cleaner
-
Attached is a quickie outline for submitting work via SVN. The document is stored in the repository (upload is from my own working copy, in fact 8-) ).
If the coders aren't comfortable with using the SVN interface for the first bit, I can do some check in/out on their behalf but it will be batched once or twice a week. Once the project gets rolling SVN use will be mandatory to avoid coding conflicts.
-
Awesome dgorsman.
I had some documentation for using Vim, and Tortoise SVN. ...ah, here it is.
[ http://www.theswamp.org/index.php?topic=30237.0 ]
-
Sorry if the format of this file is wrong but I did this on my PDA (as well as this post). I will redo it when I get on a pc if its all goofy.
Here is a documentation template for the Soldiers.
-
Is there agreement on the Psuedo code as previously posted?
Should I send my code snippet to dgorsman to be added to this SVN thing?
-
what kind of help format are we looking for here, do we want something that looks like it is part of AutoCAD (mimics the basic color scheme and formating of AutoCAD's help) or come up with our own style?
-
Is there agreement on the Psuedo code as previously posted?
Should I send my code snippet to dgorsman to be added to this SVN thing?
I would recomend that you try SVN yourself. But i dont know where dgorsman stands; i need to download SVN here so i can check out whats there (hang tight a little longer).
I didnt really sense a concensis on the pseudo code issue(s) but i will chime in and get that ball rolling again today.
-
what kind of help format are we looking for here, do we want something that looks like it is part of AutoCAD (mimics the basic color scheme and formating of AutoCAD's help) or come up with our own style?
We can do our own; that would be easier and the help docs for AutoCAD are not the best format IMO.
-
Here is how I might approach the pseudo code to tie in with my submission here (http://www.theswamp.org/index.php?topic=37930.msg429584#msg429584):
;;--------------------={ Project7 }=--------------------;;
;; Change Item(s) to Layer - PSEUDO CODE ;;
;;------------------------------------------------------;;
+-----> Cycle through Layer Table (tblnext) and construct a
| list of layer names, excluding XRef layers (not (wcmatch "*|*")).
|
|
+-----> Prompt for Selection of Objects (ssget), use "_:L" mode string
| to exclude objects on locked layers.
|
|
+--+--> Start COND
| |
...
The above pseudo code has my vote but i want to ask if there was a reason why the `layer table iterating' is done so early?
Would this work instead?
;;--------------------={ Project7 }=--------------------;;
;; Change Item(s) to Layer - PSEUDO CODE ;;
;;------------------------------------------------------;;
+-----> Prompt for Selection of Objects (ssget), use "_:L" mode string
| to exclude objects on locked layers.
|
+--+--> Start COND
| |
| |
| +--+ Condition 1: (null SelectionSet)
| | |
| | +--> Princ *Cancel* as User hasn't selected anything
| |
| |
| +--+ Condition 2: Call 7WIP:GetLayerSelection with list of layer names
| | | [sorted in Alphabetical order?] [color=red]<---- sure[/color]
| | | (null 7WIP:GetLayerSelection) [color=red]<---- what is this?[/color]
| | |
| | +--> Princ *Cancel* as User pressed Dialog Cancel button
| |
| |
| +--+ Default Condition => Valid Selection & Selected Layer
| |
| +-----+ Cycle through Layer Table (tblnext) and construct a
| | list of layer names, excluding XRef layers
| | (not (wcmatch "*|*")) and layers that are off, frozen,
| | and locked.
| |
| +--> Iterate through SelectionSet and call
| 7WIP:ChangeEntityLayer to change the layer of each
| entity to the chosen layer.
|
+--> Exit Cleanly
-
I would tend to think the second option would be better, as there is no point in cycling through the layers if the command is just going to get canceled out. My question would be, instead of just exiting if there is no selection set, should you first give the user the option to select a selection, and if it is still nil, exit out. Or is that too many extra prompts and not streamlined enough?
-
The layer table needs to be iterated before '7WIP:GetLayerSelection' is called, as this function requires a list of Layer names to display in the dialog. It could be iterated after a valid selection is made I suppose, although I thought the code would look cleaner/more readable with the construction of the list of layers outside of the COND statement.
As for :
(null 7WIP:GetLayerSelection) <---- what is this?
This is the test condition for the second COND statement, i.e. if (null 7WIP:GetLayerSelection) returns T, princ *Cancel*.
Lee
-
My question would be, instead of just exiting if there is no selection set, should you first give the user the option to select a selection, and if it is still nil, exit out. Or is that too many extra prompts and not streamlined enough?
Note that the user is prompted for a selection, then this selection is tested - this would allow both an implied selection and, if no implied selection is present, will prompt the user.
-
Ok, makes sense. I'm having some issues with how I'm seeing posts in the thread, I think its just a setting on my end.
-
Ah, gotcha.
Okay, I moved the layer iterating below object selection and this looks good to me if you guys agree too?
;;--------------------={ Project7 }=--------------------;;
;; Change Item(s) to Layer - PSEUDO CODE ;;
;;------------------------------------------------------;;
+-----> Prompt for Selection of Objects (ssget), use "_:L" mode string
| to exclude objects on locked layers.
|
+-----> Cycle through Layer Table (tblnext) and construct a
| list of layer names, excluding XRef layers
| (not (wcmatch "*|*")) and layers that are off, frozen,
| and locked.
|
+--+--> Start COND
| |
| |
| +--+ Condition 1: (null SelectionSet)
| | |
| | +--> Princ *Cancel* as User hasn't selected anything
| |
| |
| +--+ Condition 2: Call 7WIP:GetLayerSelection with list of layer names
| | | [sorted in Alphabetical order?]
| | | (null 7WIP:GetLayerSelection)
| | |
| | +--> Princ *Cancel* as User pressed Dialog Cancel button
| |
| |
| +--+ Default Condition => Valid Selection & Selected Layer
| |
| +--> Iterate through SelectionSet and call
| 7WIP:ChangeEntityLayer to change the layer of each
| entity to the chosen layer.
|
+--> Exit Cleanly
-
That would be no different to my original though, since the validity of the SelectionSet is tested after the Table is iterated.
EDIT: [After more thinking] I suppose it would be more efficient for those cases in which the user hits Esc at the Selection prompt - Agree.
-
sounds good to me.
-
great. Pseudo code is done.
Again, this group thing is fun (slow, but fun).
-
Wouldn't it make more sense to:
- Check validity of selection as it's executed (eg. if or cond)
- create layer list and feed it to dialog selection
- iterate through selection set and put objects on defined layer
(cond ((not (setq selectionset (ssget "_:L"))))
((not (setq targetlayer (7WIP:GetLayerSelection (7WIP:GetLayerList)))))
(T step through selectionset and put entities on defined layer)
)
OR
(if (and (setq selectionset (ssget "_:L"))
(setq targetlayer (7WIP:GetLayerSelection (7WIP:GetLayerList)))
)
(progn step through selectionset and put entities on defined layer)
)
or am I missing something?
-
Is there agreement on the Psuedo code as previously posted?
Should I send my code snippet to dgorsman to be added to this SVN thing?
I would recomend that you try SVN yourself. But i dont know where dgorsman stands; i need to download SVN here so i can check out whats there (hang tight a little longer).
I'd prefer coders get used to using the SVN, if only for their own education. Future projects may be much larger and require the diff-patch method rather than a simple check in/out. I think the documention/assassins should be online as well so the documentation is automatically distributed.
-
I agree about being on line, if the help documents arnt all there, you wont be able to see images, hyperlinks wont work, etc..
little confused about the SVN thing, I think I understand the procedure file, but I'm getting a page error when I go to tortisesvn.tigris.org
Se7en, When you refer to the AutoCAD help not being structured well, are you referring to the later versions, or some of the earlier ones, or both?
-
We could call the project: "Project7"
Sounds cool.
Well I like it but I may be a little biased. *lol*
Speaking of titles/names: Who can draw? Can we get a simple logo for the docs (or is that a bit too much)?
Perhaps?
(http://www.theswamp.org/screens/leemac/Project7.png)
Lee,
Can I get this png with white text/lighter text, for darker backgrounds? that way we'll have one of each.
Thanks,
-
We could call the project: "Project7"
Sounds cool.
Well I like it but I may be a little biased. *lol*
Speaking of titles/names: Who can draw? Can we get a simple logo for the docs (or is that a bit too much)?
Perhaps?
(http://www.theswamp.org/screens/leemac/Project7.png)
Lee,
Can I get this png with white text, for darker backgrounds? that way we'll have one of each.
Thanks,
I'd either set the 'Project' text to be a little more bold (washes out in some areas overlaying the "VII") and if it were me, I'd go as far to say that a "7" would look cooler. What if we follow take from hacker text and use Projec7? Although, most probably wouldn't get it.
-
the logo doesnt matter much to me. Now that you mention it, it does look like it washes over the red text. I was just thinking if we use a dark background for our help file header area, a light colored logo would be nice, and if we have any white space areas that contain the logo, the dark logo would go there.
-
We could call the project: "Project7"
Sounds cool.
Well I like it but I may be a little biased. *lol*
Speaking of titles/names: Who can draw? Can we get a simple logo for the docs (or is that a bit too much)?
Perhaps?
(http://www.theswamp.org/screens/leemac/Project7.png)
Lee,
Can I get this png with white text/lighter text, for darker backgrounds? that way we'll have one of each.
Thanks,
Sure Chris:
(http://www.theswamp.org/screens/leemac/Project7w.png)
-
I especially like the one on the left...
Fonts are ugly, but this was quickly thrown together before I leave to go get my kid.
-
I agree about being on line, if the help documents arnt all there, you wont be able to see images, hyperlinks wont work, etc..
little confused about the SVN thing, I think I understand the procedure file, but I'm getting a page error when I go to tortisesvn.tigris.org
Se7en, When you refer to the AutoCAD help not being structured well, are you referring to the later versions, or some of the earlier ones, or both?
Whoopsie, that should be tortoisesvn.tigris.org. Updated.
-
Here is my 5 minute attempt.
I was going for "hand written".
EDIT: But i dont like the "7" so...
-
*shrug* tweaked mine.
-
Am I too assume I'm being ignored or was my post (http://www.theswamp.org/index.php?topic=37930.msg430185#msg430185) just lost in logo talk?
-
I'm sorry I thought we were brainstorming.
I don't really mind any of them...I just tossed something together to help spur ideas (i thought that's what we were doing).
-
I'm sorry I thought we were brainstorming.
I don't really mind any of them...I just tossed something together to help spur ideas (i thought that's what we were doing).
Oh no, not the logo, my post about the code structure (I linked it in the text "my post"). Sorry, should have been more direct.
http://www.theswamp.org/index.php?topic=37930.msg430185#msg430185
-
I apologize Chris; I totally missed your question about the old or new help docs...I didn't mean to.
here's how I see it; you have more to say than I do about the docs because that is an area of development that I'm not that good at. however when I think of development documentation I think of the BSD developers. go to NetBSD.org and look for the manual. it is HUGE. I've heard that they spend 2 hrs on docs for every hr of coding.
we can toss around ideas about any style you like.
EDIT:
Here is a direct link to the NetBSD Guide (NetBSD is an operating system if you didnt know).
[ http://netbsd.org/docs/guide/en/netbsd.html ]
-
Wouldn't it make more sense to:
- Check validity of selection as it's executed (eg. if or cond)
- create layer list and feed it to dialog selection
- iterate through selection set and put objects on defined layer
... <snip>...
Yeah, that does make more sense. That would make error trapping a bit easier (not that it was going to be difficult to begin with). -e.g. all sub functions return a something or nil (okay, that example-error-statement sounded way better in my head --I need sleep-).
-
Hey Lee, are you gonna use subs for your DCL events? Meaning you call a sub that gets a list of Layers in the Active Drawing, not include that code in the DCL control code? Does that make sense? If so then I can easily create an OpenDCL option for the project.
Let me know, thanks!
jb
-
Hey Lee, are you gonna use subs for your DCL events? Meaning you call a sub that gets a list of Layers in the Active Drawing, not include that code in the DCL control code? Does that make sense? If so then I can easily create an OpenDCL option for the project.
I've never used ODCL before, but I understand what your saying and yes, I've structured the GetLayerSelection sub to accept a list of layer names and return the selected one - hence all is generic and nothing is hard-coded.
:-)
-
Hey Lee, are you gonna use subs for your DCL events? Meaning you call a sub that gets a list of Layers in the Active Drawing, not include that code in the DCL control code? Does that make sense? If so then I can easily create an OpenDCL option for the project.
Let me know, thanks!
jb
Cool idea!
-
OK, cool. I'll have a look back at the entire discussion(s) to see you guys are structuring this (SVN?)
I'm thinking a very minimal modeless form with no title bar, displayed at the cursor location, and utilizes a MouseMovedOff event to close.
When I get something working I'll post an example . . ..
-
OK, cool. I'll have a look back at the entire discussion(s) to see you guys are structuring this (SVN?)
I'm thinking a very minimal modeless form with no title bar, displayed at the cursor location, and utilizes a MouseMovedOff event to close.
When I get something working I'll post an example . . ..
Fantastic idea. Much easier to have it by the cursor.
(http://www.theswamp.org/screens/jbuzbee666.png) :evil:
-
Oh crap!!
-
OK, cool. I'll have a look back at the entire discussion(s) to see you guys are structuring this (SVN?)
I'm thinking a very minimal modeless form with no title bar, displayed at the cursor location, and utilizes a MouseMovedOff event to close.
When I get something working I'll post an example . . ..
Once things get spinning, let me know - I'll see about setting up an ODCL branch on the repository.
-
just throwing out an idea here, so dont shoot me. I'm not that familiar with ODCL (alhtough I have downloaded it and intend to learn it at some point) would it be possible to have a check that if the user's machine contains the runtime, it would use the ODCL box, and if not use the standard DCL box? Or is that just pointless? I assume that ODCL offers more abilities than standard DCL.
-
I apologize Chris; I totally missed your question about the old or new help docs...I didn't mean to.
here's how I see it; you have more to say than I do about the docs because that is an area of development that I'm not that good at. however when I think of development documentation I think of the BSD developers. go to NetBSD.org and look for the manual. it is HUGE. I've heard that they spend 2 hrs on docs for every hr of coding.
we can toss around ideas about any style you like.
EDIT:
Here is a direct link to the NetBSD Guide (NetBSD is an operating system if you didnt know).
[ http://netbsd.org/docs/guide/en/netbsd.html ]
Ok, I got it, I was thinking something fancier, but I see how the formatting is. I can use the help for Virtual Box as a base, as they also have theirs in CHM format, and follow this same style. either way, it doesnt really matter, as long as we aren't changing it after we get going. It is a real pain to go back and recode stuff so it all matches.
-
just throwing out an idea here, so dont shoot me. I'm not that familiar with ODCL (alhtough I have downloaded it and intend to learn it at some point) would it be possible to have a check that if the user's machine contains the runtime, it would use the ODCL box, and if not use the standard DCL box? Or is that just pointless? I assume that ODCL offers more abilities than standard DCL.
These echo my thoughts, since I also don't use ODCL :wink:
Or even, just have two programs, one using DCL, the other ODCL - you would only need to swap the DCL modules.
-
just throwing out an idea here, so dont shoot me. I'm not that familiar with ODCL (alhtough I have downloaded it and intend to learn it at some point) would it be possible to have a check that if the user's machine contains the runtime, it would use the ODCL box, and if not use the standard DCL box? Or is that just pointless? I assume that ODCL offers more abilities than standard DCL.
These echo my thoughts, since I also don't use ODCL :wink:
Or even, just have two programs, one using DCL, the other ODCL - you would only need to swap the DCL modules.
Thats one of the features of SVN. You can maintain different development branches, with the capability of merging them back into the trunk later. As an example, I used to actively contribute the MegaMek project over on SourceForge. They have a "stable" branch and a "dev" branch; "stable" is stable but somewhat out-of-date, while the "dev" branch has a lot of different updates applied from contributors which may or may not make it to the next stable release.
This is why I put the UI as a separate module here. That way it can easily be "branched" to provide an ODCL branch, or even a managed .NET DLL branch, in addition to the conventional DCL in the trunk.
-
Wouldn't it make more sense to:
- Check validity of selection as it's executed (eg. if or cond)
- create layer list and feed it to dialog selection
- iterate through selection set and put objects on defined layer
... <snip>...
Yeah, that does make more sense. That would make error trapping a bit easier (not that it was going to be difficult to begin with). ...<snip>...
What are the opinions on alanjt's recommendation?
-
Wouldn't it make more sense to:
- Check validity of selection as it's executed (eg. if or cond)
- create layer list and feed it to dialog selection
- iterate through selection set and put objects on defined layer
... <snip>...
Yeah, that does make more sense. That would make error trapping a bit easier (not that it was going to be difficult to begin with). ...<snip>...
What are the opinions on alanjt's recommendation?
I'm He's a genius! :-P
-
Yeah `genius'. Although, you may not be able to fit through doorways with that giant head...then you would be stuck at work. Yeah, `genius'.
:-P
-
Wouldn't it make more sense to:
- Check validity of selection as it's executed (eg. if or cond)
- create layer list and feed it to dialog selection
- iterate through selection set and put objects on defined layer
... <snip>...
Yeah, that does make more sense. That would make error trapping a bit easier (not that it was going to be difficult to begin with). ...<snip>...
What are the opinions on alanjt's recommendation?
I'm game
-
Yeah `genius'. Although, you may not be able to fit through doorways with that giant head...then you would be stuck at work. Yeah, `genius'.
:-P
LoL, very true.
-
OK, I've had some time and have come up with three options:
Option 1: Typical Modal Dialog Box
The user would select a layer to transfer the Object to, the Dialog would close with the user proceeding to next select the object(s).
(http://www.blackhorsedev.us//jb/modal.jpg)
Option 2: Modeless Dialog Box
This Dialog Box would remain "Open" allowing the user to run the command on multiple objects. Select a layer in the list box and then picking the "Select Object" button the focus would be shifted to the drawing space to, well, select an object(s).
(http://www.blackhorsedev.us//jb/modeless.jpg)
Option 3: Popup Modeless Form
This very simple form would "pop-up" at the cursor location. This would make a good toolbar button. Once double-clicking on a layer, the form hides, and the focus is back to the drawing to select an object(s). If the user decides to cancel the command, once the mouse leaves the form it hides.
(http://www.blackhorsedev.us//jb/layerpop.jpg)
All forms allow resizing and the last layer selected would be saved to the blackboard to be selected the next time the form is shown. The modal form is the easiest to impliment with something like (setq layerName(KWIP:SelectLayerForm)). The modeless forms will take a little bit more coding, and may need to be the main calling function (command). I'm experimenting with that now.
The Help System: if you guys want to go HTML (that's what I use) I can provide a simple modeless OpenDCL dialog that displays HTML pages.
jb
-
It depends on what you use to create them. If standard dcl can only handle modal dialogs, then that is what those users would use. If you have opendcl, I would assume option 2 and 3 were created using it, so users could choose either one. Otherwise, I'm partial to option 2
-
Oh yeah. I forgot about that - my bad Lee, sorry. OK, so ignore #1.
-
I like option #3 the best, but it would need to display at the cursor instead of at the toolbar or it would be no better than selecting objects and changing their layer with the layer pulldown.
-
That's the way it works: if you type in "LL" (the shortcut for the really long command name) the form displays at the cursor.
-
That's the way it works: if you type in "LL" (the shortcut for the really long command name) the form displays at the cursor.
Cool
-
*Thread bump*
In response to the above discussion: we will be using theSwamp SVN server for our code so managing or having several different "versions" of code will not be a problem. -i.e. having a stdDCL or an ODCL "version" will not be a problem.
jbuzbee,
you are more then welcome to join the team. You can develop a ODCL version (of course, you don't have to either; you can just "fork" ours.) However, I apologize if I missed your "sign up", I didnt mean to leave you out of the team at all. Let me know your thoughts or plans.
Chris,
How is the template for the docs coming along? What are your thoughts? Are you going to use MS Word and convert to HTML. Or do you plan on using an HTML editor; Dreamweaver, Frontpage, or whatever (I dont know if either of those exist anymore, I just tossed out those product names because those are the only two I remember)?
dgorsman,
How does our repository look? Are we good to go? Can I `check out' yet?
-
Chris,
How is the template for the docs coming along? What are your thoughts? Are you going to use MS Word and convert to HTML. Or do you plan on using an HTML editor; Dreamweaver, Frontpage, or whatever (I dont know if either of those exist anymore, I just tossed out those product names because those are the only two I remember)?
To be honest, I havent really gotten that far on it, I've been a little swamped. I'm going to use a simple tree format, similar to the example you provided. From my stand point there isnt too much of a point of going overboard on it. For the html code, I plan on using Notepad ++. I find it is easier to just code straight, than to have another program do it. It gives me more control over how the code is structured, and easier to keep it modular IMO.
-
yeah me too (Ive been swamped too, that is).
> notepad++
cool! Let me know if you can use any tools (maybe if i can write a small helper app for ya i will do what i can).
-
Repository should be good to go. Just a couple of reminders to all contributors:
- Tortoise SVN client download for those that need it: http://tortoisesvn.tigris.org/ (http://tortoisesvn.tigris.org/)
- Check and update your local copies on a regular basis
- Each function should be in its own file, in one of the sub-folders under "trunk" e.g. error handling functions go under ...\trunk\Error handler
- Develop, test, *then* commit a working copy to the repository
- The repository URL is https://www.theswamp.org/svn/public/7WIP
Chris, I've put in a "Project Docs" folder for documentation; add sub-folders there as necessary. HTML files (as text, yes?) can be posted to the repository the same way as the LISP files; image and other non-text files can also be added but will need to be checked in/out which may require some coordination. As you work through the documentation, commit operations will distribute your latest works to everybodies working copies and as people commit their changes you will also have access to the current program configuration.
Se7en, I've added a "Job jar" directory under the repository root. When you need specific tasks to be done, put them into a file there with along with any notes, requirements, etc. Might be useful to have a separate file to list all the jobs so the status can be tracked, or it can be done in the job file itself. Or, reconfigure to suit your needs.
-
Thats great.
-
Ive just checkout my copy. Very nice dgorsman.
BTW, the Job Jar is a really cool idea.
> Status
I was just thinking about how we know if someone has started a job and or has claimed it and tossing out this idea.
I `update' my copy then I go check the job jar. Could we solve the tracking issue by something as simple as renaming the file (appending your name) at this point and then `commit'?
-
Renaming a job file would be the same as creating a new file. There are a couple of ways I can think of for tracking.
The first is when somebody claims a job they can modify a line of text in the job file itself then commit the change. SVN would then flag individual jobs as having a status change. Once the job is done and the changes have been committed to the repository the job file can be updated again to indicate its status. If there are a large number of jobs this might be a bit tedious to manage.
The second is to maintain a separate file as a jobs manifest (just a list, really) with each line looking like "Job 0001 - Unassigned - Not started". When somebody claims a job they update the line in the manifest for that job like "Job 0001 - aUserName - in progress". SVN would flag the manifest as having changed, so the actual change would have to be looked at to see actual job details. Not as clear as updating individual job files but easier to see at a glance who is working on what.
With the nature of SVN, we could even have a single (or small selection of) collected job file(s) with all the jobs listed in it.
I'm partial to option two.
-
> Renaming a job file would be the same as creating a new file.
Ah, you can tell im a bit rusty with SVN things. Its been awhile.
yeah, i do like your second suggestion better too.
-
jbuzbee,
you are more then welcome to join the team. You can develop a ODCL version (of course, you don't have to either; you can just "fork" ours.) However, I apologize if I missed your "sign up", I didnt mean to leave you out of the team at all. Let me know your thoughts or plans.
I'm out, sorry guys. Good luck on your project!!
-
Didn't you have the UI built? I'm sure it would only take a few tweaks...
-
OK, this is what I have. I'm perplexed because I already have something similar, but I can't get this to work. Basically, the OpenDCL form code is designed to be called as sub in the command routine where it sets the layer name to the blackboard. The problem is it's not firing synchronously and I don't have the time to figure out why.
The short answer would be to have the OpenDCL code be the main command code and call subs in there . . . just a thought.
Sorry I don't have more time - time to go make some money.
-
Thanx jbuzbee.
I created an error handling set of files. I didnt know the best way to coordinate variable names, procedure names, etc. so i created three simple procedures. (Error_Start) (Error_End) ...and the generic error handler (Error). These are pretty generic, we can replace, redo, delete them; but i wanted to get the error ball rolling.
You will find them and a example usage in the "Trunk/Error Handler" directory.
-
I'd rather not use the SVN system, so attached is my contribution to the DCL part of the project :-)
A test function has also been included in the file for convenience.
Regards,
Lee
-
oh that's too bad. Is there any particular reason Lee; maybe someone else shares your sentiments and will be unwilling to participate in future projects. If we have the opportunity to address and fix those concerns now...
At any rate, thank you for the submission. Will you still be available for any revisions or additions?
-
Lee's code submission has been committed (added) to the project.
(revision 21)
-
oh that's too bad. Is there any particular reason Lee; maybe someone else shares your sentiments and will be unwilling to participate in future projects. If we have the opportunity to address and fix those concerns now...
I just want to keep things simple really - I'm happy to frequent the forums and participate where I can.
At any rate, thank you for the submission. Will you still be available for any revisions or additions?
You're welcome, and I'll certainly be available for revisions and additions.
-
@Lee: If you do not wish to learn SVN that is okay.
@Everyone else: dgorsman has obviously put a lot of time into developing this repository and i do not want anyone to misconstrue Lee's statement by thinking that SVN is difficult or convoluted in any way; nothing can be simpler then a right click. The TortoiseSVN interface is integrated into Windows Explorer and commands are accessed through a right click on a file or folder. For example: To add files to this repository, the tasks are very simple: Place or create a file in a folder, right click on it and "ADD" it. Then "COMMIT" the project (`commit' in this instance means to upload the file(s) to theSwamp).
A Version Control System (VCS) is used by all professional developers (and other trades). The SVN version control system is very simple to use and is widely used. There are many VCS' to choose from, each one has their benefits and quirks but once you understand the basics of one brand of a VCS you can usually start using another brand in very little time (they all have the ability to "ADD", "UPDATE" and "COMMIT" files and folders).
If anyone needs help with TortoiseSVN please do not hesitate to ask. For those of you who are in the US, if you would be more comfortable talking on the phone i would be more then happy to have a conversation to walk you through the process.
-
I didn't put *that* much time into putting it together. That should be a good idea of how simple it is.
-
sorry guys, I'm pretty much swamped until mid/late summer. If I'm lucky I'll get a few minutes here and there to put something together. It would be much easier if I had the Internet out in the field.
*edit* I just went back to the kickoff thread. Can someone provide some documentation on the project so when I do get a chance, I can have some sort information to use as a base for the template. Getting the formatting setup should be relatively simple if I have something to work from. There doesnt have to be any particular format to the documentation provided, I just need plain text, and can setup the format from there.
Boss (Decision Maker):
Se7en
Lieutenant (Package Assembly):
dgorsman
jane doe
Assassin (Documentation editor);
barc
Chris
Soldier (Coder / Documentation creation):
Lee Mac
alanjt
pBe