Is Character Randomizer for V4 still working with Poser?
Bought it today since it was in the PA sale and looked useful:
http://www.daz3d.com/shop/character-randomizer-for-v4
However, it seems to be very very broken. Under Poser9 on Mac it doesn't change any shape but throws errors instead.
The main complaint of Python are missing variables, like:
NameError: name 'vScale' is not defined
I had a quick look on the scripts and Python seems to be pretty right - it references tons of variables that are neither defined nor set to any value. So I don't think that the errors have anything to do with Mac or the Poser version 9, it is just broken.
Could anyone verify this - i.e. does it run on any version of Poser?
I reported on the bug tracker, but that is still pretty slow - so if it simply doesn't work at all, I'd like to get rid of the purchase as quick as possible...
(and probably get kicked out of the PC membership by the refund again :-/ )
Comments
I think you will find the problem lies with the fact that Poser used a newer version of Python when they released the latest version of Poser.
Nevermind, I missed the Mac part in the OPs post.
Of course it could - that's why I ask here - so if anybody runs it successful on Poser7 or Poser8 please let me know here.
However, I doubt it. As mentioned, I had a look on the script. This variable "vScale" in line 14 is obviously undefined. I can't post the script lines here (for obvious reasons), but above this line there are only default system library imports and the usual scene > figure > actor sequence.
This "vScale" simply must be defined prior within this script - appearantly with a call to the random() function. But this is missing, so the errors are understandable and completely unavoidable with this script - no matter which Python version and which platform.
I somehow can't believe that a product with this type of error would stay unnoticed for a longer time, but I have no other explanation.
PS: The M4 version of the script seems to work, and it does have a line defining vScale plus lots of wrapper calls to check for the Poser version, the correct figure etc. etc. All of this is missing in the V4 version (which looks like a small truncated fragment of a "real" working script).
So I guess this answers my question - it is totally broken. Hmm... I got 4 weeks for the refund, right? If someone starts fixing it in the near future it may be in time, but...
PS: CORRECTION! I'm actually running Poser 9 Pro... sorry for that. But it won't make a difference for this type of error.
The Python version compatibility issue, as far as I know, applies to compiled scripts (.pyc). If the script can be read then it's plain text still (.py) and so should be running.
The script gets triggered by a pose file (like "Randomize Head.pz2") which only contains one line starting "Randomize Head.py".
That one, as decribed, immediately uses multiple variables that are nowhere defined/initialized. So the failure and error message is completely unavoidable. The other scripts are the same in structure.
I searched the package if there is another "library" script that could have been imported to deliver the variable values, but there is none. In this case it could have been a simple missing "import" line, but it isn't.
The M4 scripts are three times longer with lots of data checks, version branches and (of course) all of the randomize variables.
So the (current) V4 version is crippled and non-functionable. What a pity...
I guess Murphy's Law now has two options:
a) either keep it, and it will never be fixed (and I will fail trying to copy the missing bits from M4 -> V4)
b) or return it, and as soon as I do this it will get fixed the next day but I can't buy it for the same price again...
I tend to solution b) to be honest... less work, less trouble (except for the PC subscription woes) and less damage ;-)
I am being planning to purchase this item
Works good this in Poser pro 2012 for Macintosh?
Since this is a PA product, you might want to contact the vendor regarding any questions regarding the script. You can PM 3DUniverse or email him.
You can also contact him through his own site
http://www.3duniverse.co.za/
His customer support has always been excellent.
As far as I can tell, this item is abandoned - I eventually returned it close to the 30 day deadline and got my money back.
I did contact the author (as above), and although I did both directly and once by replying in public to one of his own forum topics, I never got any answer. I also opened a ticket on the DAZ bug tracker, which is still open and untouched by now 2 months later - and I guess it will remain so till the end of times...
Before I finally returned the product, I had some closer look on the content and the problems. The V4 scripts are about 1/3 of the content that the code actually should have - it misses whole series of definitions and calls, and it does so randomly. It looks as if someone or something has ripped it into pieces at one stage of the development-to-final-product process.
The M4 scripts looked better, and some initial tests did work for me (although I didn't bother to run extensive tests, due to the state of the V4 version). I managed to manually "repair" one of the more simple V4 scripts by comparing it to the M4 one and to add/guess various code lines that were missing. This "repaired" script did work then, so the whole thing could work in principle...
However, the value-for-money bits of a package like this are in the "fine tuning" parts, and not in missing lines that are obvious enough so that you can fix it on your own - and, frankly, I don't want to spend money on anything that I'm forced to entirely re-write myself then...
I can't speak for the DAZ Studio scripts, because these are completely separate and I didn't bother and didn't test.
For that reason (and for Poser): As long as no-one from DAZ cleary states "sorry, that package was damaged, we have entirely restored it to a working version" you should stay clear off it... Sorry.
I think that daz3d has to contact with the vendor and no adding defective products (for poser) at your store. It is not good marketing strategy being bothering the Poser customers.
I had the same problem with the item 1 Click M4 Face Hair for Poser content problem why it does not run in Poser pro 2012
Is "Character Randomizer for M4" working for poser 7/poser 9 today? Are the problems fixed?
Good question - I'd love to know myself.
The DAZ bug ticket was never handled, and 3DUniverse never ever bothered to reply to my questions. So to my best of knowledge: Stand clear off it.
Or - even better - just buy it, and if it is still broken return it immediately and tell us here.
I was somehow tempted to try out as well again, once the Genesis version is out now, but (on the other side), since the customer support was/is that poor for the V4 version, one might doubt if the new stuff is reliable at all.
The Randomizer for V4, M4 and K4 should work in Poser 7/Pro and below. I don't have them so I don't know if they are compiled or uncompiled.
While most uncompiled Python scripts will work in newer versions of Poser, those that require tKinter can run into problems.
If you are using P7 or the original Poser Pro, it will work. If you buy it for a version higher than these, you may have to return it.
I'm surprised 3DUniverse didn't respond to previous questions. Usually I receive good customer support with his products.
I already mentioned that the scripts are hardly 1/3 of the size that they should be (compared to the M4 version which may work - although I didn't try to a huge extend).
When I wrote that the scripts are broken, I didn't mean broken in a sense of "doesn't work as expected", I meant "broken" in a sense of... well, imagine a car without a motor, with 3 wheels missing, without half of the windscreens and with the lower part of the oil pan ripped into pieces. That broken...
I can give you an example, here are the first 14 lines of the RandomizeHead.py file:
This is the first line that just goes "BANG" - obviously (even for coding beginners) vScale is undefined, and none of the imported libraries (see the import lines on top) are supposed to contain this. If you compare the script to the M4 version it becomes obvious that tons of extra lines are missing (in all of the scripts, not just the head) that of course also define vScale with a random value based upon other parameters (and experience with the specific model).
After all, the "plot" of a script like this is not to just set some parameters (that would be trivial and would not justify charging $30+ for it), the core magic is the logic behind "good values" and "bad values" for that. For that reason it also isn't a solution to just copy "missing bits" from the M4 script - granted that an average user could do that at all in the first place.
I apologize if I may sound rude, but I must say that the state of this item is just a shame, and I have no excuse for leaving it like that.
PS: For the Python coders...
You may have wondered why the code includes this, although "figure" is not used afterwards:
The reason is simple, the same script for M4 includes 100+ extra lines checking the correct model, the correct Geometry files, certain current values, etc. etc. etc. - all of this makes sense.
The V4 script seems to be the "core code" ripped out of an existing huge script, something like a raw debug version without all safety routines, but also missing essential fragments.
Of course you may add something like this yourself:
This will "fix" it for some more 4-5 lines until it chokes on the next error...
I told DAZ on the bug tracker (no response), I told 3DUniverse on one of his/her own threads (no response), I told 3DUniverse by PM (no response), I told 3DUniverse on their own website (no response), I gave up and got my money back...
A fix for this was submitted to the DAZ store and put live on 22 January.
Thanks for finally letting me/us know... sadly it's a bit late for me, since I returned the stuff August 2012 after waiting several weeks. Was there a specific reason for not telling me any answer back then? After all, I made clear that I would even keep the broken version if anyone (DAZ and/or the creator) would have just told me "we're working on it, and it will be fixed sometimes in the future"...
If you "just don't like me" then please tell me... I'm brave and can bear it, promise... (just kiddin' :-))
I bought it half price (one of the many surprise sales I guess) last autumn. So I won't buy it full price now anyway. Maybe "one day".
Which leads me to a curious question (to DAZ) - can I actually buy an item that I already returned once in the past, or does this conflict with the shop TOS and/or crashes the shop, my account, the universe-as-we-know-it?
Please let me know and if it's possible I will make a mental notice for the item (wishlist is currently broken... not kiddin' this time :-))
Nothing personal. We have been very busy, and this issue slipped through the cracks.
cheers,
S
You can buy the same item as many times as you wish (although why would you need two copies I don't know).
It is an situation that occurs often, - when something goes on sale, people who bought that something with more expensive price within 30 days ago will buy it again in lower sale price and ask for refund for previous, more expensive order, saving some money on the difference. For a couple of days while refund ticket is processed they possess both.
Also, the V4 version is still sparse compared to M4. I'd describe the V4 version more like a work in progress rather than a complete product. None the less, this utility (if working as described) has the potential to save me a great deal of time and aggravation towards my current project (hence the purchase). I was given the impression that issues with these products had been fixed.
Is this a PP2014 issue, or did the 22 Jan (2013) fix only apply to the M4 version? If so, the V4 version is in serious need of attention.
I'm not opposed to manual repairs, but I am not familiar with python scripts. If no fix is available from the author (3D Universe), would someone kindly steer me in the right direction to get this working properly?