PDA

View Full Version : Set Content Directory



jaf
06-05-2018, 11:01 AM
I model something in Modeler 2018.0.4 and save it to directory xdir, "Close All", "Load", and then "Set Content Directory" choosing an existing directory -- ydir. Then try to load an object, but the initial directory (xdir) is displayed instead of the content directory. Is there a logical reason it should work this way?

madno
06-06-2018, 12:00 AM
Seems so, this happens when modeler is closed between the actions:

E.g.:

- Open modeler
- make a box
- save box to a "non content" directory somewhere
- close modeler

- open modeler again
- Load -> set content directory
- Load Object

= LW is not using the content directory but the "last save" directory

- Now save object and browse to the content\objects directory by hand
- Save object
- don't close modeler
- clear object
- Load -> set content directory (another one than before)
- Save Object

= LW will now use whatever content dir you chose

Seems to me, when you save / load from a "non content dir" and close modeler, it will update its' config files with the last save dir. And this then has higher priority than the content dir selection.

Is this logical? I personally don't think so. If I chose a content dir, LW should use it (that's the reason why I select it). But there might be people who prefer it the way it is.

jeric_synergy
06-06-2018, 01:26 AM
Yep. I've been complaining about this for DECADES.

Not only that, but some plugins that load files will also cause various bits of LW/LWM to look in unusual/incorrect areas. It is WEAK.


But since software magically descends from the clouds, obviously no one is responsible and all are equally competent.

jaf
06-06-2018, 03:27 AM
As I get older, I find I have less tolerance for these small inefficiencies. I also use Terragen a lot. A couple of days ago I asked the question "why, when I start the add object process and the file selection dialog comes up with two buttons (Open and Cancel), if I select 'cancel', a warning message shows up stating 'no object specified ....' and a blank object reader is added to the node network?" Doesn't "cancel" mean cancel? The answer was "that's the way it's always worked ......"

Lightwave seems to have a lot of these mature "head slappers". I do a lot of "load object into layer" and "save layers as object" and get a lot of crashes. Been that way for me since version 9 and probably back much further than that.

Experienced user's simply learn to avoid/work around these things and new user's are perplexed. Old user's get frustrated. And developers .......

jeric_synergy
06-06-2018, 07:37 AM
Experienced user's simply learn to avoid/work around these things and new user's are perplexed. Old user's get frustrated. And developers .......
....implement FAR more complicated things (::cough:: PBR renderer ::cough:: ) while IGNORING these boring little housekeeping bugs that stick a little needle into users every damn day.

I call them speedbumps, but really it's like owning a cat that claws your leg just when you're starting to get productive.

Why is that? Are they just not sexy enough programming to be fixed???

raymondtrace
06-06-2018, 10:36 AM
"boring little housekeeping bugs" are probably more shrugged off by users and not officially reported.

Only observing from the outside, the bug reporting system does not appear well suited for duplicate reports. If something has already been reported, the more recent report is just closed, instead of being marked as a duplicate. I suspect this makes issue priority a challenge.

jaf
06-06-2018, 11:34 AM
"boring little housekeeping bugs" are probably more shrugged off by users and not officially reported.

Only observing from the outside, the bug reporting system does not appear well suited for duplicate reports. If something has already been reported, the more recent report is just closed, instead of being marked as a duplicate. I suspect this makes issue priority a challenge.

You're probably right. When I had a real job (software engineer on government contracts) it was nearly impossible to have bugs/bad design slip through the cracks -- too many people reviewing/documenting/status reviews, etc.

I really like LW -- it does what I want and seems there's usually a plugin (or work around) if LW can't do it efficiently. I just don't understand Newtek seemingly ignoring things that a new user gets told "oh, it's been that way for a long time ... here's a workaround ... Does NT marketing really not think they should do whatever needs to be done to fix these types of problems.

jeric_synergy
06-08-2018, 11:37 AM
"boring little housekeeping bugs" are probably more shrugged off by users and not officially reported.
As you might imagine, NOT BY ME. I whine about them at the top of my lungs.

Because it's FREEKIN' RIDICULOUS that they can write a g.d. PBR renderer, but they can't make CONTENT DIRECTORY or RECENT FILES work right. :devil:

And something tells me that files are loaded and saved a lot more often then huge volumetric clouds are rendered. That shiz should be bullet-proof and no-brainer, not Yet Another G.D. Lightwave Speedbump, that we have to shame-facedly explain away to our peers. :thumbsdow

++++

I don't really know about duplicate bug reports, and how they are handled-- I certainly hope that the BEST description/characterization of a bug/misfeature is retained, rather than just the first or last (not all bug reports are created equal): >>IF I WERE DESIGNING IT<<, the bug reporting system would give "weights" to bug reports, both by number of reports and a reputation value of the reporter. EG, Ryan Roye and Monroe Poteet would have the highest reputations, and bug reports from them would have more weight than from, say, moi.

raymondtrace
06-08-2018, 02:28 PM
...we have to shame-facedly explain away to our peers...

Cast off the shame. Just proudly tell peers that LW invites you to fully explore the filesystem and doesn't limit your view to a particular directory.

jeric_synergy
06-08-2018, 03:07 PM
:D That's a very nice way to say "is befuddled by the slightest deviation from corporate defined workflow", but sure, gild that jalopy. ;)

jaf
06-19-2018, 01:43 PM
Here's another puzzling thing (at least too me) with file loading. In the image, the top part shows the LW objects in a folder. The bottom part is what LW shows for recent files. The most recent is "CASE780C.lwo and the second most recent is CASE780C_v007.lwo. I wanted to use load|recent v007, but it's not shown in the dialog (neither is v006). Why is that?141980

jwiede
06-19-2018, 05:13 PM
I can 100% repro the problem with LW2018.4, incremental saver doesn't properly update the "Recent Files" entries. Can't repro the problem here with LW2015.3, "Save Incremental" updates "Recent Files" in LW2015.3, so it's a new bug (in the apparently-new "Incremental Saver" in LW2018).

I recommend you file a bug report on it.

jeric_synergy
06-19-2018, 07:26 PM
I reported this years ago.

That it hasn't been fixed is VERY weak. How difficult could it be????

Think of all the users that loaded NOT the latest versions because of this.

c.1
06-19-2018, 08:38 PM
This totally reminds me of my fulltime workplace.

Quote from a long time employee: “Wow, ******** systems are not user friendly”

Response: “really, duh.......how long have you worked here”:D

jwiede
06-20-2018, 02:41 AM
I reported this years ago.

That it hasn't been fixed is VERY weak. How difficult could it be????

Well, it apparently WAS fixed, in that it's working properly in 2015.3. It is broken (again?) in 2018.4, though.

Do you recall a prior version where it was broken?

jeric_synergy
06-20-2018, 07:51 AM
I've NEVER known a version where it worked correctly.

ianr
06-20-2018, 08:35 AM
Well, it apparently WAS fixed, in that it's working properly in 2015.3. It is broken (again?) in 2018.4, though.

Do you recall a prior version where it was broken?


A bug report in concise cool reportage to Mr. Deuce Bennett should get things moving?:newtek:

madno
06-20-2018, 09:13 AM
I sent a report a few days ago.
It got assigned to a developer, according to a reply email :-)

jwiede
06-20-2018, 05:30 PM
I've NEVER known a version where it worked correctly.

To be clear, we're talking about the "incremental saver doesn't update 'Recent Files'" problem now (per post #11 in thread (https://forums.newtek.com/showthread.php?157256-Set-Content-Directory&p=1548905&viewfull=1#post1548905)), not the "LW requesters not in Content Dir" issue. Is that what you were referring to as well (incremental save vs recent files)?

I just re-checked and LW2015.3 Mac64 definitely does update the "Recent Files" list with latest incremental save after "Save Incremental" as expected.

jaf
06-20-2018, 08:14 PM
Yeah, I didn't see the need to start a new thread on what (to me) seems like pretty basic file handling/workflow stuff.

jeric_synergy
06-21-2018, 04:43 PM
To be clear, we're talking about the "incremental saver doesn't update 'Recent Files'" problem now (per post #11 in thread (https://forums.newtek.com/showthread.php?157256-Set-Content-Directory&p=1548905&viewfull=1#post1548905)), not the "LW requesters not in Content Dir" issue. Is that what you were referring to as well (incremental save vs recent files)?

I just re-checked and LW2015.3 Mac64 definitely does update the "Recent Files" list with latest incremental save after "Save Incremental" as expected.

Just to make sure, because "aging", I fired up LWM 2015.3, and it's still there, see jpg, and I'm pretty certain I've been b!tching about this for several years. (W10/LWM2015.3)
142002

Just to make certain, I closed and relaunched LWM to make sure it wasn't some stupid "No, you're supposed to stay with the base name, increments are for archive purposes" balderdash*.

Seriously, this is DEAD SIMPLE STUFF. Why won't they fix it? And if it's working "how it's supposed to work!!1!!", they are wrong wrong wrong in the design.


*paraphrase.

jwiede
06-21-2018, 06:55 PM
Just to make sure, because "aging", I fired up LWM 2015.3, and it's still there, see jpg, and I'm pretty certain I've been b!tching about this for several years. (W10/LWM2015.3)
142002

Just to make certain, I closed and relaunched LWM to make sure it wasn't some stupid "No, you're supposed to stay with the base name, increments are for archive purposes" balderdash*.

Very interesting, are you on Windows? Also, does it work if you're saving base & incrementals within the {Content Directory}(/Objects)?

I'll test further when I get a chance. Here, just reconfirmed (again), on Mac LW2015.3, saving (base &) incrementals to {CD}/Objects using "Save Incremental" they were being entered into "Recent Files" list as expected.

jeric_synergy
06-21-2018, 07:16 PM
W10, as noted. I did simply save in whatever directory LWM threw up, didn't set one, because of course it should DEFAULT to the content directory.

Can you post a screenie if the RECENT FILES menu offerings when you do this? There should be several increments of any mesh one has saved incrementally. I'd just like to see it once before I die.
142003

jaf
06-22-2018, 12:08 PM
3D Coat and Terragen both work the way I believe LW should. Save incrementally and exiting should put that last save on the top of the recent file list. Why would one want to see a partial list of recent files?

jeric_synergy
06-22-2018, 04:42 PM
3D Coat and Terragen both work the way I believe LW should. Save incrementally and exiting should put that last save on the top of the recent file list. Why would one want to see a partial list of recent files?

I believe I had (at some point) s/w that would show the last 3 (or so) incremented versions of any given file, and then the rest of the RECENT FILES list were other files, sometimes w/multiple versions. FWIW.

OTOH, I don't want software to wait until I exit to put a saved file at the top of the list: I want it there immediately.

I'd also like "SAVE AS COPY" in LWM -- seriously, how hard can that be?? They already DID IT in Layout. C'MONNNNNN. :devil:

("Save As Copy" saves a file without changing the default name that autosaves.)

jaf
06-22-2018, 05:48 PM
Eric, I probably worded my response badly. I did mean any save increment should go to the top of the list -- just put "exiting" there thinking of my workflow as "done for the day, I'll pick up where I left off ...."

This problem got my attention when I was working a few different models and when I did a load recent file I started seeing things I thought I had already changed. Well, I did make the changes -- I just loaded the wrong file (because LW said it was the most recent.)

jeric_synergy
06-22-2018, 06:07 PM
Eric, I probably worded my response badly. I did mean any save increment should go to the top of the list -- just put "exiting" there thinking of my workflow as "done for the day, I'll pick up where I left off ...."

No worries: it's tough to be REAL precise in this medium, and after all they're not paying us to design their software.

This problem got my attention when I was working a few different models and when I did a load recent file I started seeing things I thought I had already changed. Well, I did make the changes -- I just loaded the wrong file (because LW said it was the most recent.)
EXACTLY the danger with the poor implementation that we currently have. I can't BELIEVE anybody meant for it to work this way, it's total rubbish. We'd be better off without it, because it makes it SEEM like we're working on the latest.

As above shown, I reported this in 2014. It's still on their books as OPEN. :devil:

jwiede
06-23-2018, 02:53 PM
Can you post a screenie if the RECENT FILES menu offerings when you do this? There should be several increments of any mesh one has saved incrementally. I'd just like to see it once before I die.
142003

Sure...

142014

jeric_synergy
06-23-2018, 06:12 PM
...soooooooo envious....

Looks like it works on the mac. :'(

jwiede
06-24-2018, 02:41 PM
...soooooooo envious....

Looks like it works on the mac. :'(

Well, it did work, anyway, as LW2018 appears broken again.