Page 1 of 2

Texture frustration

Posted: 12.07.2004, 00:15
by rconn2
Hi everyone. I'm a new Celestial user... been playing around with it for about a week. It's a great discovery and I commend all who've contributed to its success! So, take what I have to say below with a grain of salt (and yes, I recognize I haven't contributed) ;)

Like most new users, I want a bunch of higher res textures. So, I've been busy downloading (on a dialup).... overnight, or leaving the PC to run for hours. Then, after lengthy downloads, I _frequently_ encounter bad files! One in a while, corrupted files are to be expected. But, even with apparantly good dl's, there are still too many problems.

For example, I just dl'd levels 0-2 of textures and normals for Mars and Earth, and Moon, from John Van Vilet's -- "Johns Textures" -- site. I could only unzip _one_ of these. The rest had "unknown compression" problems. I've tried WinXP, WinZip, WinRAR, and 7-Zip... none work. Is some _secret_ method being used? While I know it's free labor, still, isn't there an obligation to check one's work so other's won't be frustrated and waste their time? I can't imagine not doing a once-over of my stuff.

Even Harald Schmidt's 32k earth textures had some corrupted files (and one of them I also dl'd using someone else's DSL and the result was the same). Luckily, WinRAR was at least able to salvage most.

Here's a suggestion: before allowing links on the MotherLode, please check that they and the files actually work. Celestial is a great program... but for new users, and I'm far more patient and techie than most, the amateurish quality and nature of the downloads seriously undermines it.

Ok... flame away... but you'd be p*ssed too (and probably have been).

-- rc

Posted: 12.07.2004, 02:10
by lostfisherman
ruh roh!

You have probably been using download managers, they are terrible, they will sometimes corrupt files, and you won't get you the stuff you want any faster than the normal save as method. I uninstalled my download manager a long time ago. If you haven't been using one, well, Johns downloads off the motherlode site MOSTLY work for me (you will need 7-zip to extract them, I think).

The Celestia motherlode site has three errors as I have found very recently, Earth level 5 part 6 (virtual texture) is faulty, Mars VT Level 5 part 8 is the same, I strongly doubt that the Mars level 4 part 3 is only 5 megs big. There are I think files that have not been downloaded correctly and put up by honest mistake on motherlode.

No flames from me, but patience and the net do go hand in hand, get the files steadily..

Posted: 12.07.2004, 19:51
by Guest
Try using Winzip 9.0

re

Posted: 13.07.2004, 08:20
by John Van Vliet
7-zip is what i use ,but they are bz2 compresed , 7zip should auto-detect the file type .

mars level 4 part 3 is 56.1 meg

Posted: 13.07.2004, 19:53
by alphap1us
Hi all,
The three files that lostfisherman brough my attention have been fixed I think. Corrupted downloads are just something you have to live with when dling big files. It really does help when you notify me as losty did. I can't read all the forum posts so it will take a very long time for me to sniff these ones out on my own. Please use the contact form to let me know of any toehr problems with the motherlode.

Cheers,
Joe[/b]

Posted: 13.07.2004, 21:26
by t00fri
alphap1us wrote:Hi all,
...
Corrupted downloads are just something you have to live with when dling big files.
...
Cheers,
Joe[/b]


Joe,

I am surprised. What would we do in my "trade" if we would encounter any corroded files in our superfast research networks?? This would be simply catastrophic...

I never have encountered any faulty transfers despite huge "transactions" in our UNIX networks. Block checking is always active in TCP/IP networks, which leads to automatic retransfer, if there was an error.

I think corrosion mainly arises due to download accelerators which are designed mainly for the "home" environment...

Bye Fridger

Posted: 13.07.2004, 22:49
by Christophe
t00fri wrote:I never have encountered any faulty transfers despite huge "transactions" in our UNIX networks. Block checking is always active in TCP/IP networks, which leads to automatic retransfer, if there was an error.

I think corrosion mainly arises due to download accelerators which are designed mainly for the "home" environment...


Indeed, the problem is not TCP/IP but HTTP, as the name implies it is designed to transfer text files, not large binary files. So called download accelerators try to work arround the limitations of HTTP but they clearly are not perfect. And that's what protocols like s/ftp or rsync are for.

Posted: 14.07.2004, 00:57
by rconn2
John V: Sorry if it came across that I was criticizing you. I appreciate your effort. Tonight I had access to a T1 and, from the motherlode, downloaded and successfully unzipped all your MarsNormalLevel's, and MarsLevel_0_1_2 and MarsLevel3.

However, I dl'd the MarsLevel4's several times, and was unable to unzip (using the latest WinRAR) any of them. I got the same error: unknown compression method.

I had similar problems with some of your textures from your web site the other day when I started this thread -- forget which ones. In that case, I used 7-zip (latest version), WinRAR, WinZip and even XP's extract... and there was the same error.

BTW, I don't use download managers.

-- rc

re

Posted: 14.07.2004, 06:59
by John Van Vliet
Hi i did not take it as a critisum . i am a very bad speller so my posts tened
to be short . It dose seam odd that 7zip is having a hard time at trying to open the files .
I just dl my own files and marslevel4 p1-4 open just fine for me . it could be that with dile-up part of the file header is getting cut, the part that tells what type of file it is ?

Posted: 14.07.2004, 16:10
by alphap1us
Hello Fridger,
t00fri wrote:I am surprised. What would we do in my "trade" if we would encounter any corroded files in our superfast research networks?? This would be simply catastrophic...

I never have encountered any faulty transfers despite huge "transactions" in our UNIX networks. Block checking is always active in TCP/IP networks, which leads to automatic retransfer, if there was an error.

I think corrosion mainly arises due to download accelerators which are designed mainly for the "home" environment...

Bye Fridger


I am not sure what you are getting at. I can't remember having a problem with a university network, or even home DSL, but you and I are on dial-up at home, where the are lots of mysterious hang-ups and errors. In this case, the corrupted downloads may have been completely my fault, but this is not always the case with slow connections

Cheers,

Joe

Posted: 14.07.2004, 17:27
by t00fri
alphap1us wrote:Hello Fridger,
t00fri wrote:I am surprised. What would we do in my "trade" if we would encounter any corroded files in our superfast research networks?? This would be simply catastrophic...

I never have encountered any faulty transfers despite huge "transactions" in our UNIX networks. Block checking is always active in TCP/IP networks, which leads to automatic retransfer, if there was an error.

I think corrosion mainly arises due to download accelerators which are designed mainly for the "home" environment...

Bye Fridger

I am not sure what you are getting at. I can't remember having a problem with a university network, or even home DSL, but you and I are on dial-up at home, where the are lots of mysterious hang-ups and errors. In this case, the corrupted downloads may have been completely my fault, but this is not always the case with slow connections

Cheers,

Joe


Joe,

I was talking mainly about the professional side, since you made such a /general/ statement that one has to live with corroded transfers once large files are involved...

Professionally, file corrosion as a consequence of TCP/IP network transfer is virtually unknown. I just wanted to point this out.

Moreover, my modem transfer is slow, but also there I get virtually no hangs whatsoever. I can have the modem run for hours without any problems /from my side/ at least. If I take my lab as an Internet provider then the modem NEVER shows any irregularities. The only drawback is that commercial I-providers are cheaper...

Bye Fridger

Posted: 15.07.2004, 03:38
by rconn2
John: I think I dl'd the bad MarsLevel4's (using a T1 not dialup) from the Celestia Motherlode site. Maybe they're ok on your http://johnscelestiapage.no-ip.com/Virt ... php?cid=11 site? I'll try from there again this week and let you know.

And thanks... what I did dl (levels 0-3 and all the normals) looks nice!

-- rc

Posted: 19.08.2004, 13:25
by Stormyman
Hi,
I had exactly the same problem. Simply install the latest Winzip and I think it will work ! You can download at http://www.winzip.com .
Sometimes the latest Winrar version and other programs are still unable to unzip some zip files.
I think then the problems will be gone :D

Posted: 19.08.2004, 23:10
by Slalomsk8er
The download problem for big files is old and easy to understand plus it can happen in small files to but this in unlikely by chance.

The error detection can only find uneven numbers of wrong bits in a TCP/IP packet!


This is one of the reasons why I started this Topic.

P2P-Networks are mostly build for big files like movies so they check the file with a Hash like MD5.
Some of them do this for junks of say 10 MB to, so you don't have to download the file again.

Bye Dominik

Posted: 20.08.2004, 14:40
by Jon R
I've tried downloading the John van Vliet textures twice and have experience the same "unkown compression" issue both times. There must be something wrong more than the process of downloading because I've been downloading big binary files for years and have never experienced any corruption. I just downloaded Win XP SP 2 (226MB) file which took 15 hours where I had to disconnect and reconnect (on purpose because my ISP had a limit of 8-hours/logon) and I had no problems.

Posted: 20.08.2004, 14:48
by Harry
Jon R wrote:I've tried downloading the John van Vliet textures twice and have experience the same "unkown compression" issue both times.

Did you use 7-zip? John wrote that this is what he used, standard zip programs may not be able to decompress these files. Can someone please confirm if these files work with 7-zip, or are really corrupted?

Harald

re

Posted: 20.08.2004, 22:06
by JohnVanVliet
ok my files are bz2

seeing as so many are having a problem
they will be converted to tar.bz2.

Re: re

Posted: 21.08.2004, 07:53
by don
JohnVanVliet wrote:ok my files are bz2

seeing as so many are having a problem
they will be converted to tar.bz2.

Hi John,

Why don't you simply use a normal Zip program / compression type, since most folks have the ability to UNZip a normal Zip file already built-into their systems? In WinZip, I use the "Maximum (portable)" compression setting and nobody I share with has ever had any problems with this.

Personally, saving another few percentage points on compression is not worth the struggle the file receiver has to go through in locating, download, installing and setting up yet another Zip program (WinRAR, 7-Zip, BZ2, or any other) -- IMHO.

Posted: 21.08.2004, 21:36
by Bob Hegwood
Sorry,

An interjection from the Brain-Dead...

7-Zip is a real pain in the Wazoo for us non-Doctors of Theoretical Physics.

Normal people use any of a wide variety of free zip programs. The one I'm
using, Zip Central, is absolutely free, is very easy to use, and
has given me NO problems when zipping or un-zipping most downloaded
files, no matter the compression method. I'm with Mr. Gee on this one.

Just my opinion, and I *know* you didn't ask for it, but why introduce even
MORE complications into an already complicated system?

Thanks, Bob

Posted: 22.08.2004, 00:43
by don
By the way folks, WinZip 9 DOES unzip John's files okay. At least the ones I've tried.