On Sat, 2002-08-10 at 13:03, Richard Atterer wrote: > Hello Nelson, > > I'm not sure about the source of the problem - you are the first > person to report something like that. What version of jigdo-lite are > you using, and on what kernel? > Sorry for the lack of information 8-)
the Jigdo version is the last: 0.6.8 I'm using for now Debian Woody Unstable, dated April 15, 2002 using now the kernel 2.4.18-k7 (got a AMD Athlon 950) the date for all the dics that I have it's not the same: 1. 15-04-2002 2. 29-03-2002 3. 29-04-2002 4. idem 5. idem 6. idem 7. idem 8. 15-04-2002 how do you figure out it, APT has a lot of files holded. > On Fri, Aug 09, 2002 at 08:36:41PM -0400, Nelson Suniaga wrote: > > Yes, I got a .iso.tmp image (why still it has a .tmp > > extension?), > > Because it isn't finished yet. Once all the data has been written to > it, it is renamed to just ".iso". > > > `/mnt/iso/pool/main/g/gnome-libs/libgnome32_1.4.1.4-3_i386.deb' > > does not match checksum in template data > > That is weird. It means that when jigdo read in that .deb for the > first time to find out its checksum, the checksum matched, but when it > later wanted to write the .deb to the .iso.tmp file, the checksum no > longer matched. Exactly! that's very strange, it's just like someone o some thing change the numbers inside the file after the process reading. > > > Really I can't get it, can someone explain me this? ah...! almost I > > forget it: all those files with wrong checksums are differents > > everytime that I run jigdo-lite... > > Very strange indeed. Are you sure the CD/harddisc containing /mnt/iso > does not have read errors? > my hard disc it's some old (Samsung with 6 GB) but anyway, how I can know if my disc it's possesed with bad spirits or bad sectors? I did it a resizing to the partitions, cause it didn't has enough space for the iso image... the image it must will be created on /opt/tmp Filesystem 1k-blocks Used Available Use% Mounted on /dev/hda5 2583624 2062476 468620 82% / /dev/hda6 514144 388736 114968 78% /home /dev/hda7 1391380 654852 722392 48% /opt /dev/hda1 1627368 1503172 124196 93% /win what do you think about it? Ah...! almost forget it: 128MB on RAM. > > and other thing: if I run jigdo-lite like a "normal user" the > > .iso.tmp image is not generated completely (only 13 MB) but if I am > > root then I get a file of 678 MB > > "Interesting." :-/ > > > This happen also with the US jigdo and template files, just the > > message it's different: > > > > Not downloading .template file - `woody-i386-1.template' > > already present > > This message means that the template data matches the checksum that is > listed in the .jigdo file... > > > Found 375 of the 1352 files required by the > > template > > zstream.cc:282: Assertion failed, `dataUnc > 0 || (status > > == > > Z_STREAM_END || status == Z_OK)' > > ...but this indicates that the template data is corrupted. > > I cannot imagine how this could possibly be a bug in jigdo - it may be > bad RAM, a broken HD... :-| > > > Question 1 : I must or not use that image? > > Whether you need CD 1 or CD 1_NONUS is your decision. Our > recommendation to end users is to use 1_NONUS. jigdo should be able to > download both. > Yeah, I can understand that, Jigdo download both, but none it's totally created. > > Question 2 : What happen with the DEBs with wrong md5sum after of > > the installation??? > > As long as there are .debs with wrong md5sums, the .iso.tmp will never > be finished; if you burn the .iso.tmp file instead of a final .iso > file to CD-R, then the installation will break in weird and wonderful > ways. > > Sorry, I cannot really help you! > Cheers, > > Richard Well... maybe the answer would be to wait another release of Jigdo (0.7.0 perhaps...? 8-) Anyway, thanks so much for all, now I know that it's not me, it's other thing... it's something unknown 8-) > > -- > __ _ > |_) /| Richard Atterer | CS student at the Technische | GnuPG key: > | \/�| http://atterer.net | Universit�t M�nchen, Germany | 0x888354F7 > � '` � -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
