1*17049c45SAxel DörflerBogus bugs (not our fault!): 2*17049c45SAxel Dörfler--------------------------- 3*17049c45SAxel Dörfler 4*17049c45SAxel Dörfler By far THE BIGGEST source of bug reports to Info-ZIP/zip-bugs is the 5*17049c45SAxel Dörfler incorrect transfer of zipfiles (or of the UnZip executable itself). 6*17049c45SAxel Dörfler ALWAYS TRANSFER IN BINARY MODE! This includes ftp transfers and *both* 7*17049c45SAxel Dörfler ends of a Kermit connection ("set file type binary"). If your copy 8*17049c45SAxel Dörfler isn't exactly the same size as the original, you made a mistake. 9*17049c45SAxel Dörfler 10*17049c45SAxel Dörfler Another common source of errors such as "compression method 8 not sup- 11*17049c45SAxel Dörfler ported" is the existence of an old version of UnZip somewhere in your 12*17049c45SAxel Dörfler path. Make sure you're using the version you think you're using; give 13*17049c45SAxel Dörfler the full path explicitly if necessary. Executing "unzip" without any 14*17049c45SAxel Dörfler options will print a help screen, at the top of which is the UnZip 15*17049c45SAxel Dörfler version number and release date; and executing "unzip -v" without any 16*17049c45SAxel Dörfler zipfile or other options will give information about what compiler was 17*17049c45SAxel Dörfler used, the target operating system, any special UnZip options, and the 18*17049c45SAxel Dörfler date of compilation--only for version 5.11 and later, though! (Also, 19*17049c45SAxel Dörfler under Unix C shell and some Bourne shells, "which unzip" will print 20*17049c45SAxel Dörfler the path of the unzip you're actually using. Under OS/2 and MS-DOS, 21*17049c45SAxel Dörfler whch21gr.zip [on Simtel mirror sites] will do the same thing; in addi- 22*17049c45SAxel Dörfler tion, "which -a unzip" will show *all* copies of "unzip" in your path.) 23*17049c45SAxel Dörfler 24*17049c45SAxel Dörfler 25*17049c45SAxel DörflerBugs (real and/or imagined): 26*17049c45SAxel Dörfler--------------------------- 27*17049c45SAxel Dörfler 28*17049c45SAxel Dörfler - [OS/2 DLL] when trying to use the REXX function UzUnZipToStem to extract a 29*17049c45SAxel Dörfler file with `&' in its name, the DLL crashes (but UzUnZipToVar still works) 30*17049c45SAxel Dörfler [Daniel H, 961215] 31*17049c45SAxel Dörfler - UnZip has problems with archives bigger than 2GB; it may print "note: didn't 32*17049c45SAxel Dörfler find end-of-central-dir signature at end of central dir" (harmless) or 33*17049c45SAxel Dörfler may not be able to seek to member files [James Lemley 970107, Iris Spaniol 34*17049c45SAxel Dörfler 970206, ...] 35*17049c45SAxel Dörfler 36*17049c45SAxel Dörfler - fix overwrite behavior: hidden/system problems?; etc. 37*17049c45SAxel Dörfler - 32-bit DOS UnZip still unable to set volume labels? 38*17049c45SAxel Dörfler - 32-bit DOS UnZip under OS/2 doesn't extract all wildcard zipfiles? 39*17049c45SAxel Dörfler [DOS box: unzip386 (ver 5.12) x:\32bit\unix\emx09a\*.zip, Hobbes 3/95] 40*17049c45SAxel Dörfler - 32-bit DOS UnZip under OS/2 doesn't set timestamp when overwriting files 41*17049c45SAxel Dörfler on HPFS partition? (go32 and pmode/w both; emx/rsx OK) [Eberhard Mattes 42*17049c45SAxel Dörfler 950726] 43*17049c45SAxel Dörfler - USE_FWRITE still causes occasional CRC errors when extracting on Pyramid? 44*17049c45SAxel Dörfler [Kevin Fritz 931102] 45*17049c45SAxel Dörfler - still NT/W95 bug with "unzip -v d:*.zip" not matching properly? [Steve S 46*17049c45SAxel Dörfler 940527] 47*17049c45SAxel Dörfler 980427: bug no longer exists, Opendir() must have been corrected by someone 48*17049c45SAxel Dörfler 49*17049c45SAxel Dörfler - when ^Z received in no-echo mode, echo is not restored (works OK if 50*17049c45SAxel Dörfler resume, however) 51*17049c45SAxel Dörfler - signal() handler disabled after first use with one of BSD/SysV? 52*17049c45SAxel Dörfler - MKS Korn shell: unzip assumes the MKS-style command-line environment 53*17049c45SAxel Dörfler options are relevant to it, but this is not the case if unzip was called 54*17049c45SAxel Dörfler by another program (e.g., from a .BAT file). A fix for this exists for 55*17049c45SAxel Dörfler Borland compilers but not for MSC, Watcom, djgpp, etc. 56*17049c45SAxel Dörfler - OS/2: for paths with one long component, the .LONGNAME EA may be saved for 57*17049c45SAxel Dörfler all components (waste of disk space): how to check?? 58*17049c45SAxel Dörfler - VMS: for extracting to other directories, only the VMS-style "-d [.foo]" 59*17049c45SAxel Dörfler format is accepted; "-d foo" should also be allowed. Long filenames are 60*17049c45SAxel Dörfler not automatically truncated to 39.39. 61*17049c45SAxel Dörfler - Novell Netware: Netware drives may clear the archive bit on extracted 62*17049c45SAxel Dörfler files under OS/2 and/or MS-DOS. UnZip always *tries* to set the archive 63*17049c45SAxel Dörfler bit, however. [pynq@uchicago, 940527] 64*17049c45SAxel Dörfler - DEC Ultrix: on long zipfiles, unzip will sometimes fail (bad CRC, not always 65*17049c45SAxel Dörfler reproducible); this is apparently due either to a hardware bug (cache mem) 66*17049c45SAxel Dörfler or OS bug (page faults?) [Igor, Jean-loup, bottom of BUGS.long] 67*17049c45SAxel Dörfler - funzip/more/decryption/no-echo bug: race condition(?) causes terminal to 68*17049c45SAxel Dörfler be "reset" to no-echo state 69*17049c45SAxel Dörfler - Macintosh (100200), Atari (020000) external file attributes not interpreted 70*17049c45SAxel Dörfler correctly (both unzip and zipinfo) 71*17049c45SAxel Dörfler - pkbug error: zipfile with incorrect csize and/or ucsize--check for end of 72*17049c45SAxel Dörfler compressed (csize) data in uncompression routines: 73*17049c45SAxel Dörfler unreduce.c: while (((outpos + outcnt) < ucsize) && (!zipeof)) { 74*17049c45SAxel Dörfler [James Birdsall, Mark, bottom of BUGS.long] 75*17049c45SAxel Dörfler - OS/2: directory EAs not restored if directory exists [Kai Uwe, KG27515@uark] 76*17049c45SAxel Dörfler (subsequent note: no way to determine which EAs are newer ==> cannot 77*17049c45SAxel Dörfler restore without user input) 78*17049c45SAxel Dörfler (update: as of UnZip 5.30, option -o forces restoring of directory EAs) 79*17049c45SAxel Dörfler - MS-DOS: Borland executables don't allow other than 80-column, 25/43/50-line 80*17049c45SAxel Dörfler screen modes (Borland bug) [Michael Stillwell] 81