home *** CD-ROM | disk | FTP | other *** search
-
- I got an email response to this topic that makes some interesting points,
- so I am reposting them and my responses here. (Don't worry, I am not
- going to make a long crusade out of this. 8-) The outdented comments
- are mine.
-
- The assertion that it is "hard to peek inside" is patently false - is:
- gtar -ztvf whatever.tar.Z
- appreciably different that than equivalent content-listing techniques
- in the other archivers you mentioned? Is:
- gtar -zxf whatever.tar.Z some.file
- appreciably different than single-file extraction techniques for the
- other packages?
-
- Don't these commands have to uncompress the entire file to get the listing?
- While that might not always matter, for some files it would matter a lot.
- Also, these don't strike me as particularly user friendly commands.
-
- I never add or replace single files in archives, so cannot comment much
- on this aspect. From experience, I prefer to consider archives as coherent
- snapshots of the state of a system at a given time, and not as active filing
- cabinets.
-
- Oh, I see. Since you don't like to use a certain function, I shouldn't
- either.
-
- Also, compressed tar has the advantage of not burdening its user with
- sub-juvenile compression terminology: "squeezed, crunched, smashed,
- hammered,
- macerated, julienned, ...". I believe that some of these other archivers
- may also use LZW compression similar to that of compress, but I can't be
- sure, because of the spurious terminology.
-
- I'll grant you a 1/3 of a point here. The terminology wears thin. However,
- you refuse here to read the manual. Many of these archivers you demean
- insultingly choose compression techniques on the fly, and each of the
- different terms indicates a different technique. The techniques are
- listed in the RTFM file 8-)
-
- And of course, there is nothing the least bit sub-juvenile about
- naming something recursively 8-) 8-) Not to mention the carefully
- considered and well-reasoned grep, awk, and finger 8-) 8-)
-
- One detractor for uncompressed tar is that it does typically carry a
- significant overhead in fixed-size directory blocks, large parts of
- which usually end up being filled with zero bytes. Compression takes
- care of this quite nicely, though.
-
- One other detractor for tar is that it does not carry a checksum across
- actual file contents - only across header contents. This is not so much
- a problem when archives are exchanged through relatively robust systems and
- media. I cannot recall a single case of file corruption out of the hundreds
- of MB of stuff that I've picked up from uunet via uucp.
-
- Thanks. I hadn't thought of those.
-
- BTW, I noticed a product called Squash! Anybody like to comment on it?
-
- In short, I think it is fine that NeXT mail uses .tar.Z or whatever
- underneath the covers. But ordinary users also need to be able
- to easily create and manipulate compressed archives for backup,
- communications, and data management. .tar.Z is not the way to go
- for them.
-
-
-