home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: gnu.misc.discuss
- Path: sparky!uunet!haven.umd.edu!darwin.sura.net!wupost!usc!cs.utexas.edu!hermes.chpc.utexas.edu!jonathan
- From: jonathan@chpc.utexas.edu (Jonathan Thornburg)
- Subject: reviews (was: Re: Unix Review of GCC)
- Message-ID: <1992Jul31.022920.4763@chpc.utexas.edu>
- Summary: "editor@ureview.com" may be a symlink to /dev/null
- Keywords: Unix Review gcc test C compiler E-mail answer respond /dev/null SPEC92
- Sender: jonathan@einstein.ph.utexas.edu (Jonathan Thornburg)
- Organization: U of Texas at Austin / Physics Dept / Center for Relativity
- References: <PATRICK.92Jul29144828@blue.ERC.MsState.Edu> <Bs7n2v.Irv@world.std.com>
- Date: Fri, 31 Jul 92 02:29:20 GMT
- Lines: 41
-
- Like almost everyone else in this thread, I agree that the Unix Review
- C compiler test was botched: the reviewer didn't read the (extensive
- and very well written) gcc documentation, didn't appreciate the value
- of ANSI C, and didn't realise that in practice free software has
- *superb* support from the net, even without Cygnus et al.
-
- However... In article [some article which I lost the attribution to],
- [someone] wrote:
- >The e-mail address for the UNIX REVIEW editors is: editor@ureview.com
-
- About 6 weeks ago I sent a letter to this address, suggesting that they
- should at least report SPEC92int and SPEC92fp separately, rather than
- a lumped "SPECMARK" (89 or 92 unspecified), and preferably report all
- the individual SPEC92 tests. I also suggested that they dump dhrystone,
- and whetstone because of their susceptability to artificial "boosting"
- by compiler tricks that don't help real-world programs. [Cf. Hennessey
- and Patterson "Computer Architecture" chapter 2.]
-
- To this date, I have recieved no reply, not even an acknowledgment
- of receipt. Perhaps "editor@ureview.com" is a symlink to /dev/null?
-
- All this said, on the whole I have found Unix Review to be the best of
- the free-subscription general-unix-workstation magazines (Unix World,
- Unix Review, Unix Today, ...). I find that their reviews and columns
- are moderately detailed and fairly good. Their C/C++, X, and networking
- columns can by no stretch of the imagination be described as for
- "computer illiterates" -- they are often very technical. Their software
- and hardware reviews give a moderate level of technical detail, and
- their reviewers at least know enough not to use awk(1) as a floating
- point benchmark (which is what Unix World once did). Moreover, they
- do occasion say *bad* things about commercial products. (Eg in a recent
- review they had some rather unkind words for SGI's emphatically non-free
- support policies.)
-
- It's not perfect, and the gcc-review botch has certainly lowered my
- rating of it, but for me Unix Review is still worth reading.
-
- - Jonathan Thornburg
- <jonathan@einstein.ph.utexas.edu> or <jonathan@hermes.chpc.utexas.edu>
- University of Texas at Austin / Physics Dept / Center for Relativity
- and (for a few more months) U of British Columbia / {Astronomy,Physics}
-