docs/master

diff ch03.roff @ 12:7ca384d68edc

Wrote some text for removal of old code.
author markus schnalke <meillo@marmaro.de>
date Fri, 20 Apr 2012 19:14:56 +0200
parents dc2bfef4cda7
children 55ec590cfa07
line diff
     1.1 --- a/ch03.roff	Fri Apr 20 19:14:30 2012 +0200
     1.2 +++ b/ch03.roff	Fri Apr 20 19:14:56 2012 +0200
     1.3 @@ -4,9 +4,119 @@
     1.4  .P
     1.5  bar
     1.6  
     1.7 -.H1 "Removal of Old Stuff
     1.8 +.H1 "Removal of Code Relicts
     1.9  .P
    1.10 -foo
    1.11 +The code base of mmh originates in the late 70s, had been extensively
    1.12 +worked on in the mid 80s, and had been partly reorganized and extended
    1.13 +in the 90s. Relicts of all those times had gathered in the code base.
    1.14 +My goal was to remove any ancient code parts. One part of the task was
    1.15 +converting obsolete code constructs to standard constructs, the other part
    1.16 +was dropping obsolete functions.
    1.17 +.P
    1.18 +As I'm not even thirty years old and have no more than seven years of
    1.19 +Unix experience, I needed to learn about the history in retroperspective.
    1.20 +Older people likely have used those ancient constructs themself
    1.21 +and have suffered from their incompatiblities and have longed for
    1.22 +standardization. Unfortunately, I have only read that others had done so.
    1.23 +This put me in a much more difficult positions when working on the old
    1.24 +code. I needed to recherche what other would have known by heart from
    1.25 +experience. All my programming experience comes from a time past ANSI C
    1.26 +and past POSIX. Although I knew about the times before, I took the
    1.27 +current state implicitely for granted most of the time.
    1.28 +.P
    1.29 +Being aware of
    1.30 +these facts, I rather let people with more historic experience solve the 
    1.31 +task of converting the ancient code constructs to standardized ones.
    1.32 +Luckily, Lyndon Nerenberg focused on this task at the nmh project.
    1.33 +He converted large parts of the code to POSIX constructs, removing
    1.34 +the conditionals compilation for now standardized features.
    1.35 +I'm thankful for this task being solved. I only pulled the changes into
    1.36 +mmh.
    1.37 +.P
    1.38 +The other task of dropping ancient functionality to remove old code,
    1.39 +I did myself, though. My position to strip mmh to the bare minimum of
    1.40 +frequently used features is much more revolutional than the nmh community
    1.41 +sees it. Without the need to justify my decisions, I was able to quickly
    1.42 +remove code I considered ancient. The need to discuss my decisions with
    1.43 +peers likely would have slowed this process down. Of course, I did research
    1.44 +if a particular feature really should be dropped. Having not had any
    1.45 +contact to this feature within my computer life was a first indicator to
    1.46 +drop it, but I also asked others and searched the literature for modern
    1.47 +usage of the feature. If it appeared to be truly ancient, I dropped it.
    1.48 +The reason for dropping is always part of the commit message in the
    1.49 +version control system. Thus, it is easy for others to check their
    1.50 +view on the topic with mine and possibly to argue for reinclusion.
    1.51 +
    1.52 +.U2 "MMDF maildrop support
    1.53 +.P
    1.54 +I did drop any support for the MMDF maildrop format. This type of format
    1.55 +is conceptionally similar to the mbox format, but uses four bytes with
    1.56 +value 1 (\fL^A^A^A^A\fP) as message delimiter,
    1.57 +instead of the string ``\fLFrom\0\fP''.
    1.58 +Due to the similarity and mbox being the de-facto standard maildrop
    1.59 +format on Unix, but also due to the larger influence of Sendmail than MMDF,
    1.60 +the MMDF maildrop format had vanished.
    1.61 +.P
    1.62 +The simplifications within the code were only moderate. Switches could
    1.63 +be removed from tools like
    1.64 +.L packf ,
    1.65 +which generate packed mailboxes. Only one packed mailbox format remained:
    1.66 +mbox.
    1.67 +The most important changes affect the equally named mail parsing routine in
    1.68 +.L sbr/m_getfld.c .
    1.69 +The direct MMDF code had been removed, but as now only one packed mailbox
    1.70 +format is left, code structure simplifications are likely possible.
    1.71 +The reason why they are still outstanding is the heavily optimized code
    1.72 +of \fLm_getfld()\fP. Changes beyond a small local scope \(en
    1.73 +which restructuring in its core is \(en cause a high risk of damaging
    1.74 +the intricate workings of the optimized code. This problem is know
    1.75 +to the developers of nmh, too. They also avoid touching this minefield
    1.76 +if possible.
    1.77 +
    1.78 +.U2 "UUCP Bang Paths
    1.79 +.P
    1.80 +More questionably than the former topic is the removal of support for the
    1.81 +UUCP bang path address style. However, the user may translate the bang
    1.82 +paths on retrieval to Internet addresses and the other way on posting
    1.83 +messages. The former can be done my an MDA like procmail; the latter
    1.84 +by a sendmail wrapper. This would ensure that any address handling would
    1.85 +work as expected. However, it might just work well without any
    1.86 +such modifications, as mmh does not touch addresses much, in general.
    1.87 +But I can't ensure as I have never used an environment with bang paths.
    1.88 +Also, the behavior might break at any point in further development.
    1.89 +
    1.90 +.U2 "Hardcopy terminal support
    1.91 +.P
    1.92 +More of a funny anecdote is the remaining of a check for printing to a
    1.93 +hardcopy terminal until Spring 2012, when I finally removed it.
    1.94 +I surely would be very happy to see such a terminal in action, maybe
    1.95 +actually being able to work on it, but I fear my chances are null.
    1.96 +.P
    1.97 +The check only prevented a pager to be placed between the outputting
    1.98 +program (\fLmhl\fP) and the terminal. This could have been ensured with
    1.99 +the \fL-nomoreproc\fP at the command line statically, too.
   1.100 +
   1.101 +.U2 "Removed support for header fields
   1.102 +.P
   1.103 +The `Encrypted' header had been introduced by RFC\^822, but already
   1.104 +marked legacy in RFC 2822. It was superseded by FIXME.
   1.105 +Mmh does no more support this header.
   1.106 +.P
   1.107 +`Content-MD5' headers were introduced by RFC\^1864. They provide only
   1.108 +a verification of data corruption during the transfer. By no means can
   1.109 +they ensure verbatim end-to-end delivery of the contents. This is clearly
   1.110 +stated in the RFC. The proper approach to provide verificationability
   1.111 +of content in an end-to-end relationship is the use of digital cryptography
   1.112 +(RFCs FIXME). On the other hand, transfer protocols should ensure the
   1.113 +integrity of the transmission. In combinations these two approaches
   1.114 +make the `Content-MD5' header field useless. In consequence, I removed
   1.115 +the support for it. By this removal, MD5 computation is not needed
   1.116 +anywhere in mmh. Hence, over 500 lines of code were removed by this one
   1.117 +change. Even if the `Content-MD5' header field is useful sometimes,
   1.118 +I value its usefulnes less than the improvement in maintainability, caused
   1.119 +by the removal.
   1.120 +
   1.121 +
   1.122  
   1.123  .H1 "Paths to ...
   1.124  .P