[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

[PATCH] clarification for ch02 of book

From: Mason Thomas <mlt_svn_at_sbcglobal.net>
Date: 2002-12-19 06:15:16 CET

Hi,
the following is a resend.

The following patch is a suggested change for ch02.xml (chapter 2 of
S:TDG). The patch itself is just suggested text, but the intent, I
think, is important. That intent is to clarify the failings of a
lock-based version control system.

The first change may reflect my misunderstanding of locking versioning
systems. As it stands ("All she can do is read the file...") the text
implies that Jane can read the file but just not obtain a lock for
editing. But what is to stop her from reading the file, waiting till she
can acquire a lock, then editing the (out-of-date) version she has? My
suggested change implies she cannot even read the file, but that may be
wrong. Either way, the current text appears to suggest that locking
doesn't prevent out-of-date problems from occurring.

The remaining changes add mini-headings to the list of problems with the
lock-modify-unlock model. I think this helps to make the failings more
memorable and/or explicit to the reader. I used <emphasis> tags, but
something else might be more appropriate.

-Mason

Index: ch02.xml
===================================================================
--- ch02.xml (revision 4134)
+++ ch02.xml (working copy)
@@ -110,10 +110,10 @@
        making changes to it. Locking a file is a lot like borrowing
        a book from the library; if Joe has locked a file, then Jane
        cannot make any changes to it. If she tries to lock the file,
- the repository will deny the request. All she can do is read
- the file, and wait for Joe to finish his changes and release
- his lock. After Joe unlocks the file, his turn is over, and
- now Jane can take her turn by locking and editing.</para>
+ the repository will deny the request. All she can do is wait
+ for Joe to finish his changes and release his lock. After Joe
+ unlocks the file, his turn is over, and now Jane can take her
+ turn by locking and editing.</para>

      <figure id="svn-ch2-dia3">
        <title>The lock-modify-unlock solution</title>
@@ -125,14 +125,18 @@
        users:</para>

      <itemizedlist>
- <listitem><para>Sometimes Joe will lock a file and then forget
+ <listitem><para><emphasis>Locking may cause administrative
+ difficulties.<emphasis>
+ Sometimes Joe will lock a file and then forget
            about it. Meanwhile, because Jane is still waiting to
            edit the file, her hands are tied. And then Joe goes on
            vacation. Now Jane has to get an administrator to release
            Joe's lock. The situation ends up causing a lot of
            unnecessary delay and wasted time.</para></listitem>

- <listitem><para>What if Joe is editing the beginning of a text
+ <listitem><para><emphasis>Locking may unnecessarily slow
+ file editing.<emphasis>
+ What if Joe is editing the beginning of a text
            file, and Jane simply wants to edit the end of the same
            file? These changes don't overlap at all. They could
            easily edit the file simultaneously, and no great harm
@@ -140,13 +144,15 @@
            together. There's no need for them to take turns in this
            situation.</para></listitem>

- <listitem><para>Pretend that Joe locks and edits file A, while
+ <listitem><para><emphasis>Locking does not prevent all conflicts.
+ <emphasis>
+ Pretend that Joe locks and edits file A, while
            Jane simultaneously locks and edits file B. But suppose
            that A and B depend on one another, and the changes made
            to each are semantically incompatible. Suddenly A and B
            don't work together anymore, and the locking system was
            powerless to prevent it&mdash;yet the locking system somehow
- provided a sense of false security, when it shouldn't
+ provided a false sense of security, when it shouldn't
            have.</para></listitem>
      </itemizedlist>

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Thu Dec 19 06:14:52 2002

This is an archived mail posted to the Subversion Dev mailing list.

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.