On Thu, 17 May 2012, Julian Fagir wrote:
I'm using write(1) on machines where people are logged in for days, weeks, months, running a tmux, as well as their write-sessions. It's annoying to not know the date when somebody talked to you. What do you think about the attached patch (write.1 modification would follow)? It just adds a flag '-t', which prepends the messages being sent with a date (i.e., you and your communication partner have to agree on using that flag to have it useful to yourself).
When you propose a change to the way something works, it's important to get agreement about the usefulness of the change before (or at the same time as) you ask people to review the actual code. At this stage of the process, man page changes, and examples of output before and after the proposed change, would be more useful then code diffs.
In other words, a better question than "what do you think about this patch to the code" would be "what do you think about this description of the changed behaviour". Review of the actual code is also useful, and that can happen at the same time, or later.
--apb (Alan Barrett)