Talk:Strace

Unnamed section

Strace was written by

• Paul Kranenburg
• Branko Lankester
• Rick Sladkey
• Wichert Akkerman

It uses the proc filesystem on SVR4 and FreeBSD and the ptrace() system call on other systems.

It runs on

• SunOS 4.x
• Linux
• SVR4 (System V release 4)
• Solaris 2.x
• Irix 5.x
• And appearantly on FreeBSD?

So...

Why is all of the above here rather than in the article itself?--NapoliRoma 19:56, 28 January 2007 (UTC)

Usage

Let's extend article with usage, e.g. I'm currently found for myself that strace able to intercept process by `pid`. See here http://chadfowler.com/blog/2014/01/26/the-magic-of-strace/

List of different hooks, tactics and methods will be great stuff. 128.73.171.56 (talk) 09:56, 12 August 2014 (UTC)

While it sounds as a great addition to the article, unfortunately such content would collide with the WP:NOTMANUAL rule. — Dsimic (talk | contribs) 10:10, 12 August 2014 (UTC)

I'm talking about different hooks, that are not mentioned in man pages. Something great!

128.73.123.228 (talk) 13:57, 12 August 2014 (UTC)

Hm, attaching to an already running process is documented in the strace man page; here's an excerpt from it:

-p pid

Attach to the process with the process ID pid and begin tracing. The trace may be terminated at any time by a keyboard interrupt signal ( CTRL -C). strace will respond by detaching itself from the traced process(es) leaving it (them) to continue running. Multiple -p options can be used to attach to up to 32 processes in addition to command (which is optional if at least one -p option is given).

Any chances, please, for more details on which undocumented hooks you're referring to? — Dsimic (talk | contribs) 22:12, 12 August 2014 (UTC)

Family

Looks like strace has great family.

E.g. See,

• http://xtrace.alioth.debian.org/
• http://www.chiark.greenend.org.uk/~sgtatham/xtruss/
• http://linux.maruhn.com/sec/xmon.html

There is no articles for all this tools, however its shall be mentioned somewhere.

128.73.123.228 (talk) 13:56, 12 August 2014 (UTC)

Comment about ptrace

The Linux sources and release notes make it apparent that Branko provided a fix for an existing implementation of ptrace (about 6-8 months before), but the nature of the fix and how it fits into the story is too technical (and since no relevant reliable source on the issue exists), would be original research. TEDickey (talk) 00:37, 15 January 2015 (UTC)

It would help a lot if you did take a second and read the references you provided. If you did, you could of course know that you needed to patch the linux kernel in order to make strace working. On the other side it would help if you did not revert text that you did not read. I did never claim that Sun provided the patch for the linux kernel, but simple porting could not make strace work on linux and this is the important information that belongs into the article to get encyclopedic relevance for the related part of the text. Schily (talk) 13:17, 15 January 2015 (UTC)

I read those sources, and more. The manpage said that Branko did some kernel development to port strace. However, none of the sources which I have found point out who implemented the PTRACE_ATTACH, etc., calls. It could be Branko, could be Linus, or (less likely), the person whose name is at the top of the file. Branko was credited with fixing a bug; the actual change is more than a bug-fix. You did read the source code, of course. Otherwise your comments are pointless. Furthermore, your comments are inaccurate, since you insist on wording the topic as "needed prtace enhancements from SunOS that allow to trace system calls". Perhaps the reader is expected to assume that the broken grammar and spelling allow some freedom to additionally infer whatever meaning you intended. As it is, the sentence lacks the normal qualifications such as "adapted from", "inspired by", "imitating", pointedly omits any mention of the actual people who did the work, and leaves SunOS as the assumed actor in the statement. That's misleading at best. TEDickey (talk) 00:24, 16 January 2015 (UTC)

Wenn Sie sich über mein Englisch beschweren, dann sollten wir am besten die Diskusion in Deutsch weiterführen, dann bin ich jedenfalls sicher, daß ich keine Grammatikalischen Fehler produziere, oder mich sonstwie mißverständlich ausdrücke... Ich befürchte allerdings, daß der Google Übersetzer ein miserables Englisch aus meinem Deutschen Text produziert und Sie damit erst Recht Probleme haben werden mich zu verstehen.

Es täte Ihnen übrigens gut, wenn Sie mal versuchen würden konsruktiv bei Wikipedia mitzuarbeiten anstatt immer nur Texte Anderer zu löschen. Zur Zeit entsteht jedenfalls der Eindruck, daß Sie kein Interesse daran haben, sich mit den Beiträgen Anderer ernsthaft zu beschäftigen.

Zurück zum Inhaltlichen: Nur weil Sie in meinen Text Dinge hineininterpretieren, die dort gar nicht stehen, ist mein Text nicht falsch - im Gegenteil. Fakt ist jedenfalls, daß mit SunOS-4.0 das ptrace-Interface erweitert wurde und damit erst das trace-Programm möglich wurde. Fakt ist auch, daß wegen der fehlenden Eigenschaften im Linux-Kern eine einfache Portierung des Programms strace nicht ausreichte um es auf Linux zum Laufen zu bringen. Wenn aber jemand in einem enzyklopädischen Text von einer Portierung spricht und die darüberhinaus notwendigen Kernel-Erweiterungen verschweigt, dann ist dies ein verfälschender Text, der so in Wikipedia wegen der Neutralitätspflicht nicht geduldet werden kann. Wenn Sie der Meinung sind, daß meine Formulierungen verbesserungsfähig sind, dann zeigen Sie doch bitte mal daß Sie konstruktiv bei Wikipedia mitarbeiten wollen und verbessern den Text. Das würde mir und Ihnen viel Zeit ersparen und hilft Wikipedia. Schily (talk) 11:04, 16 January 2015 (UTC)

For what it's worth, Bing translator uses English idioms better than your own edits. Perhaps you should simply use that. TEDickey (talk) 01:20, 17 January 2015 (UTC)

Back to the discussion: what's needed for topic development on Wikipedia is not a series of unsourced comments, but something that extracts from published documents and commentary. Comments which tell more than the sources are not an improvement. In particular, the dates at which Sun introduced PTRACE_ATTACH, when it may have been introduced into other systems, the actual people and changes made are all difficult to establish. Given that, I haven't found any suitable sources for expanding on this Wikipedia topic. There's material for "some time" writing a research article on the early history of strace, to point out the details which are and are not known exactly, which could be used "later" for sourcing some interesting comments here, but none of that exists. TEDickey (talk) 14:10, 17 January 2015 (UTC)

Went ahead and removed an unsourced statement. Unless we find some sources to support it, it's better not to include such information. As described elsewhere, Wikipedia works by summing up reliable sources, and personal memories unfortunately can't be used for that purpose. — Dsimic (talk | contribs) 13:52, 21 January 2015 (UTC)

@Dsimic: something that does not match my personal memory is a non-reliable source - and just a note: you recently removed a statement where we have a reliable source. Nothing is better than a primary source like the readme with compile instructions in the sources for the project that is described in the article.

@Tedickey: if you believe that the bing translator creates useful English from my German text, then you seem to have a really strange conception of good English. Bing just creates funny expressions for different parts of my text. Schily (talk) 14:25, 21 January 2015 (UTC)

Could you, please, specify which source you're referring to, regarding the statement I've removed earlier? — Dsimic (talk | contribs) 15:18, 21 January 2015 (UTC)