Talk:Revision Control System

First release date

Copyright date is a clue, but it is frequently abused by developers who have not actually published a program on the given date. Rather, it is more often used as a guide to when they began work on the program. The rcs manpage does not give an initial release date; a WP:RS addressing the given statement is needed rather than editor's inferences TEDickey (talk) 12:03, 21 January 2012 (UTC)

free (sic)

I found a few authors asserting this. However the closest from Tichy would be this message which quotes from a defunct (and non-archived) website: http://lists.gnu.org/archive/html/help-rcs/2009-08/msg00001.html (and I seem to recall having read this before). The various discussions make it clear that rcs was distributed in source form, free of charge from the outset. Tichy's notice in the rcs version 3 provided in the 4.3BSD contrib area (which also has emacs) made it clear that he was concerned about it being incorporated into some commercial product (for instance, hijacked by AT&T). If there's a verifiable reliable source to the contrary, it might be interesting to discuss. Discussing unpublished, or make-believe sources is not interesting. TEDickey (talk) 00:54, 11 November 2014 (UTC)

As mentioned, in 1986 a person from Techische Universität Berlin tried to convince me to convert from SCCS to RCS but he claimed that he could not give me the source because he had to pay 1000$ to get it - so I would have been forced to use the binaries in the university. For this reason, it is obvious that we need a reliable source for the claim in the WP article as it claims something that is not aligned with my observation. For the same reason, it is important to have a reliable source for the time when the RCS sources have been given to the GNU people as I guess that this is the time when RCS became freely available. Schily (talk) 12:04, 11 November 2014 (UTC)

I've seen no WP:RS which would support your comment. TEDickey (talk) 01:33, 12 November 2014 (UTC)

Given the fact that my information is from a reliable source, it is obvious that we need either a more reliable source that confirms that RCS was freely available and followed the OSS rules or we need to remove the related claim in the WP article. Schily (talk) 10:46, 12 November 2014 (UTC)

Meanwhile I was able to verify that RCS was indeed non-free in it's beginning and made (mostly) free on May 1 1989. Schily (talk) 11:12, 12 November 2014 (UTC)

I trimmed the parts which are due solely to your opinion, and asked for a WP:RS for an interesting point. TEDickey (talk) 10:08, 13 November 2014 (UTC)

By the way, your meat-puppet doesn't follow the notion that the BSD distributions were "free" software. You might suggest some reading material to him. TEDickey (talk) 10:48, 13 November 2014 (UTC)

Please use a more civil discussion style and do not attack other people. The fact that someone is not controlled by you does not verify that this person is controlled by others. Schily (talk) 13:02, 14 November 2014 (UTC)

I did you the favour and linked to the article BSD licenses so you could compare the edits of User:Schily with the text on that page. The license notice on the cited reference reads (line breaks preserved as found in the source):

/* Copyright (C) 1982, 1988, 1989 Walter Tichy
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms are permitted
 * provided that the above copyright notice and this paragraph are
 * duplicated in all such forms and that any documentation,
 * advertising materials, and other materials related to such
 * distribution and use acknowledge that the software was developed
 * by Walter Tichy.
 * THIS SOFTWARE IS PROVIDED ``AS IS AND WITHOUT ANY EXPRESS OR
 * IMPLIED WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED
 * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.
 *
 * Report all problems and direct all questions to:
 *   rcs-bugs@cs.purdue.edu
 *







*/

Notice that this is an instance of the ancestral BSD license referenced in the BSD licenses article I linked you to. It seems like you didn't even consider reading that article before pulling out your “revert” hammer. Please notice that reverting other's revisions without explanation can be seen as a form of edit warring and is a violation of the rules of Wikipedia. Now the important aspect is this part of the license:

Redistribution and use in source and binary forms are permitted provided that the above copyright notice and this paragraph are duplicated in all such forms and that any documentation, advertising materials, and other materials related to such distribution and use acknowledge that the software was developed by Walter Tichy.

This part is commonly referred to as the “advertisement clause” because it requires every redistributor of the software to make advertisement for it in all supplementary materials. Now if you read the BSD licenses article, you would find the following sentence (emphasis mine, references removed):

Two variants of the license, the New BSD License/Modified BSD License (3-clause), and the Simplified BSD License/FreeBSD License (2-clause) have been verified as GPL-compatible free software licenses by the Free Software Foundation, and have been vetted as open source licenses by the Open Source Initiative, while the original, 4-clause license has not been accepted as an open source license and, although the original is considered to be a free software license by the FSF, the FSF does not consider it to be compatible with the GPL due to the advertising clause.

The reason why the original 4-clause license has not been accepted as an open source license is the aforementioned advertisement clause, which is present in the license of RCS at that time. Thus it is completely correct to call it “partially free,” because the license is not a vetted open source license and in fact contains language that restricts the freedom to use the software in obnoxious ways. If you insist on my explanations being wrong, please do so within the next 24 hours after which I will restore the revision by user Schily unless I see further arguments from you. “not found in source” is not a valid argument when the clause referred to takes up more than 50% of the license text and is specifically referred to on other pages as “the advertisement clause.”

Finally, please stop accusing me of meatpuppery. I have been watching this page for a while now and I saw this edit. I have also been watching you for a while now because you seem to exhibit a repeated pattern of behaviour in which you revert edits that don't fit into your narrow world view and refuse to partake in any sort of discussion about the facts in a meaningful way. Just because the source others cite doesn't contain the statement they wrote down verbatim, doesn't mean that the statement isn't backed by the source. Please reconsider your behaviour. --FUZxxl (talk) 11:16, 13 November 2014 (UTC)

Oh. Are you attempting to stalk me? That's what your comment says. Back to the theme - mixing 2014-notion of "BSD" with 1980s "BSD" sounds like you ought to become familiar with the variety of licenses deemed to be "free" before the current era. In particular, applying your comment to the BSD distributions would mark those as "non-free" (rather than writing essays about how you just happened to be here, reading the change comment would help). Revisionism isn't what Wikipedia is supposedly doing. TEDickey (talk) 01:36, 14 November 2014 (UTC)

If you claim that the perception of licenses was different back then, please provide proper citation. Everything else is WP:OR. ~~I'll revert your edit now because you seem to be unable to provide tangible evidence that the properly cited evidence I provide is incorrect.~~ --FUZxxl (talk) 11:22, 17 November 2014 (UTC)

On another thought, the current version of the citation looks fine, so let's leave it this way. --FUZxxl (talk) 11:26, 17 November 2014 (UTC)

GNU maintained?

There's a mailing list via GNU, and it's a GNU project as demonstrated by its presence in http://ftp.gnu.org/gnu/rcs/ (combined with the site description at the top of the filesystem). TEDickey (talk) 00:50, 11 November 2014 (UTC)

The problem with that ftp site is that they frequently remove old versions. Try to e.g. find a gtar version from around 1989 when the FSF acquired PD-tar/SUG-tar from John Gilmore. Do you believe that they acquired RCS in 1993? From my memory, this is aprox. the time when RCS became free. Schily (talk) 14:06, 11 November 2014 (UTC)

Probably not. I have earlier versions, found readily enough through web-searches. You might consider reading the change history. TEDickey (talk) 01:34, 12 November 2014 (UTC)

Well, it may be that there was a GNU variant around 1990, but there does not seem to be a reliable way to prove that. But note that this is a minor problem compared to the fact that RCS was non-free in its beginning. Schily (talk) 11:01, 12 November 2014 (UTC)

Then we'll leave it as noting that your memory has some problem which you are unwilling to resolve. TEDickey (talk) 10:09, 13 November 2014 (UTC)

May 1, 1989

It would be nice if the source had a diff and/or pointed to a bug-tracking system to give the actual change made and the rationale (to put the change into the proper perspective). TEDickey (talk) 01:42, 14 November 2014 (UTC)

It is not my fault that there is no more fine grained history for the RVS sources. For other software that was in use by the BSD people there is a SCCS history that starts around 1979. Maybe this does not apply to RCS as well because RCS was considered to be non-free. Maybe this is a result from missing related repositories from Walter Tichy. It seems that we cannot verify this anymore and need to take the comment in the sources as is. The fact that there is the same comment for all files, verifies it as intentionally. Schily (talk) 13:13, 14 November 2014 (UTC)

One of the problems on relying on a comment as a "source" is that the actual change is not shown. For instance, in RCS 4.3 (July 26, 1990, over a year later), the same $Log$ comment is still the most recent in more than one file (rlog.c, curdir.c for instance) but the license at the top of the file is GPLv1+ (and there is no $Log$ comment which tells about this change). There is by the way an "old" directory on RCS's ftp site which might contain interesting versions (but it's unreadable). TEDickey (talk) 02:11, 14 November 2014 (UTC)

I have no ides why there is no comment in the source that is related to the change to GPL. Maybe this change was not done without official permission? We currently only have the information that RCS was non-free before 1990 and you are of course welcome to present us more detailed information in case it verifiable by a trustworthy source. Schily (talk) 13:13, 14 November 2014 (UTC)

re: done with approval (sic)

That read as an attack on the developers of rcs - you need a specific WP:RS to introduce WP:NPOV material TEDickey (talk) 23:25, 18 November 2014 (UTC)

You need to verify that there was an approval if you like to remove this text. Schily (talk) 11:09, 19 November 2014 (UTC)

WP:BLP is the appropriate guideline here (as well as WP:OR and WP:RS, of course). Making insinuations to the effect that the rcs developers may have ignored Tichy's license for RCS version 3 and simply replaced it with one to their own liking is against policy, particularly because they are easily identified in the change history and credits. TEDickey (talk) 00:24, 20 November 2014 (UTC)

I hope you eventually learn that you cannot require other people to give reliable sources that can be verified on-line from web-servers. If you don't we would need to follow your idiosyncratic definition of a reliable (sic) source and declare RCS non-free as there is no way (sic) to verify whether the conversion to the GPL has been done rightfully. Schily (talk) 18:14, 28 February 2015 (UTC)

The point to take away from this thread is that if you happen to disagree with a comment, you will accuse the other person of lying. Since you are consistent in that regard, there is little more to add. TEDickey (talk) 18:23, 28 February 2015 (UTC)

In other words: you like reliable source only for things you don't like to see in Wikipedia. Schily (talk) 11:26, 2 March 2015 (UTC)

You are repeating yourself, and not providing interesting comments TEDickey (talk) 01:14, 3 March 2015 (UTC)

Come back when you are interested in a fact based discussion and when you don't like to treat yourself different from others. Schily (talk) 10:15, 3 March 2015 (UTC)

Other editors do not spend most of their time attacking other editors. You are not one of those, and not qualified to make comments like that. TEDickey (talk) 09:09, 4 March 2015 (UTC)

So why do you then attack other users? Schily (talk) 10:04, 4 March 2015 (UTC)

I don't. Some vandals make comments on my talk page. TEDickey (talk) 01:30, 5 March 2015 (UTC)

I am talking about your stalking against me. As mentioned before: a further discussion with you makes sense after you decided to be fact based, use reliable sources for your claims and stop your stalking Schily (talk) 10:09, 5 March 2015 (UTC)

I'm not stalking: unless you contrive to change timelines around (as you have done in other recent edits), I've had visible contributions in each of these areas of dispute as long as, and often longer than you have. As usual, my focus is on addressing vandalism. See you around. TEDickey (talk) 00:35, 6 March 2015 (UTC)

Then focus on vandalism and do not approach me unless you are willing to have a fact based discussion and are familiar with the topic. Schily (talk) 09:55, 6 March 2015 (UTC)

Trimmed the licensing section

I just trimmed the licensing section. A lot of it was original research based on primary sources, and of dubious validity. E.g.,

Because there is no date information, it is not possible to determine when it was added

is the argument from silence. Maybe there actually is a source for this somewhere and the editor who added this just didn't find it. QVVERTYVS (hm?) 11:52, 27 June 2015 (UTC)

The ftp links work (for me, at any rate). Perhaps you intended a "better source" tag. Responding to your comment: file-modification times on a remote server are not a good way to determine the age of a source. Rather, the date would be part of the text. (Agreeing, there is a reasonable amount of material which may be of interest, but developing it in Wikipedia is pointless) TEDickey (talk) 13:43, 27 June 2015 (UTC)

The links actually don't work for me. QVVERTYVS (hm?) 17:03, 27 June 2015 (UTC)

They are ftp-URLs. You may be running on a network where those are prohibited. If that is the case, it is a limitation for you, but not for many others. (A better source would be nice, but that aspect is not why I suggested "better source"). TEDickey (talk) 17:25, 27 June 2015 (UTC)

I am on my own network. The FTP server sends empty responses to my browser, curl, wget and ftp(1).

The FreeBSD folks have an archive of the BSD SCCS repo, but it doesn't seem to be easily searchable... QVVERTYVS (hm?) 20:05, 27 June 2015 (UTC)

The conversion from SCCS to svn did cut off meta data and svn in general gives a worse user interface for looking through the differences than you have with SCCS. There is also a general license problem with that code: It cannot be legally distributed in the original file format without permission. Buying a license from Kirk McKusick is currently the only way to get legal access to the full meta data. I negotiated with Kirk, that a conversion to SCCSv6 plus hand crafted fixes for the defective history files (damaged by a disk crash) will be sufficient for a legal free redistribution. So I ask you for being patient until I finished networking support for SCCSv6 to be able to give free legal access to the full meta data for the project. Schily (talk) 14:10, 29 June 2015 (UTC)

this link might be useful. On the same site (ftp.mrynet.com:/operatingsystems/CSRG) there is a set of CD images, but offhand the ftp-links have been the most generally accessible. TEDickey (talk) 20:24, 27 June 2015 (UTC)

github just contains another lossy conversion from the original SCCS archives. Note that the ftp site for the csrg disk images is just another source of the first collection created by Kirk. The images have been made with a very old mkisofs version that creates FS images with bugs. Do not expect to be able to mount these images on a OS that applies heuristics in order to be able to avoid system panics from mounting disk images with junk inside.Today, Linux seems to be the only OS that still does not check ISO images for parameters that cause buffer overflows, so it may be possible to mount these images on Linux. Do not expect to be able to mount them on Solaris or FreeBSD. Also note that these images contain more defective history files than newer disk images. Kirk did press 1000 new and better sets some years ago and it would be nice if people send him money for his new better disks. He did already pay for them... Schily (talk) 15:08, 29 June 2015 (UTC)

Modes of Operation

Has anyone been able to find a source for the comment about many teams not using branches in RCS? I am not able to find anything to back that up, but I may just be searching in the wrong places. ZPDeLong (talk) 21:24, 9 February 2017 (UTC)

Zach, I am not sure about the source, but about the Modes of Operation part, I have to say that basically there is no "commit" concept in RCS, because it working with locking mechanism. I mean even if you read the original article by Walter Tichy http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.151.8450&rep=rep1&type=pdf you won't find a single word of "commit" or "commitment",so the first line I have doubt whether to address no atomic commitment or not, when there is no such a concept of commitment at all. Bardia.zamanian (talk) 05:07, 13 February 2017 (UTC)

I think the intention in that line is to compare atomic commits (commits which may contain changes in multiple files for an "atomic" change) with the RCS conception of a check-in, which is talked about at one of the sources we added. (https://www.ibr.cs.tu-bs.de/kb/revision-control.html). Unfortunately, it seems that the link has gone dead in the last week, and I'm not sure what to do about that. — Preceding unsigned comment added by ZPDeLong (talk • contribs) 17:24, 20 February 2017 (UTC)

ESR's SRC

still used as a back end for ESR's SRC, which is in use today. — Preceding unsigned comment added by 27.33.100.29 (talk) 08:30, 15 March 2018 (UTC)