Talk:Termcap

This article is rated C-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects:

Computing: Software Low‑importance This article is within the scope of WikiProject Computing, a collaborative effort to improve the coverage of computers, computing, and information technology on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.ComputingWikipedia:WikiProject ComputingTemplate:WikiProject ComputingComputing Low This article has been rated as Low-importance on the project's importance scale. This article is supported by WikiProject Software (assessed as Low-importance). This article is supported by Computer hardware task force (assessed as Low-importance). Linux Low‑importance This article is within the scope of WikiProject Linux, a collaborative effort to improve the coverage of Linux on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.LinuxWikipedia:WikiProject LinuxTemplate:WikiProject LinuxLinux Low This article has been rated as Low-importance on the project's importance scale.

Err, no. The TERMCAP variable points to a termcap database. Old_Wombat (talk) 09:25, 11 January 2012 (UTC)[reply]

No: TERMCAP can be either. If its value begins with a "/", that is interpreted as the pathname of a termcap database. Otherwise, its value is treated as one or more entries which comprise a termcap database. Using more than one entry is done rarely, but the termcap library code does (attempt to) support this usage. TEDickey (talk) 11:48, 11 January 2012 (UTC)[reply]

OK, that was not the case when I supported termcap in the corporation days (with AT&T SV.2), but if you say so, then OK. Moving on, if it can be both, then the article should say so. Do you want to put this in or do you want me to do it?

I think this highlights a wider problem, that whilst there is only one Windows XP (for example), there are as many flavours of Unix as you care to name, and they all have their slight differences, and there appears to be no consistent way of handling this within the various Unix articles in Wikipedia. Old_Wombat (talk) 07:55, 12 January 2012 (UTC)[reply]

TERMCAP either contains a pathname of a termcap database file or one single TERMCAP entry. If $TERM and $TERMCAP disaggree, the default termcap database file is read instead. Modern termcap implementations in addition check $TERMPATH which may contain a list of termcap database file names that are checked in order until a match occurs. The latter is usually used to first read more up to date entries from a file .termcap in the users home directory - BTW: while reading ~/.termcap works with the ascii termcap entries, it does not work with terminfo as terminfo uses a binary database that is not identical across different platforms. --Schily (talk) 23:10, 28 January 2012 (UTC)[reply]

hmm - I oversimplified (and misrecalled part of) this. Reading the BSD4.2 code now (AT&T SV.2 introduced terminfo and appears from previous reading to have omitted the termcap library) the BSD code uses the $TERMCAP variable as the actual value if $TERM is not set, or if $TERM happens to match the second parameter given to tgetent. But if $TERM differs from that, then it ignores $TERMCAP and uses the system's file (to find the entry given by the second parameter). In the case where it uses $TERMCAP as the data, it tries to find entries to resolve "tc=" (include) that may occur on the end of the entry. Those shouldn't occur (since it's starting by looking at the entire environment variable, and "tc=" is on the end). That will recur through tgetent again, and take a different route (using a different second parameter to tgetent) to find the included source, from the system's termcap file. (It might help to read the source, so that my explanation would be clearer) TEDickey (talk) 11:19, 12 January 2012 (UTC)[reply]

There are no WP:RS mentioned in any of Schily's edits. TEDickey (talk) 23:16, 30 January 2012 (UTC)[reply]

That's a lot of words to describe a rarely used feature, and actually is longer and provides more (unsourced of course) information than the given source (from 1992, which states that the Hazeltine terminals were very rare then. TEDickey (talk) 00:41, 21 May 2014 (UTC)[reply]

• Yes, they were rare in 1992. However, they were quite common in the late 1970s when the termcap database was first devised. The reason I didn't use an older source was because I could not find one online. Dave Cornutt (talk) 20:19, 21 May 2014 (UTC)[reply]

Termcap is just a database that allows to retrieve values for 2-char keywords. While the termcap capabilities were documented in termcap(5), they were not implemented in the termcap code but in vi(1) and libcurses(3). The latter was created from modifying ripped of code from vi(1). The file cr_tty.c that extracts the value of the "hz" capability is e.g. derived from ex_tty.c from vi. If you are looking for implementations of specific features related to specific capabilities, you need to check the code that uses termcap. So I believe that this feature should not be documented here but rather in vi or libcurses. Schily (talk) 09:39, 22 May 2014 (UTC)[reply]

There were a lot of "quite common" terminals, but a reliable source is needed to both qualify and quantify that comment TEDickey (talk) 21:16, 21 May 2014 (UTC)[reply]

• • The Wikipedia article for the IBM 3101 mentions the 1500 as one of the terminals that the 3101 emulated, along with common terminals like the VT100 and the ADDS Viewpoint. They would not have bothered with the 1500 if it were not a common model at some point. I was there in the late '70s and I can tell you that Hazeltine terminals were common. Unfortunately, there does not appear to be any sales numbers available (this is true for practically all of the 1970s-80s terminals) and the Hazeltine Corporation is long gone. Look at it this way: the 1500 was common enough that the termcap developers felt it necessary to put a special hack in to accommodate it, and that hack survives in the specifications to this day. And it is noteworthy as being one of the non-standard types of things that had to be dealt with back in the day; that lack of standardization is why termcap was created in the first place. — Preceding unsigned comment added by Dave Cornutt (talk • contribs) 02:50, 22 May 2014 (UTC)[reply]

Wikipedia is not a reliable source. The source you provided indicates that the feature can be considered obsolete, was rarely used in 1992. So far, you have provided no source demonstrating that it was common, only a comment that it was a selection in a terminal which was designed to provide several variations. Personal experience ("I was present", etc.), is not a reliable source, either. TEDickey (talk) 07:55, 22 May 2014 (UTC)[reply]

• So even though I have now provided four references, you still aren't going to accept it. Evidently there is no source that you will accept. I guess if it existed prior ro 1992 then it is simiply not meaningful to you. You can take that Millennial view if you want, but it's contrary to reality. I guess you're going to delete it. This will be the thrid time that content that I worked on has been deleted because it didn't meet some Very Important Wikipedia Person's personal critieria. I'm done with Wikipedia and I will be deleting all of my contributions. — Preceding unsigned comment added by Dave Cornutt (talk • contribs) 17:27, 22 May 2014 (UTC)[reply]

While terminfo libraries have a termcap-compatibility interface, it's still not termcap. In particular, there's no termcap file, and no relationship beyond the entrypoint names with the BSD termcap. TEDickey (talk) 18:37, 24 April 2020 (UTC)[reply]

No one provided a source during the past year, so I removed the unsupported comment about Multics TEDickey (talk) 20:26, 26 April 2021 (UTC)[reply]

The only places which I've found mentioning a connection between the (MIT-lab) ITS and Berkeley termcap are copies of this topic. Other, closer, institutions (such as Stanford's AI lab) had their own systems for managing terminals. Without a relevant source by a relevant Berkeley developer, there's no reason to keep Crispin's MIT-promoting comment in this topic. TEDickey (talk) 22:26, 26 April 2021 (UTC)[reply]

@TEDickey: There is indication that Eric Allman was involved in development of termcap forerunner called ttycap. I suggest we put that info into the History subsection. You have any objections? See source. AXONOV (talk) ⚑ 15:47, 25 April 2025 (UTC)[reply]

I think you need more information than that (there are very few reliable sources past the comment in the ncurses manpage for reset) TEDickey (talk) 18:50, 25 April 2025 (UTC)[reply]

By the way - your use of "GNU Termcap" doesn't fit into the chronology (it came 10-15 years after the events you're referring to, and never had much impact on what was actually used). TEDickey (talk) 18:55, 25 April 2025 (UTC)[reply]

History subseciton says Bill Joy wrote the first termcap library in 1978 and the reset manpage says that ttycap was written in late 1977. What's the gap in 10-15 years are you talking about? AXONOV (talk) ⚑ 19:02, 25 April 2025 (UTC)[reply]

The source you provided for GNU Termcap is dated 1998. There's another source in this topic which says 1992. Getting a reliable source for its relevance to anything other than bash and emacs is likely to be problematic. See this, in particular this, as well as the note at the end of the FAQ for more context. If you need further explanation, the bug-ncurses mailing list is the appropriate place to discuss it (talk pages are not reliable sources TEDickey (talk) 19:54, 25 April 2025 (UTC)[reply]