Talk:E (PC-DOS)

From Wikipedia, the free encyclopedia

This is the talk page for discussing improvements to the E (PC-DOS) article.
Article policies

Contents

[edit] Proposal for deletion

(User:Blaxthos proposed that this article be deleted on 12 Jan 2007. – 2*6 20:19, 12 January 2007 (UTC))

i like this article and it gave me new information of computer history. i don't know anything about E and was 9 when i saw PC for the first time - with MS Windows 3.11. but, deletion of it isn't imho necessary. 62.245.88.247 18:26, 12 January 2007 (UTC)

I have updated the article, adding an infobox and a bit more information. There are also some footnotes now, citing sources besides the PC-DOS manual itself. I'm going to remove Blaxthos' deletion template. His reasons were "Unreferenced, unverifiable, non-notable (generally singular editing programs are not notable/encyclopaedic)". Before he or anyone else re-casts it as deletable consider:

  • PC-DOS 2000 is still in active use in many embedded applications
  • how many articles in Category:Text editors and its subcats you would have to delete for the similar reasons
  • start with this one: Edit (MS-DOS) – 2*6 23:26, 12 January 2007 (UTC)

[edit] ProD

I wouldn't answer a ProD with "well look at what's happening at article X" or "if you delete this article you also have to delete articles Y and Z". This article should be able to stand (or fall) on it's own -- if you have to cite other examples (that may also be in violation of policies/guidelines) then there is no substance to your position. Also, the basis for inclusion is not how many embedded installations of PC-DOS there are; the justification for keeping an article can't be "look at article X". I suggest reading a few of our content and style guidelines:

  • Notability -- is this late-coming DOS editor really notable?
  • Content -- does this meet WikiPedia content policies?
  • Reliable Sources -- do the sources listed meet with the reliable sources guideline?
  • Verifiable -- is the information contained herein verifiable?

I am not going to re-ProD at this time, since the author is obviously actively trying to improve the article. Good luck! /Blaxthos 02:37, 13 January 2007 (UTC)

Nor would I answer it that way. I can see how you might think that's what I meant; with your supposition, "generally singular editing programs are not notable/encyclopaedic", it would follow I meant "look, mine sucks, but so do these". Look at it from my perspective; I believe the E article has value, is interesting, and is notable ... just like the other ones I mentioned. My main argument for notability was in the first bullet; PC-DOS and its E editor are still in use today. The reason I wrote the article was I had just read Edit (MS-DOS). I thought, "that's interesting; I wonder what Wiki has to say about my editor. It turned out nothing, so I be bold and fixed it.
That said, as I researched the topic, I've come to believe the greater story begins in 1984 or before with IBM's personal editor product. It's only mentioned glancingly in the article because I could only find indirect references to it with Google searches. I had an associate who had used "pe" for years, and swore by its power. Eventually my company started using PC-DOS 7 as part of an embedded application, and I began using the E editor myself -- then I realized why he liked it so much. It's a great editor for programmers. So, eventually, if I can back it up with more than my memory, this article will expand into something like Personal Editor family. It would cover the whole story from 1984, to when it became part of OS/2, and eventually to when it became part of PC-DOS. Right now all I have is the PC DOS 7 manual and bits and pieces from the web; the article covers what I could document. – 2*6 07:17, 13 January 2007 (UTC)
PE is a different editor than E. See below. The command invocation to start PE was pe. When pe2 came out, its invocation was pe2. I know. I still have a running copy. I don't know what the invocation for pe3 was, but PE is a distinct editor family from the E editor family programs, which were invoked with the commands e, e2, e3, and epm (again, see below). If you wrote e at the command line, you were probably using E rather than PE. -- Davis
interesting factoid from 1987 Usenet :

I just looked through an IBM document call The Directory (Personally Developed Software for IBM Personal Computers direct from IBM) and on page 44 under the Secion Productivity Family is a listing for the Personal Editor II written by Jim Wylie the same as the original Personal Editor. BEWARE: this booklet is a CANADIAN publication so it may be different elsewhere. Price $74.99 Cdn PArt No. 6276560. — As an avid user of Personal Editor I am extremely happy to see this new product. I cannot wait to get my copy!!

The Canadian version of "The Directory" of personally developed software for IBM Personal Computers should have been identical to the U.S. version. Personal Editor was sold for $75 in the U.S. catalog as well. -- Davis
Also there are third party products which would belong in the expanded article, such as Personal Editor 32. – 2*6 07:59, 13 January 2007 (UTC)
factoid 2, from 1982 :

- Personal Editor 1.0 -- IBM Corporation -- 1982 -- 160K

Now we know the story begins in 1982 with Jim Wylie. :) – 2*6 08:40, 13 January 2007 (UTC)
Jim Wylie is the original programmer for PE. 1982 would be the right year for his first IBM Internal distributions.

I think the point you're missing is that, although we are a community of consensus, Wikipedia has established notability requirements that determine what is appropriate -- Wikipedia is not a repository for every subject someone thinks is "useful" or "interesting". I would suggest reading the guideline referenced above; I think you should also check out what Wikipedia is not. You may find WP:ILIKEIT to be of help as well. These may help clear up why your contribution was ProD'd. /Blaxthos 08:41, 13 January 2007 (UTC)

I would hope that I've addressed the notability requirement to at least some extent below, but a problem with a notability requirement that depends on external publications is that Internal Use programs may not be widely documented in the public domain. This isn't the place to discuss this, but the notability requirement excludes a lot of important and highly influential work from consideration on Wikipedia. This isn't the same as some local rock band putting an entry for themselves here -- Davis

I'm Davis Foulger (referenced in the article). I was surprised (and pleased) to see this article posted. I don't have time now, and will come back later, but if notability is a measure, I would point out that Slickedit, which even today is regarded as one of the very best text editors on the market, is a direct descendent of E (e.g. is an e-family editor). Various versions of E were given that designation before Slickedit, but Slickedit has been recognized as such for well over 15 years now. That may be worth a mention. If there is a desire for a story on Personal Editor, which is itself arguably a descendent of Red (a CP/M editor of note), I can provide a lot of back story on it. Indeed, a widely distributed (published in PC Magazine) macro set I wrote for PE is one of the major starting points for the development of E. E was, however, a different editor from the beginning, with a different configuration language that eventually grew into a full programming language. Its hallmark was extreme speed and very large edit spaces. It was probably the first editor anywhere that fully exploited the memory space available on x286 machines, allowing documents to be edited that were larger (much larger) than 64K without resorting to paging. PE was limited to fairly small files (32K as I recall) and could be very slow with large files.

Hi, Davis! Any info you can provide about personal editor, e, or follow on products will be great to have. Especially important is external documentation, like magazine reviews or discussion about any of these editors in a book. The folks at Wiki like things to be verifiable, and they are big on references. Oh, when you say "E was a different editor from the beginning", it sounds like my "dust off its personal editor product from 1984" characterization in the article significantly understates things.
If you decide you like the Wikipedia, consider registering and working while logged in. One caution, though, it can be addictive. ("Why am I parsing links from Category:Insurance into the subcats?? -- because it's there!") – 2*6 04:32, 14 January 2007 (UTC)

I am registered and have added to several articles. ---- Davis

[edit] Notes for Broader Article

Instead of continuing to add tidbits to the "deletion" section, lets be more hopeful and use this section for info to possibly include in a broader article. First, SlickEdit already has its own article, so we should either consider making that part of the whole Personal Editor shebang, or crosslinking with it.

– 2*6 04:32, 14 January 2007 (UTC)

I love that this article has appeared, but am somewhat uncomfortable playing a self-agrandizing role, especially since the most important player in the development of the E series of editors was Clark Maurer, who is better known for his work on Slickedit. I was his first collaborator on E, and played an evangelical role through the first two (three, if you count the one we never released, ME) versions. In that, I have a deep understanding of the editor's history. Later collaborators included Richard Redpath and others. Eventually, when Clark left IBM to do Slickedit, Genero Cuomo took the lead role. There are four principle versions of the editor and several notable secondary implementations.

  • The first version, E, was released in late 1984. It was not directly programmable, but included a patch capability that allowed function keys to be modified. It is designed to be very fast (much faster than PE) and to handle very large files (which PE could not do). It looks a lot like Personal Editor and shares some of its its function and extended key set. Most of its function is derived, however, from a substantially modified function key set that I created for Personal Editor, including extended draw functionality. More than anything, it looks a bit like Personal Editor, mostly because of the positioning of its command and function key displays.
  • The second version, E2, adds an extensive command language. The language is nothing like that of Personal Editor. Personal Editor implements a tag language of a vaguely GML-like syntax and no real recursivity or other complex programming constructs. E2 implements a REXX-like p-coded language that treats the entire display as a programmable entity. Written, like E, in Turbo Pascal, it blows out Turbo Pascal's limits on program size. The pre-compiler, EI, adds the p-code to the end of the executable code and maintains it is a seperate memory segment. Most users only encounter E2 as an editor, but it can actually be used as application development system that fully controls the display. Indeed, I wrote a user interface modeling language, BBModel, in it, that we used to create a working prototype for an IBM SGML word-processor that was never released. Other systems were modeled in it as well. Another system I wrote with it, E2Survey, shared no p-code with E2 editing systems, but implemented what we would now think of as an XML interpreter of surveys. It is the success of E2 that attracts interest from the developers of OS/2. I'll come back to this below, as Microsoft Edit actually has its design roots in the effort that rises from this interest. Other interesting things that come out of E2. One is one of the very first syntax-directed editing environments that fills in an entire programming or data structure based on the initial entry of a word like "For". Another is an enhanced line drawing function (E2Draw) that allows you to draw fancy boxes on the screen. That function is eventually recreated on IBM typewriters.
  • E3/EOS2 is a vastly more functional port and extension of E2 onto OS/2 and into C. It is implemented for several operating systems, including AIX and OS/2 and is all but identical on all platforms. Memory limits go away entirely as the editor moves from the 286 to 386 platform. I do not recall if E3 (as EOS2) was ever released with the OS/2 product (it might have been), but after we got a simple editor (Edit) in place on OS/2, the focus was on building a fully functioning windowing editor. A version of EOS2 shipped for a while as a programmers editor with IBM compilers.
  • EPM is the fourth version of E. It ports E3 into the OS/2 windowing environment and extends the editor considerably. Most of the work on this is led by Genero Cuomo after Clark Maurer had left IBM. Comments elsewhere on the net (see for instance http://www.texteditors.org/cgi-bin/wiki.pl?IbmEditorFamily) indicate that Slickedit is more similar to E3 than EPM. That only makes sense. Clark recreated E3 when he created Slickedit. While Clark was deeply involved in the design of EPM and its early implementation, it was still in development when he left the company. EPM eventually ships with OS/2 as OS/2 Enhanced Editor, but it is actually the second version (maybe even the third) version of E to ship with OS/2.
  • Slickedit, although it is not an IBM product, should be considered a member of the E editor family. Early versions preserve the E2/E3 editor language (EI) almost intact, and it is still possible to run many E2 commands and EI programs in Slickedit virtually unmodified. Most of the major features that distinguish Slickedit from other programmers editors were features of E3 as well.
  • OS/2 System editor, which doesn't look like E and is much less functional (in order to facilitate detailed system testing), but is designed from E, preserves its fundamental key mappings, and contains some E3 code. I know. I wrote the spec for the editor. Clark wrote more than half of the code (and spent months in Boca doing it). I wrote the documentation that shipped with OS/2.
  • Windows Edit is a virtual clone of that OS/2 System Editor. It works exactly the same way in every detail (menus, the way editing works, etc). I don't know if its a rewrite or not, but any code IBM developed as a part of OS/2 was owned by Microsoft, and it wouldn't be the only major OS/2 code that Microsoft ported directly to Windows (while claiming that we added nothing of value). One piece of code which I know they ported directly was File Manager (which Clark and I also wrote the early versions of, including the original version, FileMan). Its not that there is a relationship between Microsoft Edit and E. Microsoft Edit, like OS/2 System Editor, is a minimally functional version of E. There is no other way to say it.

I might say more, but I think its important to note that there are tens of thousands of references to E family editors on the Internet (see for instance http://www.tangbu.com/preface.htm). Here are some additional quotes from texteditors.org: "JERED is a clone of E for Unix-like operating systems"; "pcedit32 - an 'E' editor clone". Many editors claim to be 'E' family editor clones. It should be noted, moreover, that there were other more or less direct descendent's of E that released in a variety of IBM products.

By the way, there is another place where the E editor family (particularly E2/E3) are notable to things that have been (appropriately) deemed notable on Wikipedia. E2, E3, and EPM implement a REXX-like language. It isn't REXX per se (there are some extensions), but it has been influential on REXX. When the decision was made to implement a portable version of REXX across several platforms, including OS/2 and AS/400 (I was in the room when the decision was made), the port selected was one that had been created by Sam Detweiler. It was decided, at that time, to enhance the speed of that port by adopting some of the compiler/p-code interpreter technology in E3 (Clark was a wizard at creating fast compiler/interpreter code). A substantial portion of the EI language technology was adopted into the portable REXX implementation, which adopted p-code while retaining the late binding that characterizes REXX and makes it so flexible. Sam Detweiler spent several weeks working with Clark on this in Clark's office. I helped arrange the financing for Sam's visit.

I've just modified and extended this article, in part to provide some information that relates to notability, but mostly to correct some substantial errors. -- Davis

Thanks Davis. The article is on the verge of being inverted, wherein the PC-DOS version is just one section among several in the family. A few questions, though. Which family? I.e., should it be considered the "e-editor" family or the "personal editor" family (to bring in editors like pe32)? Or both, with a split? I'm confused about the relationship between OS/2 System Editor and MS-Dos Edit. It's my understanding that Edit was actually QBasic in its editor mode, launched by a stub program called "edit.com" (413 bytes in PC-DOS 5). I've read this from several sources. This means that the "look & feel" might be that of System Editor, the code base would be substantially different. It would be difficult to imagine Microsoft patching in large sections of System Editor code into qbasic.exe, but maybe they did. (I'm not implying that would have been unethical; you already explained they had rights to use IBM's OS/2 code. Just that it seems unlikely.) Also, should we have a citation verifying the very direct relationship between E3 and SlickEdit? I know you know that's what happened :) but it feels like the kind of thing that would benefit from an external source. – 2*6 07:30, 15 January 2007 (UTC)
It's hard for me to go to sources at this point, as I'm on a mini-vacation. Please feel free to rewrite things in whatever manner seems appropriate. A general E family article may be the way to go. It would certainly make more sense than a E (PC-DOS) article. E borrows some look and feel from PE, but they have never really been a part of a singular family of editors. A user might not notice the differences if they didn't look very closely. E owes a debt to PE. They are, however, very different under the covers. PE changes very little from PE to PE2 (most notable for its subdirectory support) to PE3. It uses the same fairly simple tag language in all three versions. E, by contrast, quickly moves from a minimally configurable editor to a very powerful and highly programmable programmers editor. PE could never have handled the demands of syntax directed editing or several of the other features associated with DOS 6.1 E. E enabled such features starting with E2. It and all subsequent versions of E were readily usable as development platforms.
As for the question of Qbasic, Basic had included an editor, in DOS implementations, from DOS 1.0. It was the only editor other than EDLIN built into early DOS versions, but it was always very primitive. I was not aware that BASIC switched to a more usable editor at DOS 5, but I have no reason to doubt your sources. What should be clear is that DOS 5 considerably post-dates OS/2 system editor, and that the look and feel of the DOS 5.0 editor is not similar to OS/2 System Editor (as E is to PE), but virtually identical, right down to the silly line drawn around the text "window" (a feature I wasn't fond of in the initial implementation). It matches the spec I wrote in every detail. I have stated here and my modification of the article that there is no way to know if Edit is the same code or code rewritten from the spec. Microsoft's history suggests that they would claim the latter even if they had done the former, so there really is no way to know anything more than Edit's status as at least a clone of OS/2 System editor.  ::I've waffled a bit on exactly what ships with DOS 6. The feature set suggests that it is a version of E3. Indeed, the screen shot suggests that the version number may be E version 3.13V. There may be members of the old programming team that can confirm this, but my recollection is that the programming team associated with DOS 6 repackaged a lot our old DOS code, including Easymenu, STP, Fileman, and a version of E (probably E3) in creating the DOS Shell environment. I am familiar with the requests for the first three items, because I took the requests and sent out the code. I don't know about E. That request would have been fielded, at that time, by Genero Cuomo (or, perhaps, Brian Lewis). I did get to see the packaged result before the new DOS shipped (the lead programmer in Boca was justifiably proud of it), and made occasional use of it for a while, but I was deep into OS/2 and Web programming by then.
As for the relationship of E3 to Slickedit, the best source for confirmation would be J. Clark Maurer, who is currently listed as the Chief Technical Officer of Slickedit. ----Davis —The preceding unsigned comment was added by 71.233.73.30 (talk) 20:15, 15 January 2007 (UTC).

[edit] Rewrite and Inversion

I'm working on a rewrite and inversion of the current E (PC-DOS) article in my user area. See User:Dozen/E (text editor family). Writing it is kind of strange, with a combination of both too much and too little external information.  :) Feel free to make changes there. My plan is to get a solid version, then copy it back to this article. At that point we can Move (rename) the article to E (text editor family). A redirect would be established for E (PC-DOS) to jump directly to that section in the revised article. – 2*6 19:30, 19 January 2007 (UTC)

[edit] Fair use rationale for Image:E screenshot.png

Image:E screenshot.png is being used on this article. I notice the image page specifies that the image is being used under fair use but there is no explanation or rationale as to why its use in this Wikipedia article constitutes fair use. In addition to the boilerplate fair use template, you must also write out on the image description page a specific explanation or rationale for why using this image in each article is consistent with fair use.

Please go to the image description page and edit it to include a fair use rationale. Using one of the templates at Wikipedia:Fair use rationale guideline is an easy way to insure that your image is in compliance with Wikipedia policy, but remember that you must complete the template. Do not simply insert a blank template on an image page.

If there is other fair use media, consider checking that you have specified the fair use rationale on the other images used on this page. Note that any fair use images uploaded after 4 May, 2006, and lacking such an explanation will be deleted one week after they have been uploaded, as described on criteria for speedy deletion. If you have any questions please ask them at the Media copyright questions page. Thank you.

BetacommandBot 05:28, 26 October 2007 (UTC)

[edit] E3 and ctrl-z 0x1A

It should be noted that E for DOS and OS/2 adds a ctrl-z to the end of the file. In the case of the DOS versions, this can be defeated by adding EOFMODE=NO to the end of the E.INI file.

It should be noted that there are after-market releases of the OS/2 EPM with the control-z feature removed.

--Wendy.krieger (talk) 11:04, 12 February 2008 (UTC)