Jump to content

Talk:Gerber format/Archive 1

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
Archive 1

Untitled

I added the note about Gerber Scientific, because the first thing I wondered when I ran across the term, "Gerber File," was why it was called that. Actually, it was "Gerber Scientific Instruments", a division of "Gerber Scientific, Incorporated". The parent company still exists as of June 2005. I worked at the subsidiary responsible for Gerber format in the mid to late eighties. -- DanM

Hi DanM, do you know why there were so confusing things in aperture numbering. As I remember the counting was non-linear, instead of apertures 21 and 22 the numbers 71 and 72 were inserted or something like that? Do you know anything about that? Do you know a very common RS274D aperture file format? Best regards -- Efficiency

No, sorry, Efficiency. My involvement with this format was minimal. -- DanM

My first time, so I will just post comments, not make any changes. The apertures were D10-19, D70-71, D20-29, D72-D73. I know why, but I can not remember at this moment. When I remember at 3am some morning, I will post the reason.Dcshank (talk) 06:33, 10 May 2009 (UTC)

Merging with G-code?

I do not think that Gerber File should be merged with G-code, as "Gerber format" is a very important format on it's own. And anyway producing PCBs usually has nothing to do with milling 3D shapes. I will extinguish the proposal for merging. -- Efficiency

G-code seems to be description of languge or format of data inside Gerber File and they both be in one-to-one relation. From these articles and by googling I understand that both these fields of industry (PCB design in Gerber File, machining in G-code) use the same format. I think that using the same thing for two different goals does not grant right for two articles. Instead we shloud describe the format itself and its usage in sections rather than duplicating them. --Alvin-cs 3 July 2005 14:53 (UTC)
G-Code normally is associated with milling, Gerber format is for drawing PCBs. Gerber may be similar to G-Code, but it is shurely not the same. It's like humans and monkeys. Both have the same roots, but are different enough. I'm having to do with both - DIN/ISO G-Code and Gerber and I think Gerber is really important enough to have an own entry. Else you would try to begrudge the space-shuttle an own article as it is just a transportation medium ;-) -- Best wishes, Efficiency, 3 July 2005

Internally, how different are they? I suspect they are really the same format internally, just from the nomenclature:

  • EIA RS274X for Gerber files to make printed circuit boards
  • EIA RS274D standard for G code programming to make metal parts

Whether or not my suspicions are correct, the article should explain the difference -- or admit that the format is identical, it's just used in a different way.

If these 2 formats, in fact, turn out to be identical, then we still might choose to keep Gerber File and G-code as two separate articles, because they are used in 2 different ways. There is a precedent -- fiddle and violin are used in 2 different ways, and so have 2 separate Wikipedia articles, even though they are exactly the same instrument. --75.48.165.135 07:20, 25 September 2007 (UTC)

I added some explanations in the first paragraph. Clearly Gerber is derived from "EIA RS274D", but on the other hand "RS274X" is/was not/never an EIA standard. The EIA web site does no longer talk on these things at all. I'm not sure about the correct writing. Is is "EIA RS274D" or "EIA RS 274D" or "EIA RS 274-D"? My Gerber format handbook from 1993 write it as "EIA RS-274-D". Lothartklein 20:22, 8 October 2007 (UTC)

Actually I believe Gerber was a defacto standard prior to EIA legitimizing it. Some G-codes are similar or identical between photoplotters and most NC-contol equipement, i.e. G03 circular interpolation, but each manufacturer may ignore EIA 274D as he pleases, so even the "G-Code" is not a universal standard. IMHO G-code should go away.Dcshank (talk) 06:33, 10 May 2009 (UTC)

The link from Gerber to G-Code is confusing without any comment as to why they are linked. Perhaps something as minor as saying "G-Code is another example of a numerical control format, and the two formats have much in common" would clear up the confusion. -- James Newton 68.111.175.222 (talk) 20:13, 27 October 2010 (UTC)


HIstory

I have added my comments to other sections, if there are not objections, I will tweak some of those at a reasonable future date.

I have a lot of problems with the History section, many dates given as exact and therefore assumed to be accurate may need to be pushed back or forwards. 1) My Gerber format Guide is dated August 16, 1994 which is later than the date posited as 1993 for the last edition. 2) I was using a VAPE (variable aperture head) on a Gerber Model 32 in 1984 on a D-116 computer which is a circa 1971 computer. This head had to be around before 1986. 3) The Gerber format was the defacto standard long before 1990. It was the standard used on many other machines when I first entered the industry in 1977.

I will do the necessary research and correct the history where necessary in the near future. —Preceding unsigned comment added by Dcshank (talk • contribs) 06:56, 10 May 2009 (UTC)

Aperture

"The "aperture files" describe the shape of the apertures (image elements) used."

What is "image elements"? How should one understand the term "aperture" as used several times in this article? Thanks, --Abdull (talk) 09:47, 17 March 2010 (UTC)

Criticisms and suggestion

This article is very poor as it stands. The first two sentences contain the most useful information. The entire "usage" section is whiny and subjective: "Unfortunately", "The most practical way". "the file names are often useless", and shows no examples.

The images of "Gerber layers" tell little about the Gerber file itself. Annotated portions of Gerber files, would be much more helpful, especially if they could be related to the images. If I were knowledgeable about Gerber files I would correct rather than criticize, but in that case I would not have come to this page looking for information. I'm appealing to a knowledgeable soul to notice and improve this, I really don't want to merely complain. —Preceding unsigned comment added by 96.50.106.190 (talk) 03:51, 2 January 2011 (UTC)

Capitalisation

MOS:CAP reserves capitals mostly for proper names. The word format is not a proper name and should not be capitalised. Please stop reverting this. SpinningSpark 22:35, 15 February 2011 (UTC)

Well, that a matter of opinion. I think that The Gerber Format is a specific noun and should be capitalized in entirety. I think, otherwise, it would be Gerber's format. — Dgtsyb (talk) 23:58, 15 February 2011 (UTC)
Can you support your opinion with reference to Wikipedia guidelines? Compare Newick format, Nikon DX format, Stockholm format. SpinningSpark 19:56, 16 February 2011 (UTC)
Indeed, format is not a proper name, and normally is not capitalized. However, by common English spelling rule, it is capitalized in a title, as well as first word of a sentence. This is the point here. See e.g. the article on PDF. User:Karloman2 9:35, 17 February 2011 (UTC)
MOS:CAP — Dgtsyb (talk) 15:05, 17 February 2011 (UTC)
Ok, but what exactly in MOS:CAP do you think supports your position? MOS:CAP is against capitalisation - Wikipedia's house style avoids unnecessary capitalization; most capitalization is for proper names, acronyms, and initialisms and WP:TITLE says Use lower case, except for proper names: The initial letter of a title is almost always capitalized; subsequent words in a title are not, unless they are part of a proper name SpinningSpark 09:46, 5 March 2011 (UTC)

Requested move

The following discussion is an archived discussion of a requested move. Please do not modify it. Subsequent comments should be made in a new section on the talk page. No further edits should be made to this section.

The result of the move request was: page moved. Vegaswikian (talk) 00:43, 18 March 2011 (UTC)


Gerber Format → Gerber format — per MOS:CAP and WP:TITLE, format is not a proper name. SpinningSpark 17:56, 5 March 2011 (UTC)

The above discussion is preserved as an archive of a requested move. Please do not modify it. Subsequent comments should be made in a new section on this talk page. No further edits should be made to this section.

Other extensions

I believe .PHO is another extension used by Gerber files —Preceding unsigned comment added by 66.48.8.10 (talk • contribs) .PHO is also used, but somtimes it can be only .txt or even without any extentions

Gerber file extensions are usually a product of the EDA tool used to create them. For instance the list example in the main article is used by Altium/Protel. Typically, .txt is used to distinguish an ASCII NC drill file vs a binary encoded drill file. Most short turn PCB fabricators prefer ASCII drill files. Wamnet

How many extensions you want to list here? Every CAD-System is using its own extensions and users invent further. There are no common extensions to list here. I would also remove the incomplete list of companies/systems here. There are many more important ones. Lothartklein 20:22, 8 October 2007 (UTC)

I agree. Remove the extensions and the vendors, or list them all.Dcshank (talk) 06:33, 10 May 2009 (UTC)

Is it just me, or doesn't this article feel a little angry? It seems more interested in criticizing RS274D (it's not obsolete until everyone stops using it!) and defending RS274X (how is not being able to tell the layers apart not a problem? it allows mistakes that'll ruin your boards just as badly as messing up your aperture table does with 274D) than in being really informative about either. 71.234.76.253 (talk) 06:02, 5 March 2011 (UTC)

D is obsolete because it caters to an obsolete technology - vector photoplotters, which are no longer in use. It is obsolete because Ucamco has replaced it with X, and declared D obsolete. I think something can be called (technologically) obsolete, even if it is still used by some. Indeed, an RS-274X file describes a single PCB layer. This, it does very well. However, it does not describe the stackup, netlist, etc. It does not. Indeed, to reconstruct the whole PCB in CAM, one must assign the layer functions manually. If the CAD vendors would give clear file names that unequivocally specify the function file, this would not be such big a deal. However, all to often, they do not. RD-274-D has the same limitation. On top of that, it is not even a complete layer description, you need to link it to a wheel file or enter the apertures manually, and this for each layer. Furthermore with D you need painting to describe areas. Therefore, you are much better of with RS-274X files than RS-274-D files. PCB fabricators have a right to be angry to still receive D files instead of X. See also the external link "Improving CAD to CAM Data Transfer; A Practical Approach"--Karloman2 (talk) 23:50, 4 October 2011 (UTC)
I'm getting the sense that WP:RS is getting buried here. Glrx (talk) 21:42, 6 October 2011 (UTC)
Missing Aperture Information/File for 274-D Artwork Files is listed as a critical package completeness issue in IPC-2425 PWB Fabrication Data Quality Rating System issued February 1999. But, aside from that, what Karloman2 lists above is simply well-known in the industry. — Dgtsyb (talk) 22:02, 6 October 2011 (UTC)
Comments on RS-274D:
  • The form is archaic.
  • Although some fabricators say that they will accept RS-274D Gerbers, they all say that they prefer RS-274X.
  • The actual Gerber RS-274D specifications are lost to the web. Although hardcopies might exist somewhere, the old Gerber format manuals are either lost to the web or were maybe never generated in electronic format. Yes, it is that old. The wayback machine doesn't go earlier than RS-247X Rev. B.
  • Most CAM systems that say that they support RS-274D Gerbers probably don't. There are archaic commands in early RS-274D, such as parabolic and cubic interpolations that are probably not supported on current tools correctly.
  • The format was not so much a standard. Many photoplotters implemented the format, but, of course, each modified and embellished the language for their particular plotter (even by model). Implementations also depended on the particular brand and model of CNC controller used in the photoplotter.
  • Many of the commands were purposeful only for controlling antiquated photoplotters, and placed restrictions on coding that are no longer pertinent to photoplotters made in the last two decades or so (such as single-quadrant vs. multiple quadrant interpolation constraints of photoplotter x-y tables).
  • The format cannot be read and verified properly by various Gerber viewers.
  • RS-274D is not worth the effort to support because the use of RS-274X is so much more widespread.
Just some comments. — Dgtsyb (talk) 22:16, 6 October 2011 (UTC)

Ordering and duplicate content

The section ordering seems odd to me. Would a chrono order work better (RS-274-D then RS-274X). Also there's a lot of repetition in the article. (The first para of "History" rehashes the -D story once more.) I'm happy to try to clean it up but I don't want to invest just to see the effort reverted back. Woz2 (talk) 16:32, 6 October 2011 (UTC)

Hi Woz, I agree with you there are still repetitions in this article. However, I assume most readers will not study the whole article, but scan it quickly. I also assume most will skip the history section - readers probably are industry practitioners, with a practical need, and not industrial archeologist; (this is why I did not bother with the history section until somebody corrected a date, and then I made the style more consistent;) the article should make sense without the history section. Gerber today is X, or at least should be. Therefore X should come first. Unfortunately, D cannot be banned to the history section because it is still used sometimes; for today's practitioners D is so weird that I felt it was needed to explain why it is as it is; just nakedly explaining the format would be baffling; hence the reference to vector plotters. So I think the order is fine, and the repetitions are there for a reason. I have more of a gripe with the Alternatives section. This article should explain Gerber. Point. Of course, related formats are important to position Gerber. I would like to delete the Alternatives section, but this would be overdoing it. However, it should be as short at possible, and give a reader pointers to further information. Now, of course, you have the format wars in the PCB industry. An interesting topic, which is wider than any single format. The Gerber article is not the place to discuss this. The problem is that there is no other obvious place. It is not truly a topic for a Wikipedia article, more for a forum discussion. If would be ideal to find a good place to point to. Maybe one of these references in your - you are the main contributor - ODB++ article. OK, well, this is why I structured the article as I did. Hopefully it all makes some sense.--Karloman2 (talk) 21:02, 6 October 2011 (UTC)


OK How about making the alternates a See also section? Also cutting repetitive bits out the History and make it a bulleted time line summary? Woz2 (talk) 22:34, 6 October 2011 (UTC)

Request for third opinion on three topics

  1. I think that readers will want to know that the word Gerber itself isn't an acronym nor engineering jargon, but derives from the family name of a person. Karloman2 differs believing this info defocuses the article. As of this writing, the article reflects my preference.
  2. I think a chronological ordering of the section RS-274-D ==> RS-274X ==> Supplementing... is more logical. Karloman2 differs believing RS-274X should be presented first. As of this writing, the article reflects Karloman2's preference.
  3. I think there is too much repetition in the article especially in the RS-274-D section "The obsolete RS-274-D Gerber format..." "RS-274-D is obsolete." That section also has a sequence of five paragraphs each one of which begins with "RS-274-D..." As of this writing, the article reflects Karloman2's preference.

Thanks! Karloman2's comments are on my talk page: User talk:Woz2#Gerber format Woz2 (talk) 13:40, 1 November 2011 (UTC)

I see Woz2's points, but I think the reasons for doing it differently are stronger. Of course, I may be subjective. Let me argue my case here.

  1. I agree that is should be explained in the top section why this format is called Gerber. However, I looked at my old Gerber descriptions, and they all refer to Gerber Systems Corp., but Mr. Gerber was not mentioned. The format was called after the company, and not the person. Of course, the company was called after Mr. Gerber (who was called after his father and so on) but this is indirect. So I would only directly mention the company, and link to the Gerber company name, which links to a full article on Mr. Gerber. One can consider to add Mr. Gerber, and more, in the history section, for those that are interested in it.
  2. The historic order is indeed logical in the history section. However, most readers do not care about the history, but want to know about Gerber as it is today. For current practitioners, X is more important. See e.g. the recent open source Gerber viewer gerbv, which will only implement X. One can understand the X format without having to bother with wheel files etc. Some people will not need to know about D.
  3. I used the uniform style following Strunk&White's recommendation to "Express co-ordinate ideas in similar form". I don't think merging these paragraphs will not improve the style of the article. Each paragraph expresses one subject. I agree, there was to much repetition in older versions, but a.o. Woz2 and others rightly pruned it. I do not think there is much repetition if you skip the history section. Granted, obsolete is already mentioned in the the introductory sentence, but it is worth to repeat the short sentence "RS-274-D is obsolete." because it is such an important fact. I think it important the facts in this section are stated very explicitly. People that do not know the details of the Gerber format but use it should know these facts. This is important. Now people sometimes use D where they should use X because they do not know these facts. This has a negative effect on the PCB industry. D should normally only be used in legacy applications that do now support X. Maybe this should be stated in the article, but I felt to let the facts speak for themselves. But one should be able to understand X without reading all this, and this is the reason why I put D after X.

--Karloman2 (talk) 15:18, 1 November 2011 (UTC)
It is maybe not so important if Mr. Gerber is mentioned or not. However, it is very important that the position of D vs X is clearly explained, as it is now. IMHO this is the main value of this entry. 76.102.136.152 (talk) 16:32, 1 November 2011 (UTC)Luc Maesen

Here is my opinion on the topics:

  1. Gerber format is indeed named after the company, particularly as it was renamed Barco format and Ucamco format as the company changed hands, although noone settled for the renaming and started using RS-274X instead. Gerber format files, however, are still called "Gerber files" or just "Gerbers". There is however, not an article on Gerber Systems Corp. Maybe there should be one as the company is notable particularly for the Gerber format. I don't see a problem with mentioning in the article that Gerber Systems Corp. was named after its founder, particularly if there is no separate article on Mr. Gerber or Gerber Systems Corp.
  2. Personally I prefer RS-247X section first. This is the main thrust of the article and deserves prominence. I think that the RS-247-D section (which is a misnomer because Gerber format was a subset of RS-247-D) should be moved into a new History section along with Timeline as the old format is historical (but maybe not obsolete). I think that the Supplementing RS-274X with other data should be absorbed at the end of the Usage of the Gerber format section.
  3. I agree that the RS-274-D section needs a lot of work. This was only called RS-274D (no dash) after RS-274X arrived. It was never called Gerber RS-274-D (with the dash) as the section now uses. This section has always been problematic in this article because it has not been written with a neutral point of view. As far as I am concerned, all superlatives can be removed. There are misstatements: such as "each designer uses his own conventions for aperture files"; Standard Geber format files do not have severe limitations: some EDA programs still stroke their zones, even for RS-247X; the format was not "severely constrained" and is as usable for EDA today as it was originally–modern CAM systems make it only slightly less convenient because of aperture files; RS-274-D is not obsolete, it refers to the superset not the subset. This section needs significant rework for NPOV and inaccuracies corrected. Once that is done we can argue about repetition.

Well, that is my two cents. —Dgtsyb (talk) 01:43, 2 November 2011 (UTC)

There are also issues with the Usage of the Gerber format section. The last two paras have no refs and read like WP:OR to me. Woz2 (talk) 12:24, 2 November 2011 (UTC)

Thanks, Dgtsyb. I was WP:BOLD and attempted to implement your suggestions with the minor exception in that I didn't remove the second dash from RS-274-D. The cite from Google books seems to indicate the second dash is correct. Woz2 (talk) 13:30, 3 November 2011 (UTC)

I basically agree with Karloman2. X must come first and stand on its own. The D paragraph is correct.Hi pcb 2 (talk) 02:13, 4 November 2011 (UTC)

Thanks, Hi pcb 2! Correctness isn't the issue. Verifiablity is. See WP:V. Wikipedia isn't for original research WP:OR. The (hopefully correct) facts in the article must have been published elsewhere in reliable sources WP:SOURCES. hth Woz2 (talk) 11:32, 4 November 2011 (UTC)

About RS-274D in the history: D is still used, and deserves a section for current users. Most are not interested and will not read the history.

OK How is it now? Woz2 (talk) 11:59, 4 November 2011 (UTC)

About Supplementing RS-274X with other data should be absorbed at the end of the Usage of the Gerber format section. I think that is a good idea.

OK Article now reflects this Woz2 (talk) 11:45, 4 November 2011 (UTC)

On RS-274-D vs. RS-274D: Probably both were used. The EIA name actually had the -D, but in the Gerber documents I only see D.

I've seen both. Probably not a big deal either way. Woz2 (talk) 11:45, 4 November 2011 (UTC)

On use of superlatives: Well, the tone is angry, but the facts are generally correct, and it is imporant for the industry that they are made clear to people that do not know them.

Correctness isn't the issue. Verifiablity is. See WP:V. Wikipedia isn't for original research WP:OR. The (hopefully correct) facts in the article must have been published elsewhere in reliable sources WP:SOURCES. Also the article must be written from a neutral point of view WP:NPOV Woz2 (talk) 11:45, 4 November 2011 (UTC)

On "each designer uses his own conventions for aperture files": It would indeed be better to just say "every EDA system uses its own conventions".

OK Article now reflects this Woz2 (talk) 11:45, 4 November 2011 (UTC)

On stroking : The statement is not correct. In D one HAS to stroke, in X one has NOT. It is a limitation of the format. The fact that some X files also stroke is deplorable, but it is not mandated by the format. This is discussed in usage.

OK Article now reflects this Woz2 (talk) 11:45, 4 November 2011 (UTC)

On CAM systems make it only slightly less convenient because of aperture files: This is a problem, even if people have worked hard to automate this; if X were used, this time could have been spent on other things. More important is the stroking and the lack of negative. This results in files that are hard to use in CAM. True, some X files are as bad in this respect, but it is important to state that this has nothing to do with the format.

OK Article now reflects this Woz2 (talk) 11:45, 4 November 2011 (UTC)

On RS-274D being obsolete: Ucamco, who are the stewards of the format, calls it obsolete, so in that sense it is obsolete. The fact that some still choose to use it is really not the point.

I think Ucamco's choice of word is incorrect. Obsolete means "no longer wanted." It is a decision of the consumer, not the provider. What Ucamco mean is deprecated, status applied to something to indicate that it should be avoided. Woz2 (talk) 11:45, 4 November 2011 (UTC)

On subset/superset: Actually Gerber RS-274D is a subset of the EIA format. EIA contains #D, parabola's etc which are not part of Gerber. RS-274X is a superset of the RS-274D subset.

OK Article now reflects this Woz2 (talk) 11:45, 4 November 2011 (UTC)

Inaccuracies: It appears to me that the facts about the format are in the published and referenced D spec. Obsolete in the Ucamco spec. And in ref 8, which is a good tutorial.

Then, inline citations should be added to support each fact. Maybe use the quotation field of the {{citation}} template to make the support really specific. Woz2 (talk) 11:45, 4 November 2011 (UTC)

76.102.136.152 (talk) 04:26, 4 November 2011 (UTC)

  • I would put RS-274X second. Need some explanation that raster plotters with memory can do scratch out rather than require troublesome development of ground plane images. X also gets around the positive/negative issue. Superlatives and deprecation are troublesome. Glrx (talk) 22:19, 17 November 2011 (UTC)
The Smithsonian has an extensive library on the Gerber Scientific Instrument Company at http://invention.smithsonian.org/resources/fa_gerber_index.aspx Is the personal critique of paragraphs 2 and 3 of Section 1.1 relevant? Section 1.2 has redundant info and seems an after thought. I'm not going to touch it :-) I B D Shank (talk) 17:21, 22 November 2011 (UTC)

Third version MDA?

Here: http://www.artwork.com/gerber/ ... it refers to a third varient "...Gerber (RS274D, RS274X and MDA) data..."

I Googled around and found http://www.blakerobertson.com/files/guides/pcb/bungard/manuals/isocam.pdf

Worth mentioning? Woz2 (talk) 16:41, 4 November 2011 (UTC)

MDA came and went. IMHO I B D Shank (talk) 17:09, 22 November 2011 (UTC)

Rev by Grlx, undone by Woz2

I agree with Woz2 undoing the last edits. The points raised by Grlx may be worth mentioning, but do not belong in the header. If they are mentioned, it should be done in the history section, now called timeline. They are interesting, but irrelevant to current users and current issues. I think this article should be useful to current designers, that have to make an output in Gerber, and want to know something about the format and how to use it. --Karloman2 (talk) 07:03, 18 November 2011 (UTC)

I disagree. Woz2 reverted the edits on the grounds that it was an extensive revision. That is not a reason for a revert.
Your position about utility to current designers is inappropriate. The article is about the format and not about current users and current issues. The article is WP:NOTHOWTO use Gerber files to make your PC board. The article has a what-one-should-do advice tone.
The original/reverted introduction contains several issues.
  • Gerber format is not normally used for "drilling and milling". Typically that is done with drill and rout files. Some PC shops can take a Gerber file and use flash information to create a drill file, but that is a hack (and they like to see the drill diameter in those pesky aperture files). Admittedly, some Excellon files look like Gerber files, but that's because both are EIA RS-274 machine tool files. Article mentions Excellon; does not mention Gerber for drill files.
  • Saying that Gerber Systems developed the format is strange. MIT/EIA developed the format; Gerber selected the subset. Even the Gerber source stated subset. (I don't know, but I would not be surprised that Gerber incorporated somebody else's machine tool controller rather than build one from scratch. Bridgeport, for example, builds a vertical milling machine; if you open it up, you may find a Fanuc controller inside. Or maybe Gerber had other machine tools at the time and were just extending its product line.)
  • That Mr. Gerber started other companies is irrelevant to the lead. Mergers and acquisitions are also not important to the format.
  • The intro does not explain the need for 274X. It just changed. Why?
  • The lead and the article make unsupported claims about the popularity of 274X vs 274D. "has been largely superseded." "It was designed for manual workflow." The article tone has problems.
Generally, material in the lead should be supported by text in the article. For the most part, the info in my rewrite has support in the article. Vector vs raster is mentioned in the article (but 274X buries the significance of raster for scratch out). The MIT claim is not supported, but there are sources for del of RS-274.
There are some other items in the article. An RS-274-D file can set the units and incremental/absolute. It does not set the axes.
The article also suggests there is some magic in 274X. It makes phototools. The format does not know about component side, solder side, solder mask, and silkscreen. The article body states that, so there are still outside-the-format agreements - just like the former aperture files were outside agreements.
I'm not saying my rewrite is great, but it is better than what was there.
Glrx (talk) 01:44, 22 November 2011 (UTC)
Actually, I like the original better as a lede. It was briefer and more to the point. However, I did like your additions. I would just like to see them in the body of the article. For example, the MIT information (although present in the referenced G-Code article) could be mentioned in the "History" or "Timeline" section. Also, some of the statements of vector vs. raster plotters should appear in the RS-274X and RS-274D sections.
  • With regard to use for drill and milling data, Gerber is a popular format to convert to G-Code to drive CNC milling machines to create circuit boards by milling away the unwanted copper rather than photoresist etching. Hobbyists even use engravers for this purpose. Better for the environment. Beats working with hazardous chemicals. Good down to 12-mil from what I hear.
  • Also, it is common use to provide a Gerber file providing the board outline which is used to route the board from the panel. Gerber is also used to provide board edge milling information.
  • As regards drill data, I have run across at least one fabricator that requests Gerber data for drilling information (yes, flashes with RS-274X aperture descriptions for the drill diameter) instead of Excellon or IPC-NC-349 and does not even mention that they accept Excellon. Excellon has way too much information that is specific only to Excellon's CNC7 drill machines and a simple RS-247X flash file is in fact easier to use.
  • Also, Gerber files are not used to make photoresists anymore. They are simply a way of transferring the design data which is imported into a CAM system and outputted in whatever (usually proprietary and machine specific) format the fabrication equipment happens to need. Gerbers are even used to print to a piece of paper to used the toner-method for board etching, although most CAD tools output Postscript or PDF files for this purpose.
But please, add your information into the article body in the appropriate places. — Dgtsyb (talk) 23:49, 22 November 2011 (UTC)
Every CAM system I have used outputs Gerber which is used to make the photo tools. Low end CAD systems do not generate IPC-256. Every board shop I work with can generate an electrical test file from the Gerber files. In fact many board shops can not support electrical test from IPC-256. I B D Shank (talk) 00:36, 23 November 2011 (UTC)

I've asked for a WP:3O Some points in favor of the present version:

  • EIA developed RS-274, but Gerber developed the subset that is the topic of the article. By the way, as noted in the article it's very confusing to use RS-274 to refer to both the original and Gerber's subset.
  • It's important to know who owns a standard because of vendor lock-in. Readers will want to know who they're going to be locked into. The "merger and acquistion" phrase was just a linguistic device to join the dots between the originator the present owner.
  • The Tavernier ref is a reliable source that the original version has been superseded. His advice on page 17 is "Do not use it."
  • The MIT material in the Glrx version is unsourced. The MIT connection is also unsourced on the G-code article.
  • The Glrx version deleted mention of H. Joseph Gerber, an issue that previous third opinion weighed in on (see above) ' I don't see a problem with mentioning in the article that Gerber Systems Corp. was named after its founder, particularly if there is no separate article on Mr. Gerber or Gerber Systems Corp.'.
  • I do agree with Glrx on the " as well as the drilling and milling data" point and I've removed it from the present version.
  • I also agree that the article has an undesirable how-to tone in parts (esp. 2nd & 3rd para of RS-274X extended Gerber-->Usage) but that is a separate discussion since neither the Glrx edit nor my revert touched those parts.

Woz2 (talk) 17:40, 22 November 2011 (UTC)

I generally agree with Woz2's comments. However, the Gerber format is used quite often to specify drill and route information. (Actually, I prefer it because it is then more probably that the copper pads and drill holes register.) Excellon-like or generic NC files are most used, but about 10% is in Gerber. I googled "PCB dril Gerber" and I found this reference: http://circuitpeople.com/Blog/WhatIsAGerberFile.aspx. I am sure that there are more. I strongly suggest to restore the "as well as the drilling and milling data"

--Karloman2 (talk) 13:38, 23 November 2011 (UTC)

Good points. How about a compromise given the majority of drill data is Excellon: omit "drill and mill" from the lead, but mention it elsewhere? Either way (lead or body) is fine with me. Woz2 (talk) 13:53, 23 November 2011 (UTC)
A Google search is not an authority. Blogs are not reliable. When I look for "drill" in the proffered http://circuitpeople.com/Blog/WhatIsAGerberFile.aspx, it says something much different. It states that a manufacturer may "receive up to 9 files", but it doesn't say those files are Gerber files (the ninth file, a "README" file, would not be a Gerber file). The blog states that if the drill information is given as Gerber or DXF file, then hole sizes must be "manually translated to NC (Excellon) file(s)". The page is silent about rout information other than file #8 (which may not be a Gerber file).
Drill and mill definitely do not belong in the lead. The mention elsewhere needs a big caveat; is this article about PCB manufacture, or is it about a file format?
The comment that really being made is that fab houses are nice to their customers and will attempt to infer drill and rout information from the files they are given, but that was not an intended use of a Gerber file. If it is mentioned in the body, then it should point out that the customer's information was deficient Dea
Dear Glrx, I agree with you that the reference I proposed was not so good. However, I looked further, and the one I used in the article is quite good IMHO. Gerber files are used for drill, as confirmed by the reference and as commonly known by anyone in the PCB industry. (I actually prefer drill data in Gerber because there are less registration issues, but that may be a personal opinion, and it is not important here.) If Gerber is used for drill data, whether one is in favor of it or not, it should be mentioned next to the other PCB layers, e.g. soldermask. So I think the article is ok as it stands now. --Karloman2 (talk) 12:41, 27 November 2011 (UTC)
The advice "do not use it" does not imply it has been superseded.
If the owner of the format says "Do not use it" then, by the definition of superseded ("to cause to be set aside, to force out of use as inferior"), it is superseded. Woz2 (talk) 02:15, 27 November 2011 (UTC)
Well, no. It just means that the owner (if one can really "own" an ipso facto standard) wants it to be superseded. What the industry does in this case is something completely different. For example Barco, when they "owned" the format, tried (unsuccessfully) to supersede Gerber with Barco DPF. — Dgtsyb (talk) 09:26, 27 November 2011 (UTC)
Dgtsyb has a point, what the owner of format wants is not necessarily what happens. However, the article merely states "... is largely superseded by RS-274X". Supersede = "take the place of" according to dictionary.com. Surely we all agree this is true. It it then a matter of finding better references than only Barco, but apart from that IMHO the article is fine as it stands.--Karloman2 (talk) 12:41, 27 November 2011 (UTC)
The article should do D and then X. The article can intro D, explain problems, and show that X solves some of those problems.
Dear Glrx, The order of D and X was discussed before, and there was a consensus to put X first, as being the more important format. I also think this is better. People may be interested to know about X, and one can and should explain X without explaining problems in D and old vector photoplotting. However, I agree that it does make sense to explain these difficulties and how X was "invented" to overcome then. Good point, but I think this belongs in the history section, for those that are interested in these things.--Karloman2 (talk) 12:41, 27 November 2011 (UTC)
Glrx (talk) 18:31, 26 November 2011 (UTC)
Dear Glrx, my mild preference is D then X but when I polled a WP:3O the tie breaker was in favor of X then D 9see above) and I accepted it. Not a big deal either way for me. Woz2 (talk) 01:08, 28 November 2011 (UTC)

I agree with Woz2 that this must be mentioned, otherwize it is not clear why this format is called Gerber. I suggest to leave a "merger and acquistion" phrase, but make it shorter to accomodate Glrx's remark, and only mention the salient fact of the acquisition. The intermediate history, and it does not belong here, it can be probably be found on the Barco website. I found a reference on Barco acquiring Gerber. Some of the citations needed can be found in the Tavernier reference. I also found a Wikipedia page on Mr. Gerber. Drilling and milling; we can only mention drill in the lead, the milling being covered by the etc.; this makes it shorter but still indicates non-image use.--Karloman2 (talk) 10:38, 26 November 2011 (UTC)

I wouldn't mention Mr. Gerber in the lead (or acq of Gerber Systems by successors). I have no problem with Mr. Gerber's exploits and the subsequent acquisitions being mentioned in body. Such details do not belong in the lead.
I have trouble with the notion of "owning" Gerber format. Any company that makes photo plotters will at least consider supporting such a popular format. There are plenty of dialects out there because competitors had slightly different interpretations, and there are plenty of configuration options in PCB software to handle those dialects.
Glrx (talk) 18:44, 26 November 2011 (UTC)

A point of fact if it matters: Gerber Scientific Instrument Company developed Gerber, in 1978 it became Gerber Scientific, Inc. when it when public, Gerber Systems was/is a subsidiary of Gerber Scientific, Inc. I B D Shank (talk) 23:26, 26 November 2011 (UTC

Dear. I B D Shank. It does matter, but on the article on the Gerber company, which is really but a stub. Here, there is already a discussion if the Gerber company and Ucamco must be mentioned at all, so one tried to make it as short as possible. In my old Gerber manual it is stated that the format was developed by Gerber Systems Corp. so I think the article as it stands now is correct in this respect. (I personally would not mention Mr. Gerber for the sake of brevity, but most people seems to want to keep him in.)--Karloman2 (talk) 13:16, 27 November 2011 (UTC)

Dear Woz2, you feel the citation does not support the article. However, the press release mentions Barco Electronic Tooling Systems, and Ucamco calls itself Former Barco ETS. According to their website, ETS indeed stands for Electronic Tooling Systems. So I think the reference does support the article. Their may indeed have been intermediate steps, as you wrote, there probably where, but the article only writes "... through its acquisition...". This does not state there were no other steps, but only that the acquisition was the enabling factor. IMHO this is enough for the leader of an encyclopedic article. It explains why a Ucamco format is called Gerber. Another solution is to make it even simpler. Just state the format was developed by Gerber Systems Corp., and is now owned by Ucamco, without mentioning how and why. Maybe this is better. What do you think? --Karloman2 (talk) 13:16, 27 November 2011 (UTC)

I added a second cite that links Ucamco to Barco's ETS div -- hth Woz2 (talk) 17:59, 27 November 2011 (UTC)
This is perfect. --Karloman2 (talk) 23:44, 27 November 2011 (UTC)

Ideal or Real World?

Section 13 was getting too long.  I tried to fix the reference the Bot did not like.  Hope it flies this time.  I could swear I saw an article on Gerber Systems here a week ago.  My imagination, or it got deleted.

I'm going to leave the article to you guys, I have too much work to get done.  For what it's worth, the article already has a "This article may be too technical for most readers to understand." tag and IMHO there is too much theoretical information here and not real world information.  E.g., reference 12, I have never seen a G36 or G37 in my life, but BARCO says you better use it.  I don't know of any system that does not paint or make a reverse image for planes.  Nobody is going to listen to a Belgian company, too close to France. :-p

They are talking about IPC-2581, because about 2500 before it have been ignored.  IPC-2581 will most likely be also.  How do I know?  Been doing this for 40+ years, mostly military work.  Worked for, used, and repaired/rebuilt Gerber and Excellon machines and have used almost every CAD/CAM system there ever was(currently Cadence and Solid Works).  F.Y.I., the Excellon article is a "real" mess.

I found a copy of IPC-2221 on the web in PDF at http://www.the-bao.de/divers/ipc2221.pdf.  "That" is the standard that everyone follows.  I am not sure how it is on the web, I think I paid $350 for my copy.  I'm not sure I would trust it being there for long, but I need to ask around to see what the current copyright situation is, because IPC-2221 replaced IPC-275, which replaced MIL-C-275, which was public domain and was supposed to remain so when the military handed over PWB specifications to IPC.

Also many of the references are from companies I would stay away from, so using them as a source may be questionable(IMHO).

Happy editing.  Cheers :-) I B D Shank (talk) 18:46, 27 November 2011 (UTC)

Dear Dcshank, I do not have your long experience or deep insight. However, I am amazed you never saw a G36/G37. May I ask in which country you work? --60.32.109.162 (talk) 05:56, 19 December 2011 (UTC)

Third Opinion on Ucamco info in lead

I'm of the opinion that it's important to mention in the lead section that Ucamco recommend that Standard Gerber not be used any more. Please add your opinions here. Woz2 (talk) 16:10, 28 February 2012 (UTC)

I'm a Third Opinion Wikipedian and I've removed your 3O request because I've not been able to find any talk page discussion about this issue. All content dispute resolution methods require that the dispute be discussed on a talk page, preferably the article talk page, before resorting to dispute resolution. Mere discussion via edit comments will not suffice. If this dispute has been discussed somewhere other than here, please provide a link or links to the discussion and relist at the Third Opinion project. Regards, TransporterMan (TALK) | DR goes to Wikimania! 17:20, 28 February 2012 (UTC)
Thanks! So far the discussion, as you mention, has been in the edit comments. I invite User:Dgtsyb to make a statement here, then I'll re-list. Woz2 (talk) 17:39, 28 February 2012 (UTC)
You might want to put
==Talkback==
{{talkback|Talk:Gerber_format#Third_Opinion_on_Ucamco_info_in_lead|ts = ~~~~~}}
on his talk page since s/he may not be watching this page. Hopefully you'll be able to talk the matter out and further DR will not be needed. Remember that just a "you did" "I didn't" discussion may not be enough. Try to get into the reasons for your respective positions, taking into consideration Wikipedia policy and guidelines. Regards, TransporterMan (TALK) | DR goes to Wikimania! 20:27, 28 February 2012 (UTC) (I am no longer watching this page.)
  • Support notion but Oppose current. I have trouble with the statement phrasing in the lead section.
The quotation is from an article by Karel Tavernier who is an employee of Ucamco. Although he is an employee, that does not mean he is speaking for the company. The citation incorrectly lists Ucamco as an author; my reading is Ucamco is just the affiliation of the author. The more neutral statement is that industry experts recommend using RS-274X instead of RS-274D.
The claim that Ucamco owns both versions isn't quite right and need not be in the article. The owner of RS-274D isn't clear. EIA certainly did some broad strokes. Gerber decided to use D for tools/apertures and on/off/flash, but I don't think that created ownership. There are many manufacturers who use similar but slightly different RS-274D interpretations for their machines. Maybe they were trying to copy Gerber but made a few mistakes along the way. It doesn't matter. Certainly Ucamco cannot tell them what to do. Plenty of PCB packages can be tailored to variations of RS-274D machine that are out there. Probably the same story with RS-274X ownership. Ucamco could obsolete its spec tomorrow, but that wouldn't have much effect. I would not mention Ucamco in the lead.
Many fab houses will take either version. They also have software that handles many variations of 274D. Use RS-274X and you don't have to worry about the wrong aperture file, some D code quirks, the leading/trailing zeros problem, or where the decimal point is. There's a better chance of success. Tavernier's statement is a recommendation to avoid problems with the fab houses misunderstanding something. (But the fab house is still going to check pad sizes and drill sizes, make sure the tooling holes are there, and check a host of other issues because it wants to keep the customer happy and not make a batch of junk boards.)
I think a statement recommending 274X can go in the lead, but I'd make it more neutral. RS-274X can be a complete description of the image where RS-274D requires a side agreement about how features are specified. Consequently, RS-274X can be a more precise description. Industry authors recommend using RS-274X over RS-274D.(then cite Tavernier.)
Glrx (talk) 21:20, 28 February 2012 (UTC)
  • Support notion, can live with current words but prefer words before do/undo sequence initiated by Dgtsyb edit from 26-Feb-2012

Actually, this topic was discussed on this page in November 2011. After a long discussion, the wording before 16-Feb-2012 was agreed as a consensus. They were left untouched till the recent edits, which also indicates they are acceptable. I strongly suggest to revert to these words as a consensus. These were not words I preferred, but as these reflected a consensus, I could accept them. (I did some edits in the intro section, but they were stylistic, and I did not touch the sentence on D being deprecated.) Both ref 8 and UCAMCO (Tavernier) call Standard Gerber obsolete. (Iavernier is MD of UCAMCO, the article is published on the UCAMCO website, the same words are repeated in the spec, I think we can safely assume this is UCAMCO's position.) In November it was agreed UCAMCo used the wrong term, and that they meant deprecated. Surely UCAMCO - owning Gerber - owns on the Standard Gerber format, even if it does not own the EIA RS-274-D spec, which is much more wide than PCB. The fact that the industry may not follow UCAMCO's recommendation is not the point here. This is the reason why the Standard Gerber is deprecated, but not obsolete. It must be in the intro because it is a salient fact about Standard Gerber. It is very important for the industry. Many people use Standard Gerber out if ignorance, and may use Extended Gerber if they notice this fact. If they do, we will have made the world a little bit a better place.
I move to revert to the statement before the do/undo sequence started. --Karloman2 (talk) 21:57, 28 February 2012 (UTC)

  • Thanks, Karloman2. I too prefer the pre-Feb 26 version, but I could live with the compromise. Standard Gerber is superseded and deprecated and it's important that it's so noted in the lead. By the way, now that there are three of us debating this, we can't ask for a WP:3O because that mechanism is for a tie breaker between two contributors... so we have to settle it ourselves... Pistols at dawn? :-) Woz2 (talk) 23:17, 28 February 2012 (UTC)

Comments from a reader

"The specification can be freely downloaded" What "specification"? IMO this is referring to something that has not been mentioned yet. Perhaps it would be better being "The Gerber format specification can be freely downloaded"

Additionally, "for example by SMT placement equipment and wave soldering." who uses wave soldering for SMT nowdays? It would be better to say "for example by SMT placement equipment and reflow soldering." But why does it need to say anything about that at all?

Cadstar_User (talk) 11:28, 1 April 2012 (UTC)

Thanks for your feedback. 1) In the context of an article about Gerber format, it seems to me there's no difference between "The specification can be freely downloaded" and "The Gerber format specification can be freely downloaded" but go ahead and change it if you think it's clearer. 2) I'll add reflow. The purpose of that whole paragraph was to address an early criticism that the article lacked context for the general reader, and so that brief overview of PCB manufacture was added. Woz2 (talk) 22:01, 1 April 2012 (UTC)
Yes, wave soldering, and in particular selective wave soldering is still used. I agree with Woz2 on the other. — Dgtsyb (talk) 03:10, 2 April 2012 (UTC)
I removed the sentence about soldering etc. before I noticed there was a discussion going on. Sorry for this unintended rudeness. However, I think it not directly related to the Gerber format, and hence is outside the scope of this article. PCB assembly is discussed in the PCB article, and this article links to it in the beginning. Admittedly, the article is not useful for the general reader. Well, it is about the Gerber format. My wife will not read it. Adding a sentence about PCB assembly does not make it more generally interesting. My wife will still not read it. The article is the first port of call for people that hear about the Gerber format - typically people starting in PCB CAD and CAM- and need to know something about it. That, I think, is the purpose of an encyclopedic article. And if is encourages them to create better Gerber files, well, it will have made the world a tiny little bit a better place. Helping such people to understand Gerber is best achieved by a focused article. — Preceding unsigned comment added by Karloman2 (talk • contribs) 08:40, 4 April 2012 (UTC)
Maybe we should get an opinion from User:Curb Chain who flagged the article as being too specialized? I added the contextual intro in response that flagging WP:Jargon Articles in Wikipedia should be understandable to the widest possible audience. For most articles, this means understandable to a general audience. Woz2 (talk) 18:17, 4 April 2012 (UTC)
There are lots of references to PCB design, I don't see any problem with removing that particular sentence. — Dgtsyb (talk) 01:01, 5 April 2012 (UTC)


There's no question that the information can be found. But if a general reader wants to know "how does Gerber fit in to the fact that my electronic device exists" it's convenient to have a thumbnail of the story arc in one place. Restoring it for the general audience. Woz2 (talk) 01:29, 5 April 2012 (UTC)
The sentence was technical and introduced the concepts of reflow and wave soldering without any context other than mentioning PCB fabrication (another concept with no introduction). In fact it is not fabrication, it is assembly. A Gerber files has nothing to do with the method of assembly (i.e. type of soldering) and using the terms without introduction or explanation just clutters the main topic. — Dgtsyb (talk) 20:00, 5 April 2012 (UTC)
I don't think so. The sentence gave a proper context to the general audience. By deletion, the general reader is now left wondering "So what?" and "So what's next?" Before deletion, the terms were wikilinked to articles whose leads gave definitions and more info. See WP:LINK Internal links bind the project together into an interconnected whole. By the way, I found your comment "get real" in the Edit summary to be uncivil. Woz2 (talk) 17:46, 8 April 2012 (UTC)
I have no problem with some context being added to the article to show how Gerber files fit into the manufacture of PCB. The deleted comments about placement and soldering did not do a good job of supplying that context. The article gave a bare mention of photolithography. The context should describe more about photolithography such as a plotter reading the Gerber format and exposing film to make the phototools. Other photolithographic steps (that show the context of how the phototools are actually used) should be mentioned, such as applying the photo resist to copper clad boards, developing the resist, etching to achieve the desired conductive patterns, and then stripping the resist. The article also skipped other PCB steps such as multilayer alignment/registration, drilling, and plating holes (vias). The comments about placement and soldering thus come out of left field and were not well connected to the Gerber file. Glrx (talk) 18:57, 9 April 2012 (UTC)
There's no question there are many other improvements that could be made. That's no reason to delete my contribution about how Gerber fits in overall. Woz2 (talk) 00:44, 10 April 2012 (UTC)
I agree with Woz that the Introduction section is not terribly helpful, and that it would be useful to explain how this little format fits in PCB production. However, I do not think adding his sentence is much help. Instead of duplicating content, I feel it is best to refer to the PCB article, section manufacturing. Alas, this section does not mention the crucial role of CAM and is silent on Gerber or any other formats. I suggest we improve the PCB article and refer to it. The Introduction in the Gerber format article is not very useful. The article starts with what is in effect an introduction, referring to PCB. I suggest to remove the Introduction section altogether, and work on improving the Manufacturing section of the PCB article.--Karloman2 (talk) 23:18, 10 April 2012 (UTC)
The Introduction section is indeed better after the edits of Woz2 and Dgtsyb. Fine with me. The PCB article could, however, be improved by explaining the role of CAM. Volunteers?--Karloman2 (talk) 15:19, 12 April 2012 (UTC)
I don't like the last sentence of the present version. It seems rather odd to say what it doesn't do. It doesn't play a role in the TV series "Lost" either but we don't feel the need to mention that. If everyone else (for reasons that escape me) is against mention of soldering, then my second choice would be "For information on the next steps see ______" I volunteer Karloman2 to rewrite the PCB article :-) Woz2 (talk) 22:46, 12 April 2012 (UTC)

RS-274X as drill data

Hallo Glrx, you dont like RS-274X as a source of drill data. I am under the impression that you can transfer drill and rout information from CAD to CAM pretty well in RS-274X. (Not for driving a drilling machine of course, to do that you need a proper machine control file.) RS-274X has the practical benefit that RS-274X are usually complete and of good quality. NC files are all too often problematic: incomplete (no unit, no coordinate format, no tools, EIA encoding. But there may be things I did not think of. What problem do you see in RS-274X for drill and rout? — Preceding unsigned comment added by Karloman2 (talk • contribs) 14:35, 31 May 2012 (UTC)

Sorry, but I missed this earlier.
RS-274X is not a source of drill data. Where are speeds/feeds/drill sizes specified? Blind via drilling information? There are shops that can take RS-274X and infer simple drill information from flashes and annular ring design rules, but that is not the purpose or the intent of RS-274X. What if there are no flashes? Unplated mounting holes don't need them. Similarly, where is the router information? A technician at the PCB shop can find a board outline drawing and use it to program the router, but that is similar to a machinist looking at and interpreting a drawing. Yes, the PCB shop is going to look at all the files because most PCB designers don't understand enough of the manufacturing issues. The shop will probably do the panel layout and make sure that tooling holes are there and router won't fling valuable boards into orbit. The shop may also alter the speeds and feeds to what it believes is reasonable. But that does not mean drill and route data should not accompany the RS-274X files.
More simply, where is the WP:RS that says RS-274X is used for drill and rout? Even the lede says "de facto industry standard for printed circuit board image transfer".
Glrx (talk) 17:02, 19 June 2012 (UTC)
Speeds and feeds are not even accepted in NC drill files: there are added by the CAM system and are specific to a tool. Blind via drilling information is provided by separate NC drill files for each layer-pair. True drill diameter is not normally specified in NC drill files: it is "finished" diameter. Separate NC drill files are normally required for plated an non-plated holes. Drill sizes are adjusted for plating recipes by the CAM system. Route and milling information is not provided by NC drill files and is normally provided by Gerber "board outline" layer. There is no way that your unoptimized 1-up finished hole size NC drill file is ever going to be used directly to drive a machine unless you are doing it yourself (and want to ruin boards). There is nothing usable in the NC drill file that cannot be placed in a Gerber file instead.
The advantage of Gerber over the cheap NC drill files output by many CAD systems is that the Gerber contains dimension information, whereas the cheap NC drill file does not. This places the requirement that a Gerber "drill report" be generated anyway that identifies the placement of all holes with respect to the board outline and lists all sizes and plating requirements. Gerber is predominantly used to provide fabrication drawings and drill reports in this fashion. (I don't know whether that is highlighted in the article.)
I know of three medium to large fab houses that state in their DFM requirements that they accept drill data in Gerber format. It is just a fact. — Dgtsyb (talk) 08:09, 20 June 2012 (UTC)
I think it depends on the purpose of the drill file. I certainly agree one should not try to send a Gerber file to a drill machine, supposing the machine would accept it. Gerber is not adequate for that. However, I was referring to data transfer from designer to manufacturer, and should have said so. These files are never put directly on the machines, but always first read into the CAM system. This is in effect CAD to CAM data transfer. You do not need the feeds and speeds to tell where the holes are, and what the end diameter is. Gerber is perfectly adequate for that. The CAM system will then generate drill files, with the tool diameter, feeds and speeds, and so on. These files are then in Excellon, Sieb&Mayer, and so on, but never in Gerber. — Preceding unsigned comment added by Karloman2 (talk • contribs) 16:48, 21 June 2012 (UTC)

Gerber format for CNC

The article leaves me with a few questions. For example, is Gerber format essentially a 2D computer description or 3D? While drilling holes in a PCB is essentially a third dimension, there may not be any sort of depth control. It may be useful to apply Gerber format for 3D milling operations, such as computer numerical control (CNC) if Gerber format supports this. I recall that decades (yes decades) I worked with a Gerber plotter which must have had some sort of motion speed controls on the pen motors. Since I'm no expert here, I'd like to ask those who are to possibly comment or enhance the page for clarity.--71.10.146.139 (talk) 00:09, 30 October 2013 (UTC)

It is mentioned deeper in the article that Gerber is a 2D format. However, this indeed should be in the introduction. The introduction strangely enough starts with telling what Gerber is used for, but not what it is. Karloman2 (talk) 07:43, 30 October 2013 (UTC)
Done. Thanks for your remark. If there are more points that are not clear, please let us know. Karloman2 (talk) 08:10, 30 October 2013 (UTC)

These are interesting links, but do not fit in this article. This article is about the Gerber Format. One of the links is about the company Gerber Scientific and the other about photoplotters. Both have their own articles. I favor moving the links to the directly applicable articles. --Karloman2 (talk) 11:31, 6 July 2014 (UTC)

Common file extensions

A table with common file extensions was added to the article. I am not thrilled by this. It takes up a lot of space to give an example of file extensions/naming convention. It is just one example, there is no standard. (Ucamco should specify a standard file naming convention.) Furthermore, it documents, and therefore seems to promote, the annoying habit of (ab)using the file extension to define the file function rather than the file format. Really annoying if you have to work with such file. I know the article is just describing and not promoting, but by spending so much space on it the article seems to indicate to the superficial reader - most readers therefore - that this is somehow standard and should be used. I am very much against promoting this. A previous version of the article gave a few examples of this practice, and that is IMHO more than enough. --Karloman2 (talk) 12:26, 6 July 2014 (UTC)

On G-Code, RS-274-D

Dear Efa, let us talk about this before embarking on an edit/undo war. I admit being hasty in stating G-code is not RS-274-D. I do not know G-code in any depth. I do know D. I have a copy of the original specification. I can only say that G-Code as in the Wikipedia article only vaguely matches this old spec. I also do not see a clear reference stating that D = G-Code. Personally, I am not convinced this is true, but I am not an expert, and I stay out of that discussion. However, what I do know is that what described in the Wikipedia G-Code article has nothing to do with the current Gerber format. For current Gerber users it is irrelevant. But true, they have a historic connection through RS-274-D, whatever that connection may be, there is one. This is IMHO not worth both a link in the text and an explicit link, a link on the first use of RS-274-D in the text is the standard way of establishing that connection. I will do an edit that reflects this opinion, hoping you can agree with it. (I will also do an edit in the extended Gerber section, but that has nothing to do with this discussion. When, triggered by this discussion I read this section I noticed that it could be improved, making it closer to the specification. Karloman2 (talk) 15:11, 1 January 2015 (UTC)

Change in the name of this article

Sbmeirow changed the name of this article from Gerber format to Gerber (file format). I am not enthusiastic.

  • The argument give is similarity with other formats. It fully support the aim to conform to general practice, but the new style is not really general: other formats in the PCB industry have titles as Excellon format, Portable Document Format; not the new style. The strength of the argument is therefore relative.
  • The title must be the name of the object discussed in the article. The title Gerber (file format) gives me the impression that the name is Gerber with (file format) added to distinguish it from other objects called Gerber. The title of the spec is The Gerber Format Specification; this means the name is Gerber Format; Gerber may be used informally but surely it is not the correct name.
  • Brackets in a title is not elegant typography.

I move to restore the original title.

Comments? Objections?

Karloman2 (talk) 21:55, 1 March 2015 (UTC)

I restored the original name as there are no objections after a week. Note that the name was discussed in the first section in this talk page. I looked up the guideline for renaming pages, and none of the reasons compellingly applies:

There are many reasons why you might wish to rename a page: • The title does not follow Wikipedia's naming conventions, such as that it is not the common name of the subject or it is overprecise • The subject of the article has changed its name and the new name has come into majority use • The title has been misspelled, does not contain standard capitalization or punctuation, or is misleading or inaccurate • It needs to be disambiguated in some way to avoid confusion with an existing, similarly named topic, or it exists at a disambiguated name but should not because it is the primary topic • It is an article at a descriptive name and the scope of the article has been reduced, extended or otherwise changed • It is an article that has been created as a subpage of a Wikipedian's user or user talk space for development purposes and it is ready to be posted to the mainspace or to draft space.

The guideline for page names states "Generally, article titles are based on what the subject is called in reliable sources." The specification and other sources clearly call this the "Gerber format". ("Gerber" is of course also used for brevity if it is clear from the context what the object is. This does not apply here as it creates an ambiguity which them must be resolved. With "Gerber format" there is no problem._ Karloman2 (talk) 12:06, 7 March 2015 (UTC)

It hasn't been OVER a week, nor has anyone else commented, thus you acted too quickly. • Sbmeirow • Talk • 15:11, 7 March 2015 (UTC)
The gerber spec uses the phrase "Gerber file format" a total of 17 times inside the revision J4 document, so that is a pretty strong support for "Gerber (file format)" or "Gerber file format" article name. The most common usage of Gerber commands is being stored inside of a file. • Sbmeirow • Talk • 02:14, 8 March 2015 (UTC)

I agree with Karloman. Also in the German PCB Industrie nobody talks about Gerber as a file format, but about the Gerber format. It is a description of a program, used for programming NC machines (in early days). It would be similar to rename a 'C' file to C (file format). Where it is not a file format as such, but a language. File format can be used for describing the structure of a file itself. For example: zip files start with PK, then a number representing the amount of files contained, then the size of the first file and so on... TjibbeDeFries (talk) 08:15, 11 March 2015 (UTC)

Dear Sbmeirow, you definitely did your homework in researching the name. I agree that it is not so clear if the name is Gerber format or Gerber file format. Maybe you could complement your work by asking Ucamco what the name now is and to make it clear for all; that would be helpful. However, this does not change my position that the name of the article is best left as it was. First, the article name Gerber (file format) would be appropriate if the name of the object were Gerber; there is little support for that; the title of the spec seems to contradict it; it is a name that introduces an ambiguity whereas 'Gerber Format is unambiguous. Second, the original article name was discussed in the past and a consensus was reached. It should not be lightly changed and there are no compelling reasons to change it; there is no problem with the current name. Although I know your change makes sense in a certain context and was made with the best intentions I feel the old name is best kept. Furthermore, there are a number of external sites referring to the Wikipedia article wit the old name.Karloman2 (talk) 10:07, 15 March 2015 (UTC)

Trouble archiving links on the article

Hello. I am finding myself repeatedly archiving links on this page. This usually happens when the archive doesn't recognize the archive to be good.

This could be because the link is either a redirect, or I am unknowingly archiving a dead link. Please check the following links to see if it's redirecting, or in anyway bad, and fix them, if possible.

In any event this will be the only notification in regards to these links, and I will discontinue my attempts to archive these pages.

Cheers.—cyberbot IITalk to my owner:Online 14:54, 13 July 2015 (UTC)

Retrieved from "https://en.wikipedia.org/w/index.php?title=Talk:Gerber_format/Archive_1&oldid=1127592221"