Project:Hot air flow

From Inventing aviation
Jump to navigation Jump to search

Contents

Aero as power source

Leaving a rough note here for further development and comment. Having found an early volume of the GB Abridgments which catalogues aero patents before the 1860s, I'm seeing descriptions of the earliest British patents classified under "aeronautics". A large fraction of them deal with the application of power derived from the air to other types of transportation—namely, carriages, boats, tracks. E.g. Patent GB-1852-155, Patent GB-1849-12452], Patent GB-1843-9642, Patent GB-1853-1088; & see keyword track.

It would seem appropriate to devise a keyword for this category of invention.

Additionally, it seems significant for a history of technology and innovation, that more inventors proportionately saw aero as a means to an already known end, before later seeing it as an end in itself. Cheers, LTA (talk) 10:46, 21 February 2021 (PST)

Power derived from the air -- by windmills? There seem to have been American patents for "wind wheels" (analogous "water wheels") as well as windmills. There may be a distinction between wind wheel and wind mill but I haven't noticed it. -- Econterms (talk) 17:54, 23 February 2021 (PST)
Not windmills, but power in the form of motion derived from properly aero flyers such as balloons and kites which pull something on a track or on the ground. These were common and indexed within aero categories. LTA (talk) 14:37, 3 March 2021 (PST)
Ah. Sounds interesting. I see a discussion in Neilson's book that he doesn't want to include all the Aeronautics patents - e.g., not balloons, and maybe these also. Maybe count them as .5-relevant for now?
Tentatively calling these aero-as-power. LTA (talk) 09:20, 8 June 2021 (PDT)

Templates linking GB patents to pages in B&A (1893) and Neilson (1910)

Newly improved templates will link right to our descriptions of older British aero patents in two books which are perfectly scanned at Internet Archive. This wiki code will unpack like the references below: {{BrewerAlexander|40|p40-41}}, {{Neilson1910|71|p71}}

Mexican patents

Great find. It seems the reason they don't display at first (and come into lists doubled) is the apostrophe before the asterix, right after the template, introducing the inventor location and sources. Was this deliberate for some reason? If not, deleting the apostrophe is a pretty easy fix. Meanwhile, find/replace seems unwilling to target the string "M�xico". Also, the date granted on many or all of these came in as 1905. LTA (talk) 04:53, 30 April 2021 (PDT)

  • Yup. I tried to be careful but messed up a lot. And it made errors with an character with diacritics. I will try to clear out the extra apostrophes. -- Econterms (talk) 19:36, 1 May 2021 (PDT)
    • I'm bummed that we can't search-and-replace the strange character that came in. It can be done by hand, one by one. The strange character comes in because of some kind of dreamhost mistake and/or the Page Transfer extension. I know how to investigate when I work up the steam to do it. -- pbm

Lugar vs. Residencia

It seems likely that "Lugar" in the spreadsheet is the inventor's location, since there are many foreign applicants with known location matching the entry there. Meanwhile they almost all have México, D.F. for Residencia but I would really think that all these inventors, Charles Stanley for instance, don't live in Mexico City. Sometimes "Residencia" is another Mexican state. Could this be the location of the filer or agent? LTA (talk) 04:05, 9 January 2025 (PST) This pattern seems to differ in other parts of the spreadsheet; in another stretch of entries there is no "Lugar" but only "Residencia" and here the latter term seems likely to be the true location. LTA (talk) 04:36, 9 January 2025 (PST)

I've noticed pattern changes like that before, in our own data and in our sources, a data point meaning a particular thing, given some other factor, with different signification, in the absence of that factor. It seems that you've got this tout précisé.AvionHerbert (talk) 09:25, 11 January 2025 (PST)

One standard patent report now includes Filing date

  • I added Filing date to the kind of patent report that shows up on Thomas Sloper because he has a patent thicket and it helps to figure out parent-child relations. Maybe this would help on all patent reports? Now I'm thinking this is long overdue. But we could also create templates with short/simple patent reports. Comments welcome. -- pbm
  • I think that would quite often be handy, and most particularly as applied to inventor pages. -- jrh

Categories for archives and/or repositories

  • I'm thinking we can mark pages here about physical archives in Category: Archives, and online or electronic ones in Category: Repositories. Good? I'm inclined to use plural but I'm not sure what's best. But that helps us track places and sources. -- Econterms (talk) 18:13, 20 June 2021 (PDT)
  • Would Category: Online repositories make the distinction relative to physical archives more explicit? Plural certainly.-JRH
    • Agreed. I've created them. Let's see if it does some good. There are MANY places, both online and physical, that have primary materials of interest. I guess databases with secondary information could be marked too, as "category: databases" or something. -- pbm
    • Regarding Category: Online repositories, in a few cases I'd like to create fairly brief pages, appropriately categorized, featuring external links along with succinct references to offices, modern or historical, with which they are associated, along with references to archival facilities which hold the actual physical documents.
    • Regarding Category: Archives, a key type of secondary information in the context of catalogues or "journals" as they call them at STIC, could be marked in various ways, with AGR 2 - Cuvelier for instance being great for primary documents as well as catalogues, but all Belgian, whereas STIC is greatly international in terms of catalogues, but holds no(?) patent originals. Perhaps highlighting as an emphasis on STIC, for instance, would be handy.
      • Please proceed creatively!

Concepts of patent class and subclass

JRH, I boldly made edits to HU V/h that may not seem suitable. The internationally comparable concepts to Főosztály and Osztály appear to be patent class and patent subclass. Across countries, patents were assigned to subclasses -- sometimes one subclass, or sometimes a list of them. Patent classes are descriptive, and are made up of collections of subclasses, and generally do NOT have patent assigned to them. E.g. "machinery" or "aeronautics" or "Sport" might be classes. I would like to say that these words match Főosztály and Osztály. Is that reasonable? -- Meyer (talk) 17:36, 12 September 2021 (PDT)

PBM, I don't mind. "Főosztály V"(even sometimes translating specifically as "general department") is inclusive of "Osztály V/h". Then the Hungarian system gets into further subclassifications, though not necessarily within our aero material.

Somewhere I'd emphasize variable degrees of a subclassification process. So classes would include such elements as the Dutch "Klasse", for instance, along with various analogous forms, with subclasses including the Dutch "Groepe", for instance, along with analogous forms. Etymology might be out the window in some cases, in terms of which word designates the broader grouping, with mere convention playing a role, but yes, we're getting to the thinking of the time. Patent classification systems could use a section on this, with ongoing speculations, as well as more outlined structuring of material being handled in analogous manners across the international context, with nationally specific term pages ordered in said section.

Airship - Dirigible - Steerable Airship

Speaking of etymology, relative to mere convention, and conventions swinging between languages, we know that English "dirigible" comes form French "dirigeable" (the adjective meaning "steerable" from "diriger" - "to direct"). A Hungarian translation might get us "steerable airship"(dirigible), but is there such a thing as a non-steerable airship? I'm thinking in terms of tracking proportionate data. Other terms, when we have them, lead to more legitimate precision. A blimp is specifically a non-rigid airship. Then again, these are also "steerable". I'm sort of tied up today, but do we need some disambiguation and possibly some eloquent and well-placed confession of ambiguity on these matters?

  • I think the term was meant to distinguish from older style balloons which (1) weren't shaped for efficient steering, and/or (2) didn't have any force of motive power to move them horizontally, and/or (3) didn't have fins or tail for control. At one time I thought "rigid" was part of the definition, but it isn't. I don't know where the sharp line is; I think of it as being the cigar shape versus the balloon shape, but maybe a dirigeable always has propulsion too. In modern use "airship" would always be steerable but I don't know about that then.
Regarding: do we need some disambiguation and possibly some eloquent and well-placed confession of ambiguity on these matters? YES that sounds constructive, if we don't already have it. There are books and other references to check on these terminologies. I just checked two books on the history of ballooning and didn't see a clear definition. en.wp treats dirigible as synonymous with airship, and says: "An airship or dirigible balloon is a type of aerostat or lighter-than-air aircraft that can navigate through the air under its own power." I like that, but we also want to establish what it meant back then. -- PBM 18 Sept 2021
  • I think some of the issues come up in terms of mere usage, and interlinguistic borrowing, antique and modern. "Blimp" even pops up from time to time in Hungarian-English translation, and that should definitely be non-rigid, though I'm not sure of each case. I like the above definition, and the general equation of airship=dirigible, rigid-non-rigid being great data when we have it. I'll put something on Techtypes about this sort of issue. -JRH
  • Excellent! -- PBM
  • There seem to be few uses of airship or aéronef before the era of dirigibility. Dirigible is really a less descriptive name and probably a shortening of "dirigible airship" (arrived at in English and French). Although as you suggest "dirigible airship" is redundant, it may have been been useful for communication purposes when the technology was new and the vocabulary uncertain. LTA (talk) 09:50, 19 September 2021 (PDT)

Gustav Knäpper

  • @LTA, can you find this fellow, Gustav Knäpper, in DPMA? I suspect he patented in DE before the patents we see. --Meyer (talk) 10:46, 24 November 2021 (PST)
    • Yes indeed, he had a German patent for the same compressed air propulsion device: Patent DE-1903-160742. LTA (talk) 23:01, 24 November 2021 (PST)

Edit with form deletes unused fields

It seems this still can happen; see this linked edit. This was a US patent with the upload glitch caused by multiple tech fields. When the data for "Assigned to" was deleted in the form that field disappeared. Furthermore the unintelligible "Full specification filed date=meteo" and "Application number=instrument" were invisibly deleted. LTA (talk) 03:00, 27 November 2021 (PST)

-I've been using edit with form somewhat more lately, since the issue between edit with form and accent marks got cleared up, and I've only revisited certain pages, with plain edit, in a few cases. What I'd been noticing, before and still, is that mere edit, whether or not I'd used edit with form, would be missing some or all empty fields. However, when I create a new patent page, from scratch, I cut and paste the entire long form onto a straight blank page, initiated by way of searching for the desired patent number and adding "Patent __-____-" as need be, and all the fields remain, whether I switch to and from edit with form or not. When I was using mere edit more, on pre-existing pages, because of the now-resolved accent mark issue I often had to add at least certain groups of fields, if not all (fields that had been unused). Any field that I've added, by hand if you will, remains, even empty fields that I would leave empty.

(I just did a blank field visibility test, going into mere edit, on one page I'd created from scratch, and one page that we'd already had and that I'd edited with form, and unused fields were missing from the latter, whether or not having previously used edit with form was the cause. For me, sans the edit with form accent mark issue, edit with form simply works, and I haven't done any manual field insertion in a while, and rarely use mere edit.) -JRH

Example patent not granted

@JRH -- Can you review Patent GB-1918-169183? My practice when a patent is not granted is to leave the grant year and grant date blank. What date are you using there? -- PBM @PBM -- I've tidied it. Oddly, Espacenet gives "publication data" 1922-12-19. -- JRH

  • Ah! That sounds right actually. The document was not granted patent status but there is a date it was published to the public. We can record that date. Interesting. -- pbm

Early aero-technical development analyzed as a social network

I'm done for the month(April, that was), officially, but kicking a few ideas around. Would some variation on the above be appropriate as a structurally facilitating page? Or something like "Early aviation analyzed as a social network"? I've squeezed simple social network onto a few pages, for now. I'm very keen on data display, and user-friendly integration, and on thoughts not getting lost.

  • Good! I can add to social network. Feel free to brainstorm. There's a formal literature in which a "network" has nodes (points) and edges (lines connecting the points). The nodes might be people, and/or companies, with substantive links, or they might be patents, with citations to and from one another. Or some other collection of information can be represented this way, and there are dynamic changes which can be "measured" by summaries of the nodes, edges, links, paths, and their properties. I have a likely coauthor who can go further here. -- Meyer (talk) 16:09, 27 June 2022 (PDT)
  • It just struck me that the above "node" and "edge" terminology, of the social network is directly analogous to the “nodes” or “artificial neurons” connected via positive or negative “weights” within the AI "neural network" discussion. They even talk about the challenges presented by surprise "edge cases" ! The linearly summed "weight" combinations fit into the profoundly non-linear "network". All of our data combination has this profoundly non-linear aspect.
  • I'm keen or retaining all layers of geographical data, referenced to the individual patent (or other source), including American county data. It's a bulk data layer auto-generated(?) via histpat, and it always checks out on the original, occasionally with two histpat county allegations and one checking out on the original, with British patents on rare occasions mentioning American counties. The automatic(?) histpat data gives us something, at least, on patents we haven't visited since whatever bulk data intake. Referencing to the individual patent tracks movement, even if the data on one patent is more precisely local than that on another.AvionHerbert (talk) 15:22, 27 September 2024 (PDT)

witnesses-agents nexus

  • with at least one case of an inventor acting as a witness to a patent of another inventor, with the same witness being of the firm acting as the same patent agent. We can zone in on the specific cases, in terms of content exchange-influence between the inventors in question. I hope this sort of thing feeds traceably around our other observations and case-making regarding the influence of patent agents on the innovative process.AvionHerbert (talk) 15:22, 27 September 2024 (PDT)

Engineering credentials of patent agents

Correspondence

  • Letters tie into communication between locations, of course, in addition to the intercourse between individuals.

AvionHerbert (talk) 06:41, 6 November 2024 (PST)

  • This would tie bulk geographically-based communication flow network analysis more abstractly.AvionHerbert (talk) 08:43, 17 November 2024 (PST)

Re-examination of the 19th Century advanced search page of inpi.fr

A quick re-visit, 1800-1850, cross-referencing various terms:

  • aéro yielded false leads
  • aér and aer yielded nothing (though they got us legit data in the past)
  • oiseau, as a wildcard, yielded nothing
  • ballon gets us GREGOIRE Gaspard(non-aero), and Henri Dembinski, and Joseph-Augustin Barratte, and Auguste-François Garnier(only semi-relevant), and LOUVET(only semi-relevant), and MILLET/LUCAS(which looks dodgy)
  • update from pbm: After visit to INPI.fr I may be able to get access to a giant spreadsheet of the data made available in their advanced recherche page, meaning we would not have to use the web interface to it. In a year it is expected that more information on these older patents will be available, including their elegant diagrams. -- econterms

Legal successors filing patents

The Hungarian word "jogutódja" translates as "legal successor". There are three Hungarian incidents of this:

There could hypothetically have been an honorific aspect to posthumous filing, though Joseph Hofmann was apparently not yet deceased. Over the three cases we have a company, a presumably educated (Dr.) landlord, another (Dr.)physician, and a factory manager, filing in what they likely perceived as being their own interest. Are these indicators relative to low quality patent phenomena? Rather, these third party filings are likely indicators of a patent's being of high quality.

There should be non-Hungarian parallels, of course. A spouse, or some other next-of-kin, could be a "legal successor". A creditor could be a legal successor. It is a matter of whether the third party in question pursued this entitlement, whether they saw the patent rights as a meaningful asset, whether they filed the patent.

  • I'm reflecting on whether we should make informal inferences. We could enter "legal successor" in cases where the original inventor died and a close relative then becomes the applicant/owner during the application process. That seems useful to me, if it matches the formal Hungarian category. Then we could quickly get a report on how common these cases were, how they were apparently handled, and so forth. You can just decide this. -- Meyer (talk) 12:30, 17 January 2023 (PST)
  • I would be in favor of our doing something with this. My guess is that the designation popped up in the Hungarian protocols responsively. Interestingly, a search internal to this wiki of "legal successor" brings up other disproportionately Hungarian results, aside from the patent-generated "jogutódja" term. In theory, aside from Hungary, anything analogous, in terms of patent filings, would come up from within a small subset of "Applicant is inventor? No" material. (talk) 13:09, 17 January 2023 (PST)

New compact templates for French and British sources

Colleagues, for patents which are both on espacenet and on google, I've switched away from: {{esp|FR|391218}} and its GB equivalent to this template: {{FRpatentsources|467096A}}. This second, newer template constructs and shows both links. The A (or E) suffix has to be capitalized or the resulting hyperlink doesn't work.

When add I find it convenient to go to the google one first, use what it has in text, and only then look at the details in the espacenet one because they are on a pdf. Google also sometimes has a count of post-1948 citations by later patents.

French patents before about 1904 aren't on espacenet or google yet but I think based on my conversation with the French patent office that they're coming. They are scanning them. -- Meyer (talk) 14:04, 21 April 2023 (PDT)

Noted! I've noticed a few, even pre-1900, trickling onto Espacenet, over time. Regarding very early French material, I've been using a "1969" internal search to find "|Filing date=December 31, 1969" and-or "|Grant date=December 31, 1969" glitch material that has come up via certain data shuffles. (Of course there are a few legitimate 1969 death dates and publications.) I'm correcting everything. Interestingly, the most antique French data gives more address and occupation data. -JRH 22 April 2023

Other compact templates for patent sources

  • Swiss patents: {{CHpatentsources|73161}} ; note the lack of any letter at the end.
  • Austrian patents: {{ATpatentsources|67267}} ; note the lack of any letter at the end.
  • German patents: {{DEpatentsources|118834}} ; note the lack of any letter at the end.
  • American patents: {{USpatentsources|884432}} ; note the lack of any letter at the end.
  • Canadian patents: {{CApatentsources|126457}} ; note the lack of any letter at the end.
  • Dutch patents: {{esp|NL|13391C}}

{{TOCright}} and other tools aiding in the accessible ordering and display of data

The {{TOCright}} function was of great assistance on the page of Robert Esnault-Pelterie, among others. Given that, I noticed that mere "=== References === " registers well relative to the above-mentioned function. I inserted === Patent data followed by publication data === so as to give This person had 0 publications and 0 patents in this database. 

a place relative to TOC. This does still require scrolling through by all the patents (and there will be more, in this case), in order to see the publications report. Is there a way to address this? I've used === Other standard format data === as a heading for the name, birth date . . . and so forth form.
Were you asking how to address the need to scroll so much? I'm thinking about how to shrink or hide the list of patents. REP is a special case. He may have more patents than anyone else. I can't think of too much to do here. On Wikipedia itself there is way to put a list into multiple columns, and there is a way to hide a chunk of tech inside a show/hide box, where you'd click "show" to see it all, and "hide" to see just an introductory line. I just tested and the second one doesn't work here. It would be cool to make it work, which may be possible. There may well be a way to get multiple columns to work.
One thing you can definitely do is to shrink the font. Put <small> above the standard reports, and </small> below them. Then the characters will be smaller. This doesn't save much vertical space, though.
I think your approach to making a header above the standard reports so that they have a section in the TOC is a nice idea. We could do it for every case . . . although I think I like the compact way it looks now usually.
-- Meyer (talk) 20:28, 10 June 2023 (PDT)

REP is a special case indeed. He has so much data of various kinds that the TOC brought some order. So I added the headings. In usual cases the headings built into This person had 0 publications and 0 patents in this database. 

are very fine, patents and publications both headed, but they are as if one function, relative to my inserted headings, and TOC. In this rare case, I was wondering about some off-chance that TOC could somehow offer the option of jumping straight to publications. I'm all in favor of not applying TOC features to the majority of inventor pages. -JRH
Surprised there is no obvious option for show/hide. It does seem like there are extensions that will do this, e.g. Extension:CollapsibleSections. LTA (talk) 12:20, 11 June 2023 (PDT)

Patent dates discussion

Patent dates could have its own talk page but instead allow me to pose my little question here. Back the 1860s, US patents don't have filing dates listed at the beginning—but they do have a date of signature at the end. Is that a filing date? Or does it occur too late in the process, after the lawyers have already negotiated with the office? On Patent US-1861-32182 there's a month and a half between signing date and grant date. LTA (talk) 06:22, 4 July 2023 (PDT) P.S. These signature dates are included in a minority of patents from this period. LTA (talk) 08:15, 4 July 2023 (PDT)

It makes sense to me to use the date the inventor signs as a filing date in those cases. In cases where we have both, I'm doing some comparisons; generally they are within a week of one another. We could leave a note saying we used the inventor-signing date, but I bet there's a simple breakpoint in the 1860s when the official filing date becomes available so it can be implicit, and they are close to one another. Agreeable? Sensible? -- Meyer (talk) 11:12, 11 July 2023 (PDT)
Sounds just fine. LTA (talk) 19:12, 12 July 2023 (PDT)

Regarding Filing date & Grant date & Publication date

Filing date

Fortunately, almost all international and proprietary points of interest revolve around filing date. This applies to additions to parents within each national system, always based on filing date.

Particularly when we have originals on hand, filing dates are the most straightforward. A Hungarian patent will have its "Date of filing of the application" within the modern entered data, and this will always match the “A bejelentés napja” date shown on the original.

Priority date

Filing date also applies to the international priority date phenomenon which connects to the Convention de Paris pour la protection de la propriété industrielle. The priority date, whether or not we have seen it, is the filing date of whichever nation the patent was first filed. The enthusiasm with which each nation complies with the above provisions or displays its compliance is highly variable, relative to these 1883 agreements.

  • A British patent, such as Patent GB-1908-11948, may head with "(Under International Convention)" and give the priority date and the nation in which said earlier filing was done.
  • An Austrian patent such as Patent AT-1910-43142, will usually have its "Priorität vom . . . ", and the appropriate date, and the nation in question, and more rarely the non-Austrian patent number.
  • A Hungarian patent, such as Patent HU-1909-48795, will have its "Elsőbbsége" followed by the appropriate earlier non-Hungarian date. For better or worse, their efficiencies do not touch and are not affected by any oddities evolved on Espacenet.
  • France seems to be in decent compliance, for the most part.
  • Spain is very difficult, in that it does have data which interfaces with Espacenet, yet only partially, and erroneously. We have every reason to believe that priority date data showing up next to a Spanish patent in a column of results is actually the filing date, in Spain, and not usually, almost never, the actual priority date.

Grant date

  • Patents are “accepted” in Britain, “Délivrée” in France. In the case of many other nations, "grant date" is a semantic toss-up. Fortunately, again, filing date is the most crucial data.

Publication date

This has come to us partially by way of neat analogues relative to certain national systems, as displayed on original patent documents.

  • Each French patent will have its date “Demandée” and its date “Délivrée” and its date “Publiée”, and this last will comport with the Publication Data as displayed on Espacenet.
  • Britain actually gets more complicated. When a patent has been accepted, its Espacenet Publication Data will match the date accepted (our "granted") as displayed on the original document. Patent GB-1919-144395 is one of the overwhelming majority of cases in which the "publication data" featured by Espacenet matches the "complete accepted" shown on the British original document. I'm also highlighting the patent as another rare neat example of multiple provisional specifications well-dated and numbered under the auspices of one complete specification document. When the patent has not been accepted, there will still be Publication Data. Patent GB-1918-169183 is a handy example of this latter.
  • Spain is very difficult, again, in that it does have data which interfaces with Espacenet, yet only partially, and erroneously. We have decent reason to associate the Espacenet Publication Data with the patent's being granted.
  • Hungarian patents have a “Publication/disclosure” date entered within the modern entered data, and this data always follows the filing date by some months or years, but this “Publication/disclosure” date is never shown on the original. Rather, a “Megjelent (year). évi (month) hó (day)-én (or án)”, that is “Released (month) (day), (year)”, is given on the original, and this almost invariably comes weeks or months or occasionally years later still.

Related to aircraft?

I don't see the criteria for "Related to aircraft?" spelled out at Template:Patent. Are they written somewhere else? A while back we agreed that projectiles (bullets and grenades, mainly) would count as 0.5 or partially related to aircraft. Going through the US patents I see many "propellers" some of which could be related to aircraft, but others which are paddle-wheels for boats and have little applicability. Maybe it would be a good time to spell out the criteria. Also, considering that we have not collected a similar sample of patents (for projectiles and non-aeronautical propellers) from countries outside the US, for formal data evaluations it would probably be desirable to include only patents with "Related to aircraft=1". LTA (talk) 08:15, 4 July 2023 (PDT)

Spelling out the criteria sounds fine. Also, our count of solidly "related to aircraft" patents is still artificially low, aside from the other issues. Aside from "Yes", versus "1", there are still significant numbers of patent pages on which the field is blank.-JRH

  • Great point, this is a big inconsistency in the data. According to "Drilldown":

    Related to aircraft: Yes (6990) · 1 (6193) · Keine (2745) · No (203) · Partially (181) · 0 (131) · 0.5 (89) · .5 (87) · -1 (80) · 1? (28) · 0? (5) · frame (4) · 1911-06-27 (4) · rudder (3) · marine (3) · 1911-07-04 (3)

Within Keine here is the national breakdown:

Office: Keine (5) · ?? (1) · AH (1) · AT (18) · AU (4) · BE (26) · CA (5) · CH (4) · DE (32) · DK (3) · ES (11) · FR (274) · GB (619) · HU (5) · IT (4) · LU (1) · NL (1) · NO (1) · NZ (1) · SU (2) · US (1727)

Combining "Yes" and "1" shouldn't be a big deal but coming to terms with these blanks may take some effort. LTA (talk) 04:11, 5 July 2023 (PDT)

Criteria for partially related patents

It seems we agreed that 'projectiles' are 'partially' (0.5) related to aircraft. What about windmills? Right now, based on drilldown, it seems like we learn towards 'partially' on those. What about engines? LTA (talk) 15:59, 24 October 2024 (PDT)
Another case: Patent US-1910-1089517 is 'hydro-pneumatic' apparatus. The patent doesn't appear to mention aircraft. But its primary CPC classification is CPC B64C25/60 for shock absorbers in aircraft landing gear. I marked this 0.5 but I could see arguments for 0 and for 1 as well. LTA (talk) 20:07, 5 November 2024 (PST)

Tangential note on projectiles

Most of our "projectile" patents came from a big batch of US patents I did offline a while back. I classified these almost universally as "military; projectile" etc. It seems these were all mangled by the upload (e.g. Patent US-1900-656934) so that the projectile term doesn't appear in Tech fields, and only one patent shows up in the report at projectile when in fact there are dozens. I wonder if it is even worth the time to salvage these, when we don't have comparable patents from any other country, or whether we should put these poor entries out of their misery. LTA (talk) 15:30, 24 November 2024 (PST) Long ago I added somewhat of a 19th century bulk of French projectile patents as well, which also aren't crucial. I believe the thinking was that they tied in via aerodynamics. Regarding reports on tech term pages, I've noticed some oddities lately, quantities not showing up, that is, even pertaining to patents the patent pages of which themselves don't seem to have any other problems.AvionHerbert (talk) 11:03, 28 November 2024 (PST)

Rename Luft-Verkehrs-Gesellschaft?

Right now the page title is Luft Verkehrs G.m.b.H.. This form seems rare in the wild and I propose moving the page and standard company name to Luft-Verkehrs-Gesellschaft. I suggest this, tentatively, over Luft-Verkehrs-Gesellschaft m.b.H. since it seems there is an AG form as well which we might want to include with the same name. Gesellschaft was used as part of the name because "Luft-Verkehrs" means "Air Traffic", so, I infer, it doesn't read; it needs to be "Air Traffic Company". LVG is a standard abbreviation.
Any thoughts? We already have two dozen patents associated with this company and I suggest we change the names en masse to either of the options above. Luft-Verkehrs-Gesellschaft m.b.H. is what I see usually appearing. The search on the main page should be set up to catch a few variants, in any case. LTA (talk) 12:15, 11 August 2023 (PDT)

Luft-Verkehrs-Gesellschaft sounds fine, with a few caveats. "Luft-Verkehrs-Gesellschaft m.b.H." shows up on a spot-checked German patent, with the "G" in "LVG" and in Luft Verkehrs G.m.b.H.(in all its variants) being equivalent to "Gesellschaft". "AG" usually equals "Aktien-Gesellschaft", though occasionally "Arbeiten-Gesellschaft". So very technically, "AG" may involve a different organization type. If you have any observations on any related minutia, feel free to jump in on company types, by the way. I'm still in favor of consolidating results on one page, even Luft-Verkehrs-Gesellschaft. I'm of the school that notation, on patent pages and on the organization page, is what makes the unified bulk display kosher. -JRH

Luft-Verkehrs-Gesellschaft works for me, or LVG. Consider also Luftverkehrsgesellschaft, the phrasing on de.wp. My understanding is that "Gesellschaft" is usually part of the company name, whereas "m.b.H." is a legalistic suffix, and when one sees GmbH it is sometime all a suffix, and the G is not really part of the company's common name. (I will visit company types to help establish this in our common records.)
I am happy to edit other pages to match the rename. In most cases I'd want a patent page to show the Applicant firm that was on the original patent, but if that leads to too many names, and a complicated query/report with more than four similar names, or too many redirects, or any mess, it's okay to tweak the Applicant firm to match the name of the company page. -- Meyer (talk) 09:12, 12 August 2023 (PDT)

I'm also most in favor of keeping what is displayed on the documents we see on the record somewhere, including on the patent pages, and not getting too carried away with presumed generalizations. G.m.b.H. is an abbreviation of “Gesellschaft mit beschränkter Haftung”, and there is some inconsistency in terms of what is part of any common company name. AG, in particular, opens at least the possibility of a significant change in corporate status. One way to avoid mess, in terms of excessive reports on particular pages, is too retain at least some of the pages, with friendly and prominent though non-obtrusive mention, linking between them. We don't always know what insignificance, or significance, is behind the differentials in usage. -- AvionHerbert (talk)

Detailed disambiguation of names

The detailed lists for Phillips surname disambiguation and Thompson surname disambiguation make sense. Deep researchers in this historical field benefit from this sort of thing. Thank you, AvionHerbert. -- Meyer (talk) 08:40, 21 September 2023 (PDT)

You're welcome Meyer

Patent FR-1911-331197(?)

Presently, Patent FR-1911-331197, as is, links to a 1903 non-aero patent, unconnected to André-Julien Mahoudeau‎‎. I wasn't sure how the lead got entered into our data, so I didn't want to mess with it.

  • Got it. Moved to the name of another incomplete Mahoudeau patent. This one might have appeared because of my typo in Dec 2023. -- Meyer (talk) 17:34, 19 January 2024 (PST)

Also, Patent FR-1856-mo may possibly, though only possibly, check out as a Patent FR-1856-26857, though only possibly of Modeste-Abel Latouche. I ran into Internal Server Error when trying to save some speculations:

Note: A "système de navigation marine, sous-marine, etc." does show up, with this filing date, simply identified with inventor LATOUCHE. It would be designated as Patent FR-1856-26857, if we were to find that it has LTA applicability. This is all feasible, and with the inventor being Modeste-Abel Latouche.

This 26857 is a patent of fifteen years. This "LATOUCHE" is at this point addressed at 14, rue de Provence, Paris, département Seine, with occupation "secrétaire de la société de navigation de Paris (75056)", Classe 6.

Works of Sam Leonard Walkden

@User:AvionHerbert & @User:LTA -- you've both edited about Sam Walkden this year, which inspired me to add all his works listed in the 2nd Brockett bibliography. I'm done. If inspired, take a look at any of that and edit freely. LTA, I do think the two editions of his book should have their own own page since the second one is 90 pages longer and was published in a different year. That suggests that if someone were someday to summarize or review them, they would have historically-informative differences. So I split that page into two. I hadn't known of Mr. Walkden till now, but he seems to have been quite a focused aerodynamicist, not a journalist passing by.

FYI, I have 17700 other entries from that Brockett to add, too many to paste in one by one. There are severe glitches with non-English characters when I do a mass upload, and it takes effort, but I'll try again. -- Meyer (talk) 10:06, 26 December 2023 (PST)

It seems logical that, if a three sentence blurb in a magazine sometimes warrants its own entry as a publication, ninety pages of extra text would also deserve that. Having two entries would also enable one to clearly cite one or other other. On the other hand perhaps it would get excessive to form a separate entry for every edition of a book, as a rule.
Along the same lines I have been thinking that the Reports and Memoranda of the Advisory Committee for Aeronautics probably each deserve their own page, as they are major research contributions, each with a unique author, date, and subject. It seems a bit scandalous, actually, how hard these seem to be to find.
Now that I look at it they seem to be indexed separately in Brockett, 1921 so maybe the work of cataloguing them has mostly been done for us, already. LTA (talk) 15:24, 26 December 2023 (PST)

Aeroplane

I've added observations onto this page vis-à-vis the sprawl of interlinguistic usage tangent to "aeroplane" as a name of an entire vehicle. That is, I understand the etymology, but our data entry hasn't taken this consistently into account. Tech field entry "aeroplane" may sometimes reflect wings in particular and-or what we call airplane. I thought we should have some of this made explicit, on page aeroplane, in that ambiguity of usage, and often outright leaning to the whole-vehicle usage, pertains to a multitude of publications, along with company names, not to mention patent and other data. Likewise with hydroplane and hydro-aeroplane ; in the sprawl of casual usage, even high-end antique writers seem to have used "hydroplane" the "hydro-aeroplane". We are dealing in cases like these with multiple linguistic particles, in conjunction, each and any of which particles underwent abbreviation, and evolving signification.

Editor Vorobiev/Vorobyov

@User:LTA, it's exciting to see your findings from the Russian journals. Boris Vorobiev seems to be a central figure. I'd seen his name in Brockett but didn't know of his editorial role till you found it. He may have known the whole Russian scene.

It seems that this last name is common in Russia, and that there are 8 notables in Russian Wikipedia named Boris Vorobiev (more or less), but not ours: [1]

I'm jazzed to see a little bio of him based on sources you've found. For Cyrillic names it may be that we need expanded versions of "Standard person reports" with 10 or more alternative spellings, but that seems fine. The wiki is good at that. If I knew how, I'd write a generalized one that could take a large but varying number of arguments. There's a technique for that, detecting that the Nth argument is blank and adjusting for that. Will do it someday. Anyway it's fun to see inching forward in Cyrillic and to identify this person I can remember. It brings structure and coherence to scattered facts. -- Econterms (talk) 08:43, 3 February 2024 (PST)

I've got a copy of The Transliteration of Modern Russian for English-Language Publications, if we want to bring that into the proceedings. It gets into pre-Soviet and Soviet differentials, along with issues of dates and so forth. -JRH

  • Boris Nikitich Vorobiev is up and running with some related articles created as well. His papers are held at the Archive of the Russian Academy of Sciences and do contain correspondence, etc., with other significant people. One way to find more info short of visiting would be to see who's used this archive (fond, as they call it).
Yes it would be great to automatically 'multiply' some of these reports. I'm hoping we will get into some kind of a groove with orthography and not have the same person spelled 10 different ways in templates throughout the site. But it is undeniable that we're dealing with names spelled multiple ways in their own alphabet, in addition to all the possible transliterations! That book sounds quite applicable!
Guaranteeing some unpredictability, too, is the fact that plenty of these Russians had opportunities to spell their own names in Latin alphabets, and did so according to their own preferences. For example General Рыкачёв has -ёв (sometimes -ев, sometimes -евъ) at the end of his name like Vorobiev but signed his name "Rykatcheff" (even in English publications) instead of Rykatchev or Rykachov. LTA (talk) 15:49, 4 February 2024 (PST) P.S. Special:Drilldown/Person?Countries=RU to quickly survey the forty RU people. LTA (talk) 22:32, 4 February 2024 (PST)

Ѵ, eliminated 1917-1918, possibly factoring into Russian patent classifications

I'm currently digging into Patent RU-1906-21016. This Ѵ issue pertains more likely into the character after the "/", rather than to that preceding it.

One hypothesis is that "Group V/VIII" = RU Group V and RU Group VIII. This would be the pre-1910 classification system. There are some notes at Russia, Russian patent office, and ru that pertain to decoding the patents, and where new information could be added. Possibly there should be some redistribution of material from the main Russia article to the patent office article.
For what it's worth I've seen tons of ѣ, і, and э, pretty minimal ѳ, and no Ѵ in materials mostly 1900–1917. LTA (talk) 20:18, 8 February 2024 (PST)
PS I think these listings in Notes of the Russian Technical Society more or less confirm that the group names are traditional Roman numerals. The perception of the typed Ѵ there is astounding though. I will see if these listings will also confirm dual group classification. The headers of the categories are pretty vague and hardly descriptive of the inventions we're looking at, which probably explains why they switched to the new system. LTA (talk) 22:40, 8 February 2024 (PST)

Yes, and we are learning that administrative peculiarities practically destined to trip up the archival visitor might have offered no confusion to people engaged in the nationally specific administrative culture at the time. That is, I would never tilt a Roman numeral V so that it could even possibly resemble the Ѵ, but in context, at the time, for the caste involved, it wouldn't have presented any confusion. Seeing the two "fonts" together in one entry did stand out. I'm happy to play ball with the V/VIII until and if we find any data to the contrary. -JRH

File:Golubev-1909-GroupV.jpeg – again the Ѵ. Does this have to do with ease of typing on Russian typewriters? LTA (talk) 09:08, 10 February 2024 (PST)

Patent families with absent hypothetical original

In some cases we know about an original patent but we don't have complete information about it, in which case it gets a name like Patent BE-1909-12-24 Marie Jasogne and Theodor Dobresco. Many such names are redlinked but we could choose to create limited entries for these patents because we know they exist and therefore they are valid data. The remarkable Patent US-1913-04-21 George Raymond Lawrence also falls into this category, though it may never have been granted, because it's cited in foreign filings.
In other cases such as Patent RU-1909-27293 of Todd Churchill Woodworth there's no reference to a US original though it seems likely one existed. There is a three-way tie between RU, GB, and FR, for first filing on the Gregorian date of January 12, 1909.
We would like to tie all of these patents as a family, would we not? Should we arbitrarily assign one (I'd lean towards GB in this case) as the original and declare the rest supplementary? Would the arbitrarily chosen original then have |First filing=1, all the rest |First filing=0, and as family year the filing year of the arbitrarily chosen original?
Tangentially: the guidance at Template:Patent/Supplementary to is quite helpful but is there another place on the site, possibly reader-facing, that deals with the concept of the patent family? Patent family? Ciao, LTA (talk) 22:40, 10 February 2024 (PST) I'm keen on creating more of those parent patent pages, with unknown numbers and so forth, as needed.-JRH

Retention of IPC and CPC classifications which have been removed from particular patents on Espacenet

Long ago, we set this, the retention, as somewhat of a precedent, when revisiting patents. I'm wondering about this. Aside from the great interest in the antique classification systems, it seems that said retention is a preservation of data which the team at Espacenet has deemed imperfect, a preservation of "erroneous" which is also inconsistent and undocumented in its nature.-JRH

We've decided to remove, when Espacenet has removed, on a "moving forward" basis, when revisiting the appropriate patents in the course of other things. This has to do with among things the use of links, as "sources", of data points which are no longer verified via the link.-JRH
If we no longer have a source connecting a patent to some IPC/CPC, it's sensible practice to remove our link to. But this is a judgment call; if it seems useful to keep it one can leave a note saying why. This issue comes up with the somewhat mysterious '2700' CPC classes that appeared in some kind of transitional phase. -- Meyer (talk) 16:55, 24 February 2024 (PST)

Corporate evolution cases

  • Siemens & Halske Aktiengesellshaft seems to have been solidly corporate in 1900, getting into aero much later.
  • George Holt Thomas, bulk filer, collaborating with "lesser" filers, though they might have been more key to the innovation ; occupation data all through the complex tangent to his collaborations, feeds into this, with smaller "complexes", tangent to this one.
  • Westinghouse, vastly corporate, international, tying into aero, partially via Leblanc
  • Atelier, the page, in terms of theory, and "Atelier" and "Ateliers", as keywords, practically by definition, zone into cases on the innovation-industrialization cusp
A central case in France would be the Grands Ateliers aérostatiques du Champ de Mars, an internationally known workshop in Paris which eventually transformed into Société de Constructions Aéronautiques / Société anonyme Astra. The Compagnie générale de navigation aérienne, which held the Wright license in France, was also part of this corporate family. Chadeau may have relevant information and analysis. LTA (talk) 04:33, 3 April 2024 (PDT)

A structured approach to patterns of non-aeronautical innovation as it connects to aeronautics and aviation

Patent US-1914-1360694, of Elmer Ambrose Sperry and the Sperry Gyroscope Company, makes reference to an earlier Sperry patent. It also, by way of serial numbers, makes reference to two other American patents, which were incidentally filed after Patent US-1914-1360694, though filed before the granting of Patent US-1914-1360694, and of course before the printing of the document. These patents may have been filed by neither of the Sperry inventors, and not even necessarily by any inventor working with Sperry Gyroscope Company. The patent making the reference is solidly aero-applicable. One of the patents to which refers is a compass, not necessarily aero-applicable. The other tangents further into electronics. Particularly with the serial number factored in, I didn't find the patents in question, and maybe we don't need them.

I'm interested in:

  • a structured approach to non-aero data when it comes up tangent to aero-applicable data, a structured approach to threads of this sort that we find, and have documented, even if we don't do the follow-through into the non-aero ; I'm wondering if there's a way for us to tag these sorts of things, treating the phenomenon as a variable, to be processed algorithmically, for later display.
  • the fact of a patent document's making reference to later-filed patents, though this is rare, the patent document therefore taking on a retrospective aspect and highlighting a cross-fertilization between technologies ; this also highlights a systematically proprietary and governmental tracking of the innovations generally, and of the cross-fertilization between technologies. This is more nuance in the nexus between administrative culture and the grander principle of innovation itself. -- AvionHerbert

Honing our interpretation of German patent administrivia

Examining the weekly reports published under the title Patentblatt shows the different phases of the process more clearly.

  • A) Upon filing patents are given a code which usually seems to have a letter and a five-digit number. It has recently dawned on me that the letter is the first letter of the applicant's last name. I don't know if the numbers are serial within last name categories, or reset after a period of time, or what. It doesn't hurt to record these numbers within our system as they are useful for tracking applications before they receive final patent numbers, or if they never do.
    • It would make sense to record this code in "Application number" but this field seems to be used already, when the data are imported (from espacenet?) The number preloaded into this field does not seem to add much, e.g. for Patent DE-1902-140705 the Application number field has "140705D ". Application ID is also populated. Should we overwrite "Application number"? Put this code in "serial number"? Or record it outside of the template?
  • B) Patents are indeed granted ("erteilt") before being published ("ausgegeben") typically with a delay of over a month. The Patentblatt, always published on Wednesday, lists granted patents released on the Monday of that week in the Reichsanzeiger.
  • C) It would be convenient if we could make a blanket statement that the filing date is always one day before the "Patentiert im Deutschen Reiche vom" date appearing on the published patent. On Patent DE-1909-228073 the "patentiert vom" date is a Sunday indicating Sunday is not skipped for the purpose of this system. Is there a date when this one-day-delay system begins?

Thanks in advance for any comments or insights on these or other matters related to the German data; the earlier we have a clear interpretation the more consistent our data will be overall. I am hoping to discover a number of missing German parents from international families, including possibly applications that were not ultimately approved. Ciao, LTA (talk) 14:55, 25 June 2024 (PDT)

This is all noted. I'm glad to hear of indication that Sunday (and presumably Saturday) is not skipped in the "Patentiert im Deutschen Reiche vom" being one day after the German filing date. I've been employing that protocol, and referring to patent dates whenever I touch German patents. I've never noticed it not applying or being contradicted by internationally cross-referenced data. Would ein Putsch into the earliest German patent data be a way of zoning into the starting date of said protocol? -JRH
If page patent dates were to be re-structured in any way, moving some of the German administrivia elsewhere, could some trace of the "Patentiert im Deutschen Reiche vom" observations be left on the page? I've loaded the site with references thereto, based on that. Also, the blurb on the old reasoning behind using year granted had left a few things out, and has become dated. Should we remove that blurb? -JRH
Please feel free to continue organizing and editing that page if you feel so inclined. LTA (talk) 14:27, 6 July 2024 (PDT)
Glad to learn from the Patentblatt. Thanks for looking in detail at the German ones, which I have found tricky.
(A) It might be good to include the code with the letter and five-digit number in our records. Let's at least document it. But I observe that we are tracking a variety of serial numbers, applicant numbers, INPADOC ids, and more, and that we get nothing out of them yet. It's work. So maybe document a few but let's not put them in the template yet until we have more free time.
(B): I didn't know this! Please document these grant and publication dates, under erteilt and ausgegeben. I've pushed to treat ausgegeben (issued) as meaning granted, and it sounds like that's wrong. If we should be moving the data between fields, all right, let's do it. But on the patents themselves we only see ausgegeben. There might be an analogue to other patent systems in which a gets approved on the technical side and is then "granted" only if the filer pays a fee. Some British patents seem to have been approved but then aren't granted because the filers don't pay a "sealing fee". Could that be what's happening here? Actually I am on good terms with some pros in the German patent history business and could ask once we have precisely formulated questions.
(C): The one-day difference between our data sources mystifies me, but seems stable. I could ask about it. -- econterms
Suggestion: Pages about German patent lingo might go into this category so we can rediscover them and knit them back together: Category:German patent administration‏‎

Confluences between various techfields and other terms

  • Pivot seems to be more or less interchangeable with ball and socket (or ball-and-socket), and these seem to lean into the juxtaposition of portions of aircraft, such as rigid portions relative to non-rigid portions (even as these terms apply to heavier-than-air). They lean into flexibility. The keywords pop up across publication and patent classification data. -AvionHerbert

Tech issues to address in fall 2024

I'm swamped for a few months but in the fall let's upgrade our MediaWiki and Cargo and other extensions. It's stressful and perhaps risky to do that and requires focus. We have some specific broken things that could be fixed by that set of upgrades. Please add to the list so we're watching the right issues when we attempt it.

  • Drilldown gives a bad error: [2] (Template:Patent, filing year field empty)
  • Cargo "join" function doesn't work, but perhaps would work after an upgrade. We need this to get some indirect queries to work, as they apparently do on other sites
  • In that same late-2024 window I think I can upload lots of publication entriess, hopefully finishing off the giant Brockett 2 bibliography.
    • This would be tremendous!
  • Edits by template delete empty fields and overwrite others (1 changed to Yes etc.); likewise uploads don't automatically use the full template. It's a minor issue but a widespread one; can we (and should we) cause the system to favor the use of the full template by default? And—probably a separate question—can and should we get the template editor not to automatically modify fields that aren't being edited?
  • We agreed on a few changes to the Patent template, here: Template talk:Patent

How is the "Other techtypes related to ______:" function affected by evolutions of usage ?

AvionHerbert (talk) 08:02, 17 July 2024 (PDT) ; this would have to do with Espacenet IPC and CPC data entry which we don't control.

Consolidation of page hydro-aeroplanes and page hydroaeroplane

For now I'll simply make the connections work, by adding hydroaeroplane as needed. AvionHerbert (talk) 19:40, 4 August 2024 (PDT)

I changed the page to a redirect; the main page already has results for the hyphenated term (and shows the two patents that were coming up on the report for the other page). Is that the outcome you had in mind? LTA (talk) 19:49, 26 August 2024 (PDT)
It was the plural (and the hyphen) that cropped up somewhere, and I wasn't sure if it all got connected. AvionHerbert (talk) 13:00, 1 September 2024 (PDT)
It makes sense to add the keyword hydroaeroplane on source pages that have aerohydroplane, which presumably means the same thing. We can make redirects to treat these as synonyms. -- Econterms (talk) 16:23, 16 October 2024 (PDT)

"New Espacenet" versus "Classic Espacenet"

  • I'll be laying off for a few days, but I got an email from Espacenet. Patent CH-1910-54062 is another case of a modern type-entered title, on Espacenet, not matching the aero-relevant original. It will likely be fixed within a few days. In the meanwhile, and with practically no delay, they sent me a link to the same case, on "new Espacenet", which is fine, though it seems to have the diagrams and not the text pages. They've told me dodgy stories before, alleging that on some "new" or otherwise special Espacenet, that all families are based on the title of the English-language patent in whatever family. Of course not quite all patents are in families, and not all families include a patent in the English language, and I've never seen a non-English aero-patent title functionally translated into English, but rather English language titles that have nothing to do with the original. They always fix it. The above escapades just add a step to the process. I always thank them for handling antique material, knowing that they are dealing with huge volumes of new material that carries a lot of contemporarily proprietary weight.AvionHerbert (talk) 13:10, 19 September 2024 (PDT)
  • "Classic Espacenet" is set to be decommissioned in the course of the year. The new Espacenet may be an improvement in some ways. This will likely affect the programmatic structure of the necessary reference links. Espacenet entry page and Espacenet Advanced search page AvionHerbert (talk) 06:20, 20 September 2024 (PDT)
  • All right. I'd hoped to stay with Classic but we'll adapt. -- Econterms (talk) 15:07, 20 September 2024 (PDT)

Brazilian patent info

It looks like Brazil's INPI put up information on historic patents in 2023.

Have you seen this before? I haven't seen how to get to to the underlying documents yet, but it's clear there are foreign filings in Brazil and I'd love to see them. -- Econterms (talk) 05:09, 3 November 2024 (PST)

French serial number

Just noticed at Patent FR-1917-493411.22317 there's a confusing reference in the patent text to le "brevet principal déposé le 13 juin 1916 sous le no. 84.082". In the header the original is given as a more familiar looking no. 493411. The latter was in fact filed on 13 June 1916. Could this be a rare reference to a French serial number assigned before the final number to identify a pending application? LTA (talk) 20:56, 4 November 2024 (PST) It very well could be something similar, and possibly analogous in some underlying way, with the fluke of extraordinarily rare mention on patent documents, or it may not have been consistently used. My interest in American serial number phenomena has to with the American system pointedly and publicly calling the numbers serial numbers, and using them as apparently the sole means of one American patent making reference to another, in the cases of patents' being "renewed", and in the cases of patents' being "divided", as we know. We deal with various numbers which are generated relative to some series, or serial processes. I've mentioned this French case among oddities on page serial number.AvionHerbert (talk) 00:11, 27 November 2024 (PST)

HistPat, American county data, along with matters of England, Wales, Scotland, and Ireland

  • Is there a way to re-import this data, directly into field "|Inventor location="? The data is overwhelmingly county data, and it checks out overwhelmingly restive to the original documents. My hunch is that peak data would cluster around counties we'd associate with particular major cities, with second and third tier clusters occurring around further-flung counties, with dynamic and unsought (and thereby non-bias-confirming) relations being displayed between the alpha clusters and the known major cities. It would secondarily aid in hinting at location change on the part of the inventors, and, once we get patents factored into reports displayed on location pages, it would offer a bulk of displayed volume, chronologically-ordered, with various other relations being explicit based on the nature of the display.AvionHerbert (talk) 08:01, 5 November 2024 (PST)
We could roughly try this using a find-replace technique but there may be a cleaner way to do it using the original spreadsheet that was uploaded. Meanwhile do we have any refinements of the standards for entering American locations?

|Inventor location=Dickinson, ND; Stark county; North Dakota; FIPSloc=38089

Would "Stark county, ND" be better? Or going the other way, it could be

|Inventor location=Dickinson; Stark County; North Dakota; FIPSloc=38089

based on the premise that the terms can be used together. Is this a good way to include the FIPS code? Could it be the bare number? Or something else? What is the loc? FIPS comes up in web searches but searches for "FIPSloc" lead mainly to Inventing Aviation.
Also, meanwhile, reports of patents by location will become possible once we recreate the table. A table replacement was initiated about 48 hours ago and it's been hanging at ~15,740 (out of 17,812) all day. LTA (talk) 19:40, 5 November 2024 (PST)
If I may add to the above general discussion under this header, I tentatively suggest that England, Wales, Scotland, and Ireland be included within "Inventor location" (rather than "Inventory country" or not at all). Although they are officially countries it makes sense to place them within the "country" United Kingdom (as represented by its official abbreviation, GB) as the administrative level which also operates the patent office. It makes sense to include them since they are they largest administrative subdivisions of the country and, while England will account for a large majority of filings, searches and reports using the other three countries should prove meaningful. LTA (talk) 14:35, 24 November 2024 (PST)
  • Regarding American states, I was initially semi-favorable to the postal codes, then, in some contexts, I noted that, in isolation, problems would come up. We wouldn't call California "CA", for instance, as a separate element, because of CA-Canada. I realize that this wouldn't come up with San Francisco, CA, for instance, as a combined element. I'm a bit interested in the factor of non-Americans seeing the data, matters of presentation. Ultimately having |Inventor location=Dickinson, North Dakota; Stark County, North Dakota; North Dakota in the field would save and connect the most data, redundant appearances aside, but, if

    |Inventor location=Dickinson, ND; Stark county; North Dakota; FIPSloc=38089

    , for instance, would allow the bulk importation, that would not be bad.AvionHerbert (talk) 00:45, 26 November 2024 (PST)
  • I'm all for England, Wales, Scotland, and Ireland, perhaps in location, leaving GB in "country" as an administrative concession to Espacenet and so forth. I have nothing against UK. I'm most keen on the location field, and particularly in patent-by-patent response to data as it is presented on each original document, translating French "Angleterre" to "England", for instance, and putting it in the location field, would be a very modest move.AvionHerbert (talk) 00:45, 26 November 2024 (PST)
  • Retaining all the locational data elements, from each document, as specific or as general as they may be, and never projecting specific but arbitrary data onto patent pages, is very important, in terms of legitimate referencing, in principle, for one thing, and in terms of locational changes which we can chronologically trace, and viably reference, and compare in the aggregate.AvionHerbert (talk) 00:45, 26 November 2024 (PST)

Approaches to the new Locations system, continued

There are diverse issues that will come up as we develop and coordinate Locations so I'm adding some structure here rather than continuing to mash the topics together.

Designations

At the lower level, do all of these fly: borough? municipality? village? At the higher level do we accept the UK's designation of England, Wales, Scotland, and Ireland as "countries"? Is "Württemberg" properly designated a Kingdom? (With the German Empire?) How standardized do these need to be? Partly I'm asking because it grates a little against our designation of the patent-office-having administrative levels as "countries".LTA (talk) 15:22, 26 November 2024 (PST)

I'm keen on all these items. That "levels" differ in quantity, and complexity, and occasional intranational inconsistency, across and within the international context, is something reflected even on patent documents. I don't see a problem with adding the words, factoring the wildcards into our thinking behind it. This will add flexibility to the structure, and the "patent-office-having administrative levels" will still exist as structural facts. Even "nation" might not clear anything up. So I'm conservative in terms of far per se standardization gets pushed, and-but I am keen on new words, whether or how far we flesh various of them out. We're ultimately dealing with all the factors of semantics, and variable usage, not to mention translations. I implicitly assume that a New York "borough", for instance, will likely not have any precisely analogous relation to some German locality translated as a "borough". I even see American and British counties as not very neatly analogous, but they don't have to be, and some "countries" have patent offices, and others don't. None of this can be avoided, but I don't see a problem.AvionHerbert (talk) 23:18, 26 November 2024 (PST)

It seems fine to have all these levels. They will be different for different countries but that seems okay. -- Econterms (talk) 23:26, 10 December 2024 (PST)

Patent agents

Having just looked at a group of Chicago patent agents, I am reminded of the importance of the agent's location, as we often see applicants using agents in nearby cities. However we don't have an agent template that would be usable to generate "Agents by location"—and it may not be worthwhile to create one. Just wanted to record the idea of this possibility. LTA (talk) 21:39, 9 December 2024 (PST) Yes, we often have good location data on agents, British, of course, Belgian and Swiss at least down to the city level, variability on the American front, and with patent agents tying into a complex on other levels. I think of it every time I touch a patent agent page. Without the semi-unplanned report displays that flow from template features, then someone would have to "drill down", aggressively, seeking the pertinence in the furtherance of some a priori agenda related to location. A standardized template-based approach, and the resulting displays, would bring smaller players into the conscious spread, with other relations popping out even during the processing. I'd be all for it.AvionHerbert (talk) 10:19, 10 December 2024 (PST)

OK, it would be good to get those reports and logical to use that segment of our data. Before we go creating a new table perhaps we could take the radical viewpoint that patent agents are Persons? But the problem there is that some are Companies and really the status as Agent is more important than membership in either of these other groups. So perhaps a new table would be the best approach. LTA (talk) 21:03, 10 December 2024 (PST)
Agreed. My mind is elsewhere but feel free to create such a table. It is okay to have two templates on a page -- a Person and Patent agency, or a Company and a Patent agency. Then the page shows up in two tables, which is okay. What else could be attached to a patent agency, aside from name/names, country, and location(s). There are a few with multiple locations but I'm not sure there were any international firms. They have some evolution of their own, with changes in name, mergers, etc. We might treat these as alternate names of each agency, analogous to other alt names. In some cases we might treat a merger as the creation of a new firm.
I anticipate publish-able research coming from our information on patent agents someday. I haven't seen a real study of this industry in our period. I understand Gabriel Galvez-Behar studied it in a narrative history in his thesis and book. Our data can be more broad and statistical. I'm also curious if we can ever show that inventors get some advantage from having patent agents nearby, or having contact with them before entering the aero field. -- Econterms (talk) 23:26, 10 December 2024 (PST)

A multi-table option sounds good, or a new table, with individual agents sometimes merging, and firms occasionally splitting. There are a few British patent agent firms with offices in New York, Haseltine, Lake & Co. and Alfred Ernest White, maybe others, making them international, though the data has generally come in via British patent documents, so they would come up in association with multiple location. There is a complex between individuals and patent agencies and companies, indeed, with some firms filing, themselves, and facilitating the filing of others, using the same corporate designation, with phenomena of internationalism gaining in documentation and structure.AvionHerbert (talk) 07:07, 11 December 2024 (PST)

I will create a template and table next chance I get. Name, Locations, Affiliates? (Possible affiliates? Affiliates is too strong, I want to suggest other people and firms that may have overlapping membership.) LTA (talk) 07:31, 11 December 2024 (PST)
Got the ball rolling on this, Template:Agent, see Washington, DC for a readout of Agents by Location. I'm not committed to any of the wording used here so please feel free to make or suggest changes. LTA (talk) 23:59, 12 December 2024 (PST)
I think the inclusion of people and firms which may be connected is great, with the related tentativeness usually being articulated on the related pages. I've test-inserted "|Accreditations=A.I.Mech.E." onto page Alfred Ernest White. This and related/semi-analogous matters of status pertain to patterned technical know-how, beyond litigious clout on the part of agents.AvionHerbert (talk) 07:17, 13 December 2024 (PST)
Brilliant! Thanks for bringing it to life. Accreditations is ideal! I added it to the Agent template. And it will make sense in a Person template too perhaps! This will be another way to track kinds of change over the technological history -- more certifications over time, I expect. -- Econterms (talk) 20:30, 13 December 2024 (PST)
I'd be all for it also being included on the Person template.AvionHerbert (talk) 00:46, 14 December 2024 (PST)
I'm ruminating on how to frame it for Person. We want to include perhaps in the Person version of the concept . . . work-related certifications . . . academic degrees . . . but also memberships in professional societies, or not? One issue is whether to use the word Accreditations or something else. I'm not jumpy about that -- the right word might not exist, and this word is okay -- but let's decide jointly what we expect to include there, covered and what word to use. Culturally I think the German tendency at that time would be to emphasize a certification, and the British tendency might be to count membership concepts more. I'm thinking we want to include both? -- Econterms (talk) 08:25, 14 December 2024 (PST)
In our modern anglophone usage "accreditation" hints most directly at academic degrees, with the German Dipl.-Ing. fitting in neatly, and being a great link across the inventor-agent spread. Some of the British "status" labels also cross this spread (whereas Chartered Patent Agent, for instance, does not, of course). We're getting into connotative semantics. So, "chartered", and "accredited", and "certified", and "registered", and so forth, all have to do with institutional acknowledgement or recognition of status, with some partially unknown range between bought status and intellectually earned status being at play. I'm personally partial to "accreditations", the agent and-or person being accredited by some institution, in recognition of something. In terms of what could be included, I would lean towards expansiveness, in general. This involves what I've been thinking of as imperfect indicators here, these degrees and so forth meaning something, even though we know that various countries engage in less and-or less consistently documented variations on these things, the data being valuable and yet also kept in proportionate check. Treating the field expansively would somewhat counter any artificial precision, regarding "status", that could arise from narrow inclusion of the most official titles(?).AvionHerbert (talk) 14:27, 14 December 2024 (PST)
Agreed! I like this a lot. A simple Accreditations element for persons will cover the territory nicely, in the Person template. Let's document the concept at the page Accreditations and I can add it to the Person template. If we wanted to describe something the person was a member of, we could put that in the Affiliations field. -- Econterms (talk) 08:20, 26 December 2024 (PST)
I see the logic behind keeping accreditations distinct from affiliations, professional memberships being in the latter, although I could have gone the other way on this. We've also been using the pre-existing affiliations field to cover to specific companies to which individual inventors are connected, which is different, so there could be a tangle, although maybe a manageable one. I've always left the names of individual inventors with whom another co-files as a separate thing, not within affiliations, and not necessarily needing a field, although I wouldn't be against one, and I've patent-referenced referenced said cases along the lines of collaboration.AvionHerbert (talk) 00:02, 6 January 2025 (PST)
I suppose membership in the Institut de France would be an accreditation? LTA (talk) 22:51, 25 January 2025 (PST)
I'd like to call it an accreditation, even though it involves "membership". Chartered and various designations go beyond specifically academic degreed accreditation. I'd also like to include the various British memberships, which skew to agents with a minority overlap of interest into inventors.AvionHerbert (talk) 10:04, 26 January 2025 (PST)
Memberships quite often apply to patent agents. Calling them affiliations would leave them out of the existing patent agent template.AvionHerbert (talk) 09:03, 7 January 2025 (PST)

The challenges of bulk treatment

  • I have bulk replace text capacity, on one computer, regardless of the browser used. Field "|Inventor location=" is partially the issue. Inserting, via edit, patent-by-patent, is not a problem, but that is not bulk. Identifying it, on the macro level, as "empty", in particular, is a challenge in that the field doesn't exist until inserted. If we had a bulk of "|Inventor location=Paris, France", for instance, bulk replacing them with "|Inventor location=Paris; Seine" would not be difficult. The challenges are the empty or non-existant fields. Back on the micro level, edit with form doesn't seem to include the location field, by the way, but I don't mind not using edit with form, and the macro level challenges are more key.AvionHerbert (talk) 10:41, 25 December 2024 (PST)
    • There is probably a way to use magic for some of our repeated phenomena such as Inventor location: Seine. Maybe we could work on this together at some point. Edit with form should work now but I urge that it not be used (for now) on fully templated patents because it eats the extra field. LTA (talk) 05:55, 26 December 2024 (PST)
  • American county data is still of bulk intrigue, in that we have that data text-digitally sourced. Is there a way to have that data bulk-inserted into the field? State postal codes, in this context, are fine, though "CA", for instance, in other contexts, brings up issues.AvionHerbert (talk) 10:41, 25 December 2024 (PST)

US location protocol continued

How are we feeling about the postal codes now? How many potential ambiguities are there beyond CA? I notice MA is Morocco, MD is Moldova, ME is Montenegro, MN Mongolia, MO Macao, MS Montserrat, and MT Malta. Some of these countries didn't exist and the others aren't heavy hitters but would this level of conflict be enough to induce us to switch to fully spelled out names? Have we decided in general these codes are acceptable after a comma in US locations but not as standalones? LTA (talk) 05:49, 26 December 2024 (PST)

Regarding the last question, I'd say yes, "San Jose, CA" is a reasonable use of such a code. But CA on its own is too ambiguous. If you'd like to switch to something more spelled out, . . . please propose it. I see the problem with using MA, MD, ME, MN, MO, and so forth as states. We could switch to three-digit codes for countries. There's a standard list: https://en.wikipedia.org/wiki/ISO_3166-1_alpha-3 -- Meyer (talk) 12:58, 26 December 2024 (PST)
Okay I've started moving to the postal code consistently, after the comma, that is, replacing text, when appropriate, watching out or odd cases of links and so forth which use the full state name after the comma, moving pages as needed. In terms of American counties, yes this applies, and I've used the lower-case "c", in "county", both the lower-case "c", and the postal code, comporting with the bulk data, as is, imported from HistPat.AvionHerbert (talk) 19:42, 26 December 2024 (PST)

Updates to Person template

I made three updates to the Person template. Let me know if you have any objection to applying them to the Person table. I have in mind to run that tonight. The updates: Added Accreditations as we discussed above; added "Family name" so as to enable easier sorting by "last name"; and moved the birth and death dates down in the template so as to draw attention to substantively more interesting things -- occupation, location, and tech areas. For slightly more detail on the changes see my recent update to Template talk:Person. -- Meyer (talk) 12:54, 26 December 2024 (PST)

I like the "Family name" phraseology, factoring in Hungarian originals and so forth. Should we set a protocol straight off regarding "von _____" and "de ___" and "du ____" and "di ____" names, treating them as included in the "Family name"? I think across all of our data we have two or three wildcard cases (very few) of a possibly "actual" family name with a semi-dubious nobility-suggesting preposition thrown in, but we can handle it.AvionHerbert (talk) 22:32, 26 December 2024 (PST)
These changes all make sense to me.
  • Regarding "von ____" I suppose this depends upon the intended purposes of the family name field. Alphabetization? In many cases I think "von ____" is the only family name, not an additional one. I could see a case for either "Zeppelin" or "von Zeppelin". "Deutsch de la Meurthe" is an exceptional case; perhaps so is "de La Vaulx".
I can't think of any faux or dubious "von ____" cases either. "von Zeppelin" is a tough call, because he and his work, and more, is known via the "Z", but he may be the only "von ____" case of that happening. For him to be an exception would be workable, with some "von ___", and "van de ___", "van den ___" and so forth otherwise clustering in "V" alphabetically. I'd go with "Deutsch de la Meurthe", in that case, and "de la Vaulx" AvionHerbert (talk) 11:33, 27 December 2024 (PST)
Punched in "von" and saw Wilhelm von Bezold. I looked at some other sites including w:de:Wilhelm von Bezold expecting them to call him von Bezold but to my surprise they just use Bezold. So does this period obituary by Arthur Berson: Berson, 1907, William v. Bezold. LTA (talk) 06:58, 28 December 2024 (PST)
I'm also surprised. In bulk terms, I'd go with the "d... ___" and "v... ___" cases, including the prepositional usages in the "family name". I don't think there will be any overall determinable "norm", even in terms of lower- versus upper-case. "de La ___" always looks awkward to me, for instance. In terms of basic inclusion, perhaps there's some sort of both-way linking manner in which we could have everything covered. Even an imperfect family name capacity will be better than all given-name-based alphabetization.AvionHerbert (talk) 07:11, 28 December 2024 (PST)
  • On a side note I see the replacement table has 2,358 entries compared to 2,354 in the old one. So I would think it's done computing and that we've added four entries.
  • Finally it would be good if we can automatically add a blank |Family name= if not also accreditations to the existing ~2350 templates that don't have them, since we know we'll be able enter information for this field in just about all of them. Maybe we can figure out how to do that tomorrow. LTA (talk) 08:37, 27 December 2024 (PST)
Automatically! Yes, that would be good. The order within the template doesn't matter much. I'm free to meet today. -- Econterms (talk) 09:42, 28 December 2024 (PST)

Let's tag 'divided' and 'renewed' patent applications

Let's tag "Divided" and/or "Renewed" patents applications so we can later review them together as a group, if we don't have a practice to do this already. Here's a quick and dirty way that doesn't require lots of work or deep reflection. I put {{Divided}} {{Renewed}} in the discussion of Patent US-1916-1210376. That'll do for now. Later we can see "What links here" from those templates to find a set of these cases. My interest is someday to track back to these cases to understand the administrative practice and infer when the innovations really occurred and how these concepts relate to them. And in practical terms: are those more or less common in certain fields, or time periods, or with certain patent agents or examiners; is there some pattern or meaning to them that is not obvious from the stated policy of the patent office? is it costly? does one of the "Divided" patents then get approved and the other rejected? etc.

I wouldn't object to having a Category or two for these terms if it would help. And I think Renewed and Divided are US-only terms, but would be happy to see counterexamples or evidence. -- Econterms (talk) 09:42, 28 December 2024 (PST)

I'm all for keeping the terms distinct. "What links here" has been a feature of which I've been thinking more generally as well. What if renewed, or divided, could just be straight linked, from each appropriate patent, and also from serial number, on which I've been fleshing out the issues?AvionHerbert (talk) 10:29, 28 December 2024 (PST)
Good. Let's keep them distinct and yes, maybe simple wikilinks will be better. Perhaps "divided patent" and "renewed patent" should just be Techtypes. They are administrative categories, not technological ones, which puts a twist on this keyword concept, but if we do it this way, we can have a page for each of definitions and examples using our existing infrastructure, and that may be enough. -- Econterms (talk) 11:04, 28 December 2024 (PST)
That would be fine.AvionHerbert (talk) 11:18, 28 December 2024 (PST)
Excellent. Please someday review Divided patent applications, a new Techtype. It really applies to applications with serial numbers, as you say, not to granted patents. I see your nice list of examples at Serial number. I didn't create the analogous Renewed patent applications Techtype yet, but if this one seems to work either of us can do that. -- Econterms (talk) 12:11, 28 December 2024 (PST)
Patent US-1911-1420609 should be showing up on the divided patents page but isn't for some reason. I don't think it's my cache. I wouldn't think it's the capital letter ... ? LTA (talk) 12:47, 28 December 2024 (PST)
I think there's a misunderstanding in our seeing "applications with serial numbers" in contrast to "granted patents". Serial numbers are not replaced by what we more generally call "patent numbers". The serial number is replaced only by another serial number, in these minority cases of division and renewal. Patent US-1907-1072663, which I just picked randomly out of the overwhelming majority, has the serial number 360379, and has been granted.AvionHerbert (talk) 13:02, 28 December 2024 (PST)
There's quite a complex here. Patent US-1912-1081794, itself, uses the filing date 7 November 1913 and serial number 799816. It refers to 674009 and filing date 1912-01-29, which we may not have, that application being "renewed". The earlier document may have been superseded and expunged. In that case, this patent keeps the titular year 1912, and any other document referring to 674009 and filing date 1912-01-29 gets pointed towards Patent US-1912-1081794. Patent US-1912-1081792(684824) doesn't use the either "renewed" or "divided", but, within its text it refers to Serial Number 674009, filed 29 January 1912, and it also refers 694825, filed 1912-03-19. That is, it refers from 684824, to the later number in the "serial", 694825, albeit 694825 and 694824 were all filed on the same date. There's also a per se back and forth among the references.14:58, 28 December 2024 (PST)

International Directory of Patent Agents

I'm seeing references to an International Directory of Patent Agents, published by (company) William Reeves, in London, in years 1893, 1897 and 1901. Can we find an online copy? or a physical copy? We can inch toward making a helpfully definitive database of patent agents. I hadn't meant to, but that seems to be a nice open niche. Expert David Pretel wrote "The number of agents registered in this directory increased from about 2,200 in 1893 to more than 4,000 in 1901". If that's the scale, it would be thinkable to transcribe the whole thing on Wikisource and make a database with them all. -- Meyer (talk) 10:24, 4 February 2025 (PST)

Wow that's a nice obscure one.

It doesn't look like this will be easy to find but it does make me wonder if there might be other comparable documents. LTA (talk) 21:16, 5 February 2025 (PST)

Retrieved from "http://econterms.net/aero/index.php?title=Project:Hot_air_flow&oldid=314624"