Latest revision as of 22:37, 27 June 2017

This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

naming conventions

Per our IRC chat discussions, can we come up with a formality as to how we want to sub-organize various thing such microprocessor and ICs in general, given that many (most?) are numbered and collisions are way too common to use wikipedia-style disambiguation.

I think we've agrees the use of sub-pages is the right direction:

• company
  • /family
    • /chip

So for example, the Intel 4004 would ideally be located at: intel/mcs-4/4004. This will resolve the already-growing ambiguity issues we're getting. Even for the Intel 4004, the support chips 4001-4003 have conflict with the famous 4000 series which include those chips, albeit missing 4004 (intentional?).

To expand on that, the naming rule could be generalized into computers, systems, and programming language (although those we've already been following this naming convention for some time).

• concept
  • dependent concept
    • specialized instances of the dependent concept.

Additionally, we want to use appropriate {{XXX title|title}} and {{XXX|topic}} to display appropriate titles and links, although in general things such as "Intel 4004", "4004", and "i4004" should all link correctly to the respective article. --ChipIt (talk) 23:29, 23 December 2015 (EST)

LGTM, I think it covers it all. --Jonathan (talk) 00:27, 24 December 2015 (EST)

Yup, all sounds about right --David (talk) 15:58, 25 December 2015 (EST)

Semantic MediaWiki

Any thought about installing Semantic MediaWiki? This will help us dramatically. Would also allow us to generate lists and compare different computers, microprocessors, and whatever else we might consider adding. Essentially all the values we're putting in those infoboxes could be compared and contrasted. Thoughts? --Jonathan (talk) 02:04, 26 December 2015 (EST)

It's been installed for some time now, just not enabled since I never got around to actually testing it. I've just enabled it, go nuts with it. --David (talk) 07:23, 26 December 2015 (EST)

UPDATE so it seems to be fully functional, I've converted crystal well and all its processors into using it so that should be a pretty good starting point! -David (talk) 22:37, 28 December 2015 (EST)

resources

any has access to this book or other similar books? http://trove.nla.gov.au/version/45481409; 1978, English, Book, Illustrated edition: Microprocessor data manual : from Electronic design / edited by Dave Bursky. --ChipIt (talk) 12:36, 3 January 2016 (EST)

"model name" redirecting

Hey guys, in an attempt to give some wiggle room for API queries, I want to setup a redirection policy. This includes redirecting common names, partial common names, model number (so long there are no ambiguities). I want to prioritize the redirection of cpuid because that's something we're particularly interested in for API purposes.

The following redirects should be done for all pages:

1. Full common name
2. Full common name, all lowercase
3. Common name, just family/model, if applicable
4. Common name, just family/model all alowercase, if applicable
5. Model number
6. Model number all lowercase
7. /proc/cpuid name

E.g. Intel Core i7-4770R:

• Common: intel/core i7/i7-4770r
• Intel Core i7-4770R
• intel core i7-4770r
• Core i7-4770R
• core i7-4770r
• i7-4770R
• i7-4770r
• Intel(R) Core(TM) i7-4770R CPU @ 3.20GHz

--David (talk) 15:44, 8 January 2016 (EST)

I went ahead and created WikiChip:model numbers based on folding@home dump table. Should give us a nice starting coverage. --ChipIt (talk) 18:02, 8 January 2016 (EST)

What about sSPEC numbers? Shouldn't they be a redirect as well? -217.248.190.1 15:01, 7 January 2017 (EST)

Yea, sSpec should redirect as well. at the moment you can actually search for by sspec parts direct here s-spec. In fact, sSpec, qSpec (we have qualification sample info too), and part numbers redirect too. We have a bot that should be doing that automatically as we add that info. Right now I think User:David is getting ready to switch over the default Skin to our new skin as well. So he's a bit busy with all that stuff. --ChipIt (talk) 16:22, 7 January 2017 (EST)

articles for years/decades

I've been thinking and I think we should start creating articles for individual years with all the notable things that happened. It would make collecting that stuff easier if we start early with it. Any thoughts? --ChipIt (talk) 04:45, 28 April 2016 (EDT)

Sounds good I'll start filling stuff in as I go along. --At32Hz (talk) 06:26, 30 April 2016 (EDT)

mass DB of CPU chip info

I collected a mass amount of CPU info to create a similar site to yours. The website never worked out, but i still have the data. the data is public domain, so feel free to use it. It is stored in comma-seperated-values files, so its is easy to import to what ever. Any you can probably set up a script to import it to you mediawiki website.

• Archived website: http://www.happytrees.org/cpu-db/

  This is the code for the website: https://github.com/zymos/cpu-db/tree/master/website

• Raw Data(in CSV spreadsheets): https://github.com/zymos/cpu-db

I also tried to make a mediawiki version, and may be able to run that again for show you. I hope this is useful. contact me if you want any help importing it, or anything else. ZyMOS (talk) 16:00, 21 May 2016 (EDT)

THis is apparently the script i used to import the CSV files to MediaWiki. Also i dont recall the details of it. https://github.com/zymos/cpu-db/blob/master/cpudb_cvs2mediawiki.pl Contact me if yoiu have questions and i can try to remember what i did ZyMOS (talk) 16:05, 21 May 2016 (EDT)

Hey ZyMOS, these CSV files are pretty nifty. I'll see if I can import them into our {{mpu}}/{{mcu}} templates when I get a bit of free time. --At32Hz (talk) 16:37, 22 May 2016 (EDT)

CPU images / Chip manufacture logos / chip package outlines

I probably have 400 cpu/mpu/SoC photos, of mostly good quality. A ton of chip package outlines of medium quality, and IC manufacture logos of low quality. If you want i can upload them, but id want to script it, and need to have permission to use a bot. Check them out below.

• http://www.happytrees.org/chips
• http://www.happytrees.org/main-images/chip-v2/
• http://how-to.wikia.com/wiki/How_to_identify_chip_packages
• http://how-to.wikia.com/wiki/Category:Chip_package_image
• http://how-to.wikia.com/wiki/How_to_identify_integrated_circuit_%28chip%29_manufacturers_by_their_logos
• http://how-to.wikia.com/wiki/Category:Logos

ZyMOS (talk) 14:58, 27 May 2016 (EDT)

Hey ZyMOS and welcome! If you want, you can go for it. I've added you to the 'Bot' group so you should be able to automate it (we have the generic {{Information}} template). We might want to make a template that has some related info: Model, Family, Datecode, etc.. Preferably with semantic properties embedded similar to how we have in the MPU template as that will allow the images to be queried/searched/dynamically shown on a pages (if you have that info associated with the images, otherwise we could add that later no biggie). --David (talk) 19:05, 27 May 2016 (EDT)

I will probably have to script it with simple info, Manufacturer and part number. It will take me a while to get the script working, its very old. ZyMOS (talk) 21:10, 27 May 2016 (EDT)

Datasheets upload

One more thing i can contribute. a bunch or CPU/MPU datasheets. I can script upload them too, but it wont be a very intelligent script, so details will have to be inputed manually. BTW im impressed at some of the databooks you found, i know how difficult old ones are hard to find. ZyMOS (talk) 10:08, 28 May 2016 (EDT)

• http://www.happytrees.org/main-files/datasheets/

Datasheets are a bitch to find! I actually have a whole treasure trove worth of data books from the 60s-80s. I scanned a bunch of databooks (some I sent to bitsavers) and a ton of datasheet years ago. Interestingly enough, a lot of the datasheets I scanned and uploaded online somehow found their way to those online datasheet websites (they even slapped their own watermark on it as if they actually own it or something!). I have a lot more manuals and stuff for some of the more rarer 4 bit to 16 bit chips out there. Lots of stuff from ebay and craigslist! My main issue is scanning them. Atm I don't have the means to scan and I can't drop $1K on one of those fancy machines so I've been waiting for something like the Czur scanner (an Indiegogo project) to come out. --ChipIt (talk) 17:29, 3 June 2016 (EDT)

You can send them to archive.org They will scan them but they have to cut the bindings and i doubt they will give them back ZyMOS (talk) 01:58, 4 June 2016 (EDT)

another resource i found: https://github.com/larsbrinkhoff/awesome-cpus

Interesting. Out of curiosity, how are they not reaching hosting limit with GitHub? Even our modest amount of docs have reached GBs. --David (talk) 02:11, 29 July 2016 (EDT)