At 03:03 PM 1/24/2006, Brad Velander wrote: >Abd ul-Rahman, > You suggest a fine idea but over the years I have come to > one conclusion about common user libraries. There are just too many > unique details that people want or have need of in their library > parts. This includes but is by no means limited to: Line thickness, > pad sizes, in some cases pin numbering (SOT-23 parts?), company > part numbers/details, layer usage, parameters and fields, just > plain personal preferences and then we also fall prey to > fabrication issues with certain vendors.
First of all, most of these issues can be addressed in a proper shared public library. That libraries can be shared is obvious: many are. Many companies have internal libraries shared and used by many designers. Companies, of course, can have design standards, which is one reason why it's possible for them to have common libraries. It is possible to have a public library without having a common design standard, but it will be far simpler and the parts more generally useful if there *are* common standards. So the first task of a library association would be the adoption or formulation of standards. I'd assume that parts would be submitted to the library by a designer and would be tagged by the contributor as conforming to standards, not conforming, or unknown. The submission process should be simple and quick and easy. Other tags would include a confirmation that the symbol/footprint was actually used in a project and served without modification. Additional details might be included as options (wave solder vs. reflow, etc.) That's how parts would initially get into the library. But those parts would start out as unverified. Another user, needing a part, might pick up the unverified part and check it over. If an obvious error is found, that would be submitted as a correction. If not, if the part is actually used, again, that designer could submit a tested validation. Each part would have a history showing validations and corrections, like a wiki page on the part. I thought this out to a certain point before and some of it is coming back to me. When a whole design is tested and found good, the set of symbols and footprints would be extracted by a designer who cares to do so and submitted as validated. So parts would accumulate multiple validations. A part that had been stable with no correction and many validations, I would say, I could use without further ado.... as long as I found the original design standards adequate. I learned a long time ago that trying to make everything perfect, i.e., what *I* think is perfect, was a huge waste of time. There is a point where good enough is good enough, and making it "better" is worse, since it impacts cost, time to market, and general aggravation. A shared library would not be perfect, for sure. But the argument I'm seeing here against shared libraries is essentially the same argument as might be made for hand-crafted nuts and bolts, works of art indeed, compared to standardized parts. The pinout problem would be totally eliminated if the shared library is keyed and parts named by manufacturer part number. Given that, it doesn't matter what the pin numbers are, though standards would be nice to avoid confusing the poor technician. (And, indeed, more attention in the standards could be put to the needs of techs, such as helpful silkscreen marking and consistent solder-side pin identifiers). > I definitely don't knock your idea or enthusiasm for such a > project but for the aforementioned reasons I would abstain. The beauty of Free Associations is that everyone is free to abstain. I'm not proposing that the User Association collect dues from every member and then spend it on making libraries. But it would not be at all a bad idea for those users who cared to do so to contribute to a fund that could hire a librarian. The fact is that if enough of us did it, we'd save money. And it would be up to us if we wanted to share the parts with the rest of the world. > This is why I also never use any canned libraries (well very > rarely use them as a starting template), by the time I modify them > and check their accuracy I usually can design them from scratch in > an equal or lesser amount of time. Personally checking every part is a waste of time. It's necessary in the current environment, because you have no confirmation from a large body of designers that the part is correct. Think about what I've proposed. If it cost you *very* little in time to contribute parts, very little to validate them (you'd simply run a process on a completed design that you were satisfied with), you might consider making the contributions. And if enough designers did that, even a few dozen might be enough, we'd have a common library that would be good enough to use without checking. Yes, every part would have, necessarily, an identification of the designers who created and validated them, not possible to massively forge (you'd have to crack many passwords). A part suddenly appearing validated by a lot of unknown designers would be very, very suspicious. Not likely to happen, though. Perhaps the idea wouldn't work. But I'm not suggesting a massive investment. Indeed, just a little interest and activity could make it happen. And when it started to happen at all, it could snowball. I've thought at times of asking for Altium support for this project. Perhaps they'd like to support it, I don't know. But the main thing needed is not money, just a little effort distributed over many people. ____________________________________________________________ You are subscribed to the PEDA discussion forum To Post messages: mailto:[email protected] Unsubscribe and Other Options: http://techservinc.com/mailman/listinfo/peda_techservinc.com Browse or Search Old Archives (2001-2004): http://www.mail-archive.com/[email protected] Browse or Search Current Archives (2004-Current): http://www.mail-archive.com/[email protected]
