On Sun, Mar 14, 2010 at 1:58 PM, istvan_v@xxxxxxxxxx <istvan_v@xxxxxxxxxx> wrote: > On 02/18/2010 01:11 AM, Devin Heitmueller wrote: > >> Yeah, my plan at this point was to submit a PULL request once I felt >> the driver is stable > > For those particular cards that my patch adds support for, it seems to > be stable, and I have been using it for months. Perhaps stability issues > in xc4000.c are specific to the PCTV 340e and its dib0700 I2C problems ? Different people have different standards of quality. For example, I've done essentially no analysis into the tuning performance of the current driver - validating different frequency ranges and modulation types or bandwidths. I've done no testing of tuning lock time, minimal application validation, and no effort toward making sure the power management works. Sure, I can just throw the driver upstream as-is, but I've been hesitant to merge something with questionable quality, as it reflects poorly on my reputation. Right now it's in a development tree because it's what I would consider "alpha quality", where only advanced users will install it and they know to "proceed at your own risk". And none of the above is related to the problem with the dib0700 i2c master. All that said, the situation hasn't been helped by the fact that I'm working on five different projects currently (as102, drx-d, xc4000, em28xx, and now ngene), nor the fact that I'm also the maintainer for a variety of other tuner products (and the significant support burden that creates). Bear in mind that I am aware of the frustration that when someone has patches and cannot get the maintainer to find the time to review/comment/merge. I've been in that situation myself more than once. I'll try to go through my tree and see if I can get something upstream this week which you could build on. Once that is done, you will need to break up your huge patch into a series of small incremental patches (with proper descriptions for the changes), since there is no way a single patch is going to be accepted upstream which has all of your changes. Also, you should *not* be submitting board profiles that are completely unvalidated. I saw your email on Feb 19th, where you dumped out a list of tuners that you think might *possibly* work. You should only submit board definitions for devices that either you have tested or you have gotten a user to test. It is far worse to have broken code in there (creating the illusion of a product being supported), then for there to be no support at all. When users complain about a particular board not working, you can work with them to get it supported. Devin -- Devin J. Heitmueller - Kernel Labs http://www.kernellabs.com -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
