[Live-devel] How to contribute code?

[BONNEAU Guy]

Ross,

|
|When I say that I don't support old versions of the code, that also
|means that I don't provide any *links* to old versions of the code.
|So, while anyone is, of course, free to keep around any old versions
|of the code that they wish, I won't be providing or publicizing (on
|the "live555.com" web site or elsewhere) any links to these.
|
|The reason is quite simple: When a new version of the code is
|released, it's usually to fix bugs.  I can't, with a clear
|conscience, endorse a copy of the code that I know is buggy.

I agree with you and understand your point about supporting old release.
However this being said my point of view is a source repository is not
about supporting old release. It is about knowledge and information. And
for this purpose it can only benefit the whole community using the
live555 library. A few years ago I had to implement the full SMPTE
2022-1 specification to the live555 library. For this purpose I had to
dig deeper in the library. To gain knowledge about the how and why of
some part of the code I really missed an history insight from a source
control repository. 

For professional development when you jump into a code with no a priori
knowledge an history repository is a great professional tool. It can
spare you of many hours if not days of struggling into understanding the
code.

Remember you wrote the live555 history. It is in your brain. For this
reason it is quite understandable that you don't need an history
repository. You own it. You write the book. But I didn't and this is
also shared by many others people.

I hope you will reconsider your decision about providing an external
link to an history repository for the benefit of the community and
everyone working with the live555 library. 

|
|However, there was recently a rare exception to this - a new version
|that did not fix bugs, but instead was a major reimplementation of
|existing functionality (the "RTSPClient" class).  Although this was
|intended to improve the code, it also introduced some new bugs (most
|of which now appear to have been fixed, although there remain some
|unverified reports of problems with RTP-over-TCP).
|
|I gave serious thought to making an exception for that particular
|version, and releasing it as a separate, 'experimental' release,
|separate from the usual 'latest' release ("live555-latest.tar.gz").
|In hindsight, perhaps I should have done that.  It would have allowed
|people to avoid encountering bugs in the new "RTSPClient"
|implementation, but it also would have meant that these bugs would
|not have been found as quickly (because few people would have ended
|up trying the 'experimental' release).  (Note that the new
|"RTSPClient" implementation has so far also exposed a bug in at least
|one other server (camera) implementation, as well as a bug in our own
|"openRTSP" application code - neither of which would not have been
|found nearly as quickly (if at all) had we not made people endure the
|temporary pain of dealing with this new implementation.)
|--
|
|Ross Finlayson
|Live Networks, Inc.
|http://www.live555.com/
|_______________________________________________
|live-devel mailing list
|live-devel at lists.live555.com
|http://lists.live555.com/mailman/listinfo/live-devel