This is the first time I have gotten to the point of wanting to submit code to (what I would consider) a major project. Specifically, I have made a small change to oggenc/audio.c. I will explain the change if necessary, but that us not the purpose of this post. Primarily, I want to know how to submit changes? Mail it to someone? Via CVS commit? If CVS, how do I get access? Also, I have had trouble logging in anonymously via CVS. When I try to log on with the instructions on the cvs.html page, I get an error: "cvs [login aborted]: recv() from server xiph.org: Connection reset by peer" Any help on these fronts is appreciated. - Jeff Smith _________________________________________________________________ Get your FREE download of MSN Explorer at http://explorer.msn.com --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/ To unsubscribe from this list, send a message to 'vorbis-dev-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
> This is the first time I have gotten to the point of wanting to submit > code to (what I would consider) a major project. Specifically, I have made > a small change to oggenc/audio.c. I will explain the change if necessary, > but that us not the purpose of this post. > Primarily, I want to know how to submit changes? Mail it to someone? > Via CVS commit? If CVS, how do I get access?Genereally only fairly active and somewhat trusted developers are allowed direct commit access. If you regularly submit patches, access it not a problem. The easiest way is to post the patch here, along with a description. One of us will apply the patch, or we'll discuss it here if there are some changes or anything like that.> Also, I have had trouble logging in anonymously via CVS. When I try to > log on with the instructions on the cvs.html page, I get an error: "cvs > [login aborted]: recv() from server xiph.org: Connection reset by peer" > Any help on these fronts is appreciated.You likely have no reverse dns. xiph.org is paranoid, so if you don't have reverse lookups, it might not work. jack. --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/ To unsubscribe from this list, send a message to 'vorbis-dev-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
I apologize for posting to both groups. I knew which to send it to, but I somehow missed that little thing. So much for first impressions. Thanks for the hints about Reverse DNS. I will check that out soon. I didn't know how you wanted the patch, so I am just inserting the diff output here. If you need something different (or for future reference), let me know. 24c24 < (((buf)[3]<<24)|((buf)[2]<<16)|((buf)[1]<<8)|((buf)[0]&0xff)); ---> (((buf)[3]<<24)|((buf)[2]<<16)|((buf)[1]<<8)|((buf)[0]&0xff))27c27 < (((buf)[1]<<8)|((buf)[0]&0xff)); ---> (((buf)[1]<<8)|((buf)[0]&0xff))30c30 < (((buf)[0]<<24)|((buf)[1]<<16)|((buf)[2]<<8)|((buf)[3]&0xff)); ---> (((buf)[0]<<24)|((buf)[1]<<16)|((buf)[2]<<8)|((buf)[3]&0xff))33c33 < (((buf)[0]<<8)|((buf)[1]&0xff)); ---> (((buf)[0]<<8)|((buf)[1]&0xff))337c337 < unsigned char buf[16]; ---> unsigned char buf[18];352,353c352 < if(len!=16) < { ---> if ((len!=16) && (len!=18)) {358,359c357 < if(fread(buf,1,16,in) < 16) < { ---> if(fread(buf,1,len,in) < len) {369a368,371> if ((len==18) && (READ_U16_LE(buf+16)!=0)) { > fprintf(stderr, "Warning: Unrecognised format chunk in WAV header\n"); > return 0; > }Reasoning: Some programs seem to treat the 'fmt ' section of a standard, uncompressed WAV slightly differently. Some specs show an extra word, cbSize, which tells how many extra bytes there are in the section. This word may be thrown out with normal uncompressed WAVs, making the size of the 'fmt ' section to be 16. But these programs don't, leaving it size 18. If the data in this extra word is anything besides 0, there is an inconsistency. Specific items: The ; on the macro definitions is redundant (take a look at 'gcc -E audio.c' if you need more convincing). If you do not use the macro at the *very* end of a statement, it causes errors. This change should be effected regardless of whether the rest is. But other parts of my changes rely on this. increasing buf size from 16 to 18. The size of the alternate header is one word longer. if ((len!=16) && (len!=18)) Check for both header sizes. if(fread(buf,1,len,in) < len) 16 and 18 are the only possible values of len, if we got this, far and we take care of both. if ((len==18) && (READ_U16_LE(buf+16)!=0)) The extra word must contain the data 0x0000. - Hope I didn't get too wordy - Jeff Smith _________________________________________________________________ Get your FREE download of MSN Explorer at http://explorer.msn.com --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/ To unsubscribe from this list, send a message to 'vorbis-dev-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
I realized the same changes needed to be made to OggDrop's audio.c file. The reasoning is the same, of course: a slightly different interpretation of the WAV format specs for vanilla, uncompressed WAVs. I still need to know how you want the diffs/patches sent, and if I need to resend the two diffs in that format. Here is the straight diff for OggDrop's audio.c: 24c24 < (((buf)[3]<<24)|((buf)[2]<<16)|((buf)[1]<<8)|((buf)[0]&0xff)); ---> (((buf)[3]<<24)|((buf)[2]<<16)|((buf)[1]<<8)|((buf)[0]&0xff))27c27 < (((buf)[1]<<8)|((buf)[0]&0xff)); ---> (((buf)[1]<<8)|((buf)[0]&0xff))71c71 < unsigned char buf[16]; ---> unsigned char buf[18];94c94 < if(len!=16) ---> if((len!=16)&&(len!=18))100c100 < if(fread(buf,1,16,in) < 16) ---> if(fread(buf,1,len,in) < len)111a112,115> if((len==18)&&(READ_U16(buf+16)!=0)) { > fprintf(stderr, "Warning: Unrecognised format chunk in WAV >header\n"); > return 0; /* Weird format chunk */ > }- Jeff Smith _________________________________________________________________ Get your FREE download of MSN Explorer at http://explorer.msn.com --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/ To unsubscribe from this list, send a message to 'vorbis-dev-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
At 06:01 PM 3/1/01 -0600, you wrote:> This is the first time I have gotten to the point of wanting to submit >code to (what I would consider) a major project. Specifically, I have made >a small change to oggenc/audio.c. I will explain the change if necessary, >but that us not the purpose of this post. > Primarily, I want to know how to submit changes? Mail it to someone? >Via CVS commit? If CVS, how do I get access? > Also, I have had trouble logging in anonymously via CVS. When I try to >log on with the instructions on the cvs.html page, I get an error: "cvs >[login aborted]: recv() from server xiph.org: Connection reset by peer" > Any help on these fronts is appreciated.Unfortunately CVS tends not to be terribly verbose in its error messages. It's possible that your host isn't able to be reverse-resolved - and the cvs server is set up (for security reasons) to reject connections in this case, I believe. Monty may be able to give a more definitive answer here. As for code - you should send a patch to this list (vorbis-dev. NOT vorbis. It is not appropriate to send mail to both lists in most cases, like you did). Someone (since this is oggenc, almost certainly me) will commit your patch if it is appropriate. Michael --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/ To unsubscribe from this list, send a message to 'vorbis-dev-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.
1) On the reverse DNS thing, I am behind a masquerading firewall at my university. Is the reverse DNS problem likely with my box, the school's firewall, or neither? 2) Monty: I think that was the entire diff file, it was just a straight diff not a diff -u. I'm still pretty green at this (diff's, patches, and CVS). 3) Michael: the standards docs I've looked at on the net are fuzzy on the issue of the header. Let me know what the project is using as an official guideline and/or where to obtain it. And I will try to figure out what programs my friend was using that created those funny headers. - Jeff Smith _________________________________________________________________ Get your FREE download of MSN Explorer at http://explorer.msn.com --- >8 ---- List archives: http://www.xiph.org/archives/ Ogg project homepage: http://www.xiph.org/ogg/ To unsubscribe from this list, send a message to 'vorbis-dev-request@xiph.org' containing only the word 'unsubscribe' in the body. No subject is needed. Unsubscribe messages sent to the list will be ignored/filtered.