I remain attracted to the Linux version numbering scheme, as described for example here... http://www.ussg.iu.edu/hypermail/linux/kernel/9804.3/0241.html It...
... Ok. Lets go w/ it. See below for further comment. ... I'm fine w/ the above except the piece where we add labels of heritrix-0_1_x and heritrix-0_3_0...
Michael Stack
stack@...
Jan 5, 2004 6:01 pm
225
... I can see dropping the 0_1_x, but the 0_3_0 serves a valuable purpose. Namely, as soon as the release version 0_2_0 is done, but before another dev version...
... Michael has pointed out to me that we can "assign the next dev number" immediately, by committing updates to project files reflecting the new version...
I just finished making our first release. Below I list what I did making the release. Its got holes in it. We can add to it as we get more practise...
Michael Stack
stack@...
Jan 5, 2004 10:07 pm
228
Congratulations & I'm sorry on Heritrix making Slashdot today. In classic Slashdot style, they got a key detail wrong (Heritrix != Wayback). But that leads...
... Thanks! Slashdot attention often comes a little earlier than ideal, but we'll take it. :) Thw WM is older code of a different pedigree (Perl & C and Alexa ...
The following changes are planned for the current Heritrix 'crawl.log': (1) Split log into crawled.log (successes) and uncrawled.log (URIs ineligible or unable...
Comments below. ... Sounds good. Why the multiple spaces between fields? Makes logs ~30% bigger than then need to be.| ... Sounds good. ... You make the...
Michael Stack
stack@...
Jan 10, 2004 12:11 am
232
Hi there, I'm interested in contributing some time to this project. What would be a good thing to do? I thought about starting by writing some junit tests, and...
Contributions would be great James. Unit tests for utility code sounds would be a good place to start. You want to sent in your contributions as patches at...
stack
stack@...
Jan 10, 2004 9:19 pm
234
Contributions would be great James. Unit tests for utility code sounds would be a good place to start. You want to sent in your contributions as patches at...
stack
stack@...
Jan 10, 2004 9:20 pm
235
We need the crawl operator to be able to specify usernames and passwords to enable access to password-protected sites, whether those credentials are requested...
... The fields up to the URI are justified with extra spaces for easier human readability: all lengths, statuses, and URIs line up for at-a-glance comparisons....
Your a good man James. I'd suggest sending the contributions to the list. Good stuff, St.Ack...
Michael Stack
stack@...
Jan 16, 2004 9:27 pm
239
I mentioned this in a private email to Gordon, but though since it wasn't getting any use that I'd announce here that I set up crawl08 to host the new test...
Dear Parker. I have tried to install this test suite on our own test installation. However, one or more files seems to be missing in the testsuite, which were ...
Hi all, here's a patch with some unit tests in it. It's not complete for the module, but should give you an idea of the style of tests I'm writing. Also I'm...
... Here is a proposal for an extension of the arc format in order to accommodate more detailed metadata. An entry for a http response currently looks like...
Dear All. On http://crawler.archive.org/dependencies.html there is a link to file:///home/stack/cc/checkout/ArchiveOpenCrawler/lib/stataclasses.jar Shouldn't...
BACKGROUND/REQUIREMENTS: We adopt the general definition of checkpointing from a Mercator paper: # Checkpointing is an important part of any long-running...
... I've been thinking of something very similar; some comments below. ... I would include these headers, both in the captured request and in any checksum...
Soren, This means that the requisite PHP PEAR DB libraries are not installed. I believe this can be fixed by using a newer version of PHP, though you can as...
Thank you for the high quality contribution. I committed your patch and added you as a Heritrix contributor. See below for response to your remarks. ... ...
Michael Stack
stack@...
Jan 20, 2004 11:09 pm
248
Thank you. Fixed. Yours, St.Ack...
Michael Stack
stack@...
Jan 20, 2004 11:14 pm
249
... No problem... Can I clarify what the current header used for code is, since it seems to be different in different files. You prefer: /* ${NAME} * * $Id:...
Hi guys, While writing tests, I've encountered a nasty bug in AbstractLongFPSet and subclasses. The test code is as follows (using a MemLongFPSet for example) ...
Hi there, here's what I've done so far for LongFPSet. It's gotten held up because of the bug I've found, but this is mostly just javadoc changes, with some...
See below. ... Yes, there are two header types. We're trying to fix that as we go. The one to use is the one you sketch above (I updated the developer doc. ...