tag:blogger.com,1999:blog-36118108.post3123972085933535005..comments2023-12-06T00:17:28.519-08:00Comments on Creative Chaos: When should a test run unattended? - IIIMatthewhttp://www.blogger.com/profile/05956714498778698672noreply@blogger.comBlogger7125tag:blogger.com,1999:blog-36118108.post-8340412898936237772009-07-27T01:03:10.317-07:002009-07-27T01:03:10.317-07:00Sesli Sohbet Sesli Chat Diyet Makyaj Sesli Sohbet...<a href="http://www.seslistanbul.com/" title="Sesli Sohbet, sesli chat, sohbet" rel="nofollow">Sesli Sohbet</a> <a href="http://www.seslistanbul.com/" title="Sesli Sohbet, sesli chat, sohbet" rel="nofollow">Sesli Chat</a> <a href="http://www.pembeli.net/" title="Makyaj, Güzellik, Egzersiz, Diyet" rel="nofollow">Diyet</a> <a href="http://www.pembeli.net/" title="Makyaj, Güzellik, Egzersiz, Diyet" rel="nofollow">Makyaj</a> <a href="http://www.sepsesli.com/" title="Sesli Sohbet, sesli chat, sohbet" rel="nofollow">Sesli Sohbet</a> <a href="http://www.sepsesli.com/" title="Sesli Sohbet, sesli chat, sohbet" rel="nofollow">Sesli Chat</a>Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-36118108.post-17329020754583198182008-11-05T07:58:00.000-08:002008-11-05T07:58:00.000-08:00merhabasesli sohbet sitemize bekleriz.merhaba<BR/><BR/><A HREF="http://www.zirvesesli.com" REL="nofollow" TITLE="sesli sohbet">sesli sohbet</A> sitemize bekleriz.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-36118108.post-75584374068145351642008-10-31T11:03:00.000-07:002008-10-31T11:03:00.000-07:00Did I say that in the podcast? I don't think I agr...Did I say that in the podcast? I don't think I agree with it. The stumbling block for automation is always maintenance of tests as code changes. Most of the reason I was wrong about unit-level TDD in 2001 was that I assumed maintenance cost and unpleasantness would lead to programmers abandoning tests. What I didn't realize was that programmers would change the way they programmed so that (a) maintenance costs would go down, and (b) the value of tests to them would go up. The combination made test maintenance palatable.<BR/><BR/>Open source tools have something to do with the cost going down - so does fast, cheap computers - but I think that wouldn't have been enough without the change in design style.<BR/><BR/>So what I suspect we need for whole-product style tests is more a change in larger-scale architectural style than better tools. But I'm recently wondering whether such style changes wouldn't help exploratory testing to the point where it would cut away at the market share of automated whole-product tests.<BR/><BR/>Tools are still important because they make it easier for a common style to infect a large community.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-36118108.post-15325519612012512372008-10-31T06:46:00.000-07:002008-10-31T06:46:00.000-07:00I'd be very interested to hear more about the slid...I'd be very interested to hear more about the slideshow tests you use. We have been talking about this exact problem here at weplay (it is actually on my to do list to post to the agile-tester list about this topic). At times we have problems with visual regressions and are seeking better coverage. While doing exploratory testing it is not uncommon to come across something that has gone out of wack in IE6 for example. Anything you would be willing to share about the tools/process you use?<BR/><BR/>Jeremy<BR/>www.weplay.comMr. Owenshttps://www.blogger.com/profile/00910839237597220839noreply@blogger.comtag:blogger.com,1999:blog-36118108.post-73679800163061314512008-10-30T10:32:00.000-07:002008-10-30T10:32:00.000-07:00I wish we kept better track of our regression metr...I wish we kept better track of our regression metrics that are caught by the unattended tests. I would guess on average, 5 per iteration. They can be very important bugs that I know for sure we would not have found any other way. Also, major refactorings can trigger dozens of failures, so refactoring support is a big value of the tests.Lisahttps://www.blogger.com/profile/10230090963033880060noreply@blogger.comtag:blogger.com,1999:blog-36118108.post-74247261383658994502008-10-30T10:27:00.000-07:002008-10-30T10:27:00.000-07:00hello Lisa. I'm not coming to conclusions yet. I'...hello Lisa. I'm not coming to conclusions yet. I'm saying the unattended strategies are most valuable for regression, and asking the question - how many bugs are regressions? For that, I'll need to do some empirical research from a bug database (and have permission for it), ideally several.Matthewhttps://www.blogger.com/profile/05956714498778698672noreply@blogger.comtag:blogger.com,1999:blog-36118108.post-39336172252921589842008-10-30T09:33:00.000-07:002008-10-30T09:33:00.000-07:00I like the question "when should a test run unatte...I like the question "when should a test run unattended", that is a good specific category of tests.<BR/><BR/>I'm a bit confused on your conclusion. Are you saying unattended tests are only useful for regression, but your regression tests aren't catching many bugs?Lisahttps://www.blogger.com/profile/10230090963033880060noreply@blogger.com