There is also a rebuttal from a former Fog Creek employee.
Yeah, and I still think this article is ridiculous on this site also. The linked post amounts to misguided armchair speculation and doesn’t even capitalize either products' name correctly. I’m not going to flag it, because it’s not in any of those categories, but I’d kill for a “horribly researched” or “factually inacrruate” flag.
IMHO, I posted it in hope of further discussion, in (or because?) spite of the source’s quality. I also know @tedu was a former Fog Creek employee, maybe he’d have thoughts on it?
Why not link straight to the comment then? Atlassian devs also eloquently weighed in.
I think gecko addressed it pretty well. I don’t know how “fast” Jira moved, but FogBugz changes took a long time for other reasons. For example, Joel was very opposed to custom fields (and I tend to agree with him), but customers wanted them, real bad. Writing FogBugz in Java was not likely to change his mind any faster.
Not mentioned is that people had the ASP source to FogBugz from the beginning (hard not to ship it), and there were definitely customers making changes to it. No, not a marketplace, but that’s because there wasn’t a stable API. FogBugz 4 (all ASP, all the time I believe) could have loaded plugin ASP files out of a directory, but nobody built that feature. It was however buildable, using the technology at hand. In a language that 90% of customers were familiar with. It didn’t happen because it didn’t happen.
Everyone who has ever worked in Wasabi has stated that it didn’t slow things down, yet people keep repeating this claim. There were plenty of things not to like about Wasabi, but the Wasabi haters never mention them because they have no knowledge. Sigh.
Since the article is just pure opinion without many facts, here are some more opinions:
Every time a customer makes me use JIRA for their projects, I want to put a fork in my eye. JIRA can do so much, but I never know what I am supposed to do.
Atlassian as a “unicorn” may be a stretch. As much as I like Mercurial and Bitbucket, I think that git/Github is going to win.
Touting the Java stack, as the article does, isn’t really a big selling point for me.
The rebuttal/comments link is much better than the article.
For what it’s worth, Atlassian agrees with you about mercurial. Take a look at the logged out bitbucket homepage, you will find not one reference to their support for mercurial.
Yeah, I should have said “GitHub has already won.”
Take a look at the logged out bitbucket homepage, you will find not one reference to their support for mercurial.
Well, kinda sorta indirect:
SourceTree is Atlassian’s free desktop client for Bitbucket. Harness the full power of Git and Mercurial in a beautifully simple application.
Card-carrying Mercurial apologist here. Two big companies I know currently back Mercurial: Jane Street and Facebook. I think it’s safe to say that git is great riiiiiight up until the moment you have to crack open one of the one hundred C files and add a feature that your giant company needs.
As for Atlassian, I don’t know. I at least hope they’re not a unicorn. A unicorn is a nasty, pompous animal.
I think people can get misled by what’s popular in the open-source community and assume that anything that isn’t currently “cool” is dead — if people judged databases the same way they judge VCS’s, they’d say Oracle doesn’t exist at all, right?
To a first approximation, is that so false? Oracle is so immensely expensive that I could well imagine them only having a few thousand customers (I have no idea what the actual popularity is, but I imagine Oracle keeps that secret just like their performance numbers). Whereas a decent proportion of shared hosting accounts have MySQL instances.