Slashdot Comment Responses #1 (On Cardkeys and Trust)

Sorry for misattribution

Alex Moskalyuk — Wed, 25 Aug 2004 05:52:00 GMT

Josh, I was the one who submitted your post to Slashdot. Sorry for the misattribution, this is embarassing, I was thinking you were Rob Mensching (http://osdir.com/Article380.phtml), the original Microsoft open-sourcer, and had you confused with him. My apologies.

Alex Moskalyuk

re: Slashdot Comment Responses #1 (On Cardkeys and Trust)

josh ledgard — Wed, 25 Aug 2004 05:58:00 GMT

Alex: No problem. It only caused me to scramble a bit and send mail to Rob and the members of his team that I'm working with. They seemed cool with it and are just happy to see any conversation around this stuff. I'll explain in some of the next few posts I make, but really we want to start small here before making some really huge announcements and screwing up how we go about it. We, as a company, need more experience in this space first.

re: Slashdot Comment Responses #1 (On Cardkeys and Trust)

Frans Bouma — Wed, 25 Aug 2004 07:16:00 GMT

Btw, "Slashdotted" is the term used for the situation where a webserver goes down because of the enormous amount of traffic from slashdot.org because it has an item on the frontpage linking to your site.

re: Slashdot Comment Responses #1 (On Cardkeys and Trust)

Ron — Wed, 25 Aug 2004 13:15:00 GMT

Josh,

I certainly hope you don't get fired for your efforts. I like what I see in many of the MSDN blogs and look forward to seeing just how this all develops. Keep up the good work.

Ron

re: Slashdot Comment Responses #1 (On Cardkeys and Trust)

Chris Garty — Wed, 25 Aug 2004 14:28:00 GMT

Keep up the great work Josh.
The .Net development community will definitely benefit from your efforts.

re: Slashdot Comment Responses #1 (On Cardkeys and Trust)

Peter — Wed, 25 Aug 2004 16:20:00 GMT

Trust is hard to get and easy to lose.

re: Slashdot Comment Responses #1 (On Cardkeys and Trust)

Norman Diamond — Thu, 26 Aug 2004 03:21:00 GMT

> We (Microsoft) have spent years damaging
> what had formerly been very good
> relationships with the developer communities
> through our lack of developer engagement,
> transparency, caring, and general air of
> arrogance.

I think that gets a silver medal.

Gold would have been for a similar comment on damaging relationships with all customers that formerly had been good or had potential for being good, by RENEGING ON WARRANTIES, lack of engagement of customers who wished to report BUGS under the asserted warranties, and lack of transparency, caring, and general air of arrogance.

re: Slashdot Comment Responses #1 (On Cardkeys and Trust)

Jim Dabell — Thu, 26 Aug 2004 12:32:00 GMT

> But if I do get fired, and you need a CEO… you can send mail to jledgard(at)gmail.com. J

Be careful, Microsoft's weblog engine screws up smileys. That looks like a 'J' to many people, not a smiley.

http://annevankesteren.nl/archives/2004/08/ieblog

re: Slashdot Comment Responses #1 (On Cardkeys and Trust)

exc@hexatron.com — Thu, 26 Aug 2004 13:27:00 GMT

Getting down to cases...

Microsoft removed code optimization from all but its most expensive C compilers.

As a developer, I don't see this as arrogance, or lack of caring or transparency.

I see it as an act of pure contempt. And I see pure contempt as the only valid response to it.

re: Slashdot Comment Responses #1 (On Cardkeys and Trust)

Mike Capp — Thu, 26 Aug 2004 16:42:00 GMT

Josh: You're getting there. It's glacially slow, and inordinately frustrating, and I don't think you'll ever be able to complete the process under current upper management, but you're getting there. Ten steps forward, nine steps back. The blogs help, especially the apparent freedom from political interference. Meaningful efforts towards C++ standards compliance have helped even more.

Fully agree with the comment you quote from "Undisclosed" regarding the deleterious effects of "hand waving marketing speak". After my own mini-rant here I finally threw realism to the winds and filed an issue on the Product Feedback system (FDBK14337 if you care and can fight your way in) and the response, while impressively polite considering my rather nasty tone, included the following gem:

"Microsoft tries very hard to respect the boundaries between commercial software, shared source, and open source, and using Bugzilla would not be appropriate for us."

If there's a faster way to lose all credibility with developers than content-free guff like this, I don't know what it is.

exc@hexatron.com : That's certainly been true in the past, but it's changing. The VC++ Express beta is free and optimizes decently (bar a few newer features like whole-program). Too early to say whether the final release will also be free, but it'll certainly be cheap, and I can guarantee it'll be free after the next Eclipse release. Annoying that it doesn't include the Platform SDK, though. (MSDN demands that I not only run IE but install a raft of excitingly risky new ActiveX controls just to download a nasty web installer. Are they *trying* to drive people away? I want a zipfile. Give me a zipfile. This shouldn't be a technical challenge.)

What angered me more was the fact that as soon as their C++ compiler finally started compiling C++ (7/7.1) they immediately stopped selling it. (Apart from the "Standard" aka crippleware version.) You had to buy the whole of Visual Studio, including stuff like Visual Basic and ASP.NET that I'd frankly rather die than own. Product tying at its very worst; that was when I went over to Mingw and the GCC toolchain and discovered that hey, this stuff actually works pretty well. But again, they've seen (or been forcibly shown) the light.

re: Slashdot Comment Responses #1 (On Cardkeys and Trust)

josh ledgard — Thu, 26 Aug 2004 17:01:00 GMT

Mike: I looked at your bug. I should say bugs, because you list several specific issues.

The largest development cost we had for this system was actually in the replication of data to our everyday internal bug database. (Eventually it will be the same thing we ship with team server.) We would have incurred this cost regardless the use of the bugzilla front end/database because we would not have switched to using bugzilla internally.

RE: Browsing and logging into passport... these are known issues. The read part of the site shouldn't force you to log in. We do want to require unique ID's for voting and interacting though, so that will always require a log in.

General use of passport. It sucks, people here know that passport, is broken in a lot of ways, but it's a sign in technology that we can use and will continue to use of several of our sites. The biggest reason for this IMO... as a user I don't want to have to have different logins for each Microsoft application I connect to. (and this is just me speaking now..) It's frustrating to have to remember a different username for channel9, GDN, Asp.Net forums, and any other MS property. The passport team has actually started work on new versions that will be much easier for people to use. Regardless though, we still don't have it implemented 100% correctly.

re: Slashdot Comment Responses #1 (On Cardkeys and Trust)

Mike Capp — Thu, 26 Aug 2004 17:48:00 GMT

Josh - thanks for the reply and for taking the time. This is *exactly* the sort of response that builds trust - you get vastly more respect for acknowledging imperfection than for pretending you're perfect when everyone can plainly see you're not. Now, if you can just clone yourself a few hundred times...

Glad to hear that the Passport issues are getting attention. Another aspect of this problem that didn't occur to me before is that the read-barrier shuts out the Google bot, and Google is still by far the best way to find information on MSDN.

Re the fixed replication cost - fine, although that doesn't actually answer the question. A fixed cost is neutral; it's no argument either for or against reinventing the wheel versus using an off-the-shelf frontend.

re: Slashdot Comment Responses #1 (On Cardkeys and Trust)

josh ledgard — Thu, 26 Aug 2004 18:36:00 GMT

Mike, one note. I feel that I've noticed the forced log in for reading before, but I honestly couldn't get it to force me to log in today when I tried several searches and bug browsing. If this still happens for you the team would be very interested in knowing some more details... browser... what exactly you were doing etc.

Yes, I did dodge the issue a bit about the front end. Honestly I'm sure the idea may have come up (I wasn't involved early in the project) and I can imagine that someone felt it was too risky to work with that code. It's one of the attitudes I'd like to change around here. It doesn't just have to do with OSS, but in general there is a bit of arrogance I've noticed here about building things ourselves. Even for internal tools teams often seem to choose building their own internal document storage site (for example) despite the fact we ship one called sharepoint. And if that one wasn't good enough why wouldn't we just buy a better one for our teams use??? I don't have a good answer other than knowing I'd like to change it for no other reason than making us more efficient so I (selfishly) can see a better piece of the profit.

re: Slashdot Comment Responses #1 (On Cardkeys and Trust)

Mike Capp — Thu, 26 Aug 2004 19:46:00 GMT

Dammit! It's most definitely happened several times today, because I've been cussing it, but since reading your latest post I can't reproduce it for the life of me. And it recognises me now, even after deleting every Passport cookie in sight! Do you have mystical powers, or did someone's cat just brush past the "workProperly" switch?

For what it's worth, I use Moz 1.6 at home and 1.7 at work. Not doing anything special; it's happened on every single visit until now. I'll shout it happens again. (I should warn you that if you keep this up your blog is liable to become the new Product Feedback frontend...)

> in general there is a bit of
> arrogance I've noticed here
> about building things ourselves

Heh. That's not a Microsoft thing, that's a software engineer thing. Just hit us with something heavy, it usually does the trick.

re: Slashdot Comment Responses #1 (On Cardkeys and Trust)

Mike Capp — Thu, 26 Aug 2004 21:41:00 GMT

Happens again now - came back after a couple of hours, new dial-up connection, navigate to http://lab.msdn.microsoft.com/productfeedback/default.aspx and I get the login page without the username (email) being prefilled as usual. That is, it's now reflecting the fact that I deleted my cookies.

Can't do it repeatably, though. (That is, deleting Passport cookies, quitting the browser and reconnecting dialup with a different IP) does not appear to be sufficient. Looks as if the two-hour wait is also significant.

re: Slashdot Comment Responses #1 (On Cardkeys and Trust)

josh ledgard — Thu, 26 Aug 2004 23:57:00 GMT

I'll pass the dialup, cookies and browser information along to the site owners. I still haven't been able to get this to happen from multiple machines.

Microsoft and open source

Digging .NET — Tue, 31 Aug 2004 01:23:00 GMT

Microsoft and open source

More on Open Source at Microsoft via Josh Ledgard's blog.

when setup isn't just xcopy — Mon, 27 Sep 2004 08:55:00 GMT

Imagine a blog entry where I discuss Open Source at Microsoft via posts on Josh Ledgard's blog.

Scooblog on Corporate Transparency - Part 1: What Comprises Transparency?

scooblog by josh ledgard — Wed, 29 Dec 2004 07:45:00 GMT