Home > Windows > Microsoft shifts researchers to Windows unit Microsoft shifts researchers to Windows unit Eugenia Loli 2004-08-03 Windows 18 Comments Microsoft is shifting about 70 technical staff from its research unit into its Windows effort as the company gears up for Longhorn, the next major release of the operating system. About The Author Eugenia Loli Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker. Follow me on Twitter @EugeniaLoli 18 Comments 2004-08-03 8:31 pm Microsoft’s new security initiative and code auditing process seemed to slow down the company a bit these days. The deadlines they are missing have a domino affect on other products, they seemed to have bitten off more than even they can chew with Longhorn. 2004-08-03 8:39 pm “they seemed to have bitten off more than even they can chew” ms management got overly excited about all the software people might be interested in purchasing so they started to many projects with to few programmers. 2004-08-03 8:59 pm The few thousand coders they’ve got on Longhorn may be too ambitious for the kind of stuff they want in it. However, I often wonder which parts are making it the hardest. Are they fighting through building on a bad base? Are they having trouble with a specific one of the few new big features for Longhorn? Or are the programmers just being slow or over maticulous? I understand they’ve got a lot to work through, but a large part of the features they are implementing should be already done. They own their own database software (right?), so that should be a large part of WinFS done. They already do directX, that’s a good chunk of what is needed for Avalon. Just keep in mind, they have thousands of developers working on Windows. But I’d bet they’re doing backwork on security problems, fighting through writing patches between writing new features? Whilst supporting all their major stuff: Windows Office and now .Net. 2004-08-03 8:59 pm At the last place I worked at we were building a multi-media solution that was made of a player, authoring tool (that’s what I wokred on), and server. There were 3 teams of devs, plus a small group of “R&D” devs that worked on nothing but developing the latest and greatest codecs, either for video, or for advnacved 3D effects, or whatever. At one point (since we were running low on cash and other devs had already been laid off) some of these guys started to directly contribute code into the core production code. And the minute that happened we had problems (we had other problems as well, but this didn’t help ). For example, one of the R&D guys code took a filename as a chararacter array, which he had hard coded to be no more than 20-30 characters long. And sure enough this caused all sorts of hell once the code made it’s way into production and started really getting tested. It took a while to track that down. When asked why he had hard coded something like this (no joke here) his response was to just blow the other person off and ask why on earth would anyone have a filename longer than 30 characters! 2004-08-03 9:11 pm “Just keep in mind, they have thousands of developers working on Windows.” Many times adding developers to a project is counter productive. The overhead of communication and managment problems can seriously hinder and evenen risk the project. I can recommend Agile Software Development by Alistair Cockburn if you are intrested in project managment. 2004-08-03 10:11 pm i think not. core teams at ms generally run about 20 devs or less (this doesn’t include PM’s, testers, etc). granted there might be dozens of core teams working on one project, but saying there are thousands of devs is simply outlandish. also, they have to maintain backwards compatability with current builds, etc. i say take your time, get it right. plus, if they released it tomorrow, who would upgrade? chances are by the time they ship, more entities will be willing to upgrade. 2004-08-03 10:32 pm I wonder if they ‘knew’ about this drop in their own research, when they offered grants for research? Gates said “This is the place where the kind of advances that will drive the economy will be coming from.” In http://www.osnews.com/comment.php?news_id=7901 I thought he meant the USA and CS not MS and Longhorn! 2004-08-03 10:54 pm Development of Longhorn is broken up into the core areas; if you visit the blogs of the WinFS guys, you can tell that the Indigo and Avalon teams seem like a completely different alien lifeform to them. This is probably a good thing. The research developers are probably going to be likewise chopped up and allocated to Indigo, Avalon, WinFS, and all the other modular teams that work on Longhorn. I hope they put the most researchers to work on Aero Glass, though; I’m not exactly enamored with the taskbar anymore. 2004-08-04 7:46 am I wonder if they have ever read “the mythical man month”. Adding large numbers of new developers to a project often means that it takes a longer time to build, rather than speeding it up. As these new developers have to learn the code and the only people to teach them are the developers already working on the project, meaning they have less time for coding. Looks like Longhaul is going to get delayed again, it is rapidly turning into Microsofts Copland (Mac OS 8). 2004-08-04 7:51 am …nine women don’t make a baby in one month! 2004-08-04 7:53 am ..that they finish XP SP2 properly before going for the burn on Longhorn ! 2004-08-04 10:11 am I think even Microsoft realise by this stage that throwing developers onto a project won’t automatically make it quicker. I suspect that, in order for them to pull people off research, they must have decided that they are running under capacity on some (all) of the other project teams, and poaching researchers is the fastest way to bring experienced Windows developers onto the team. Given some of the output of MS Research, you could reasonably expect these new developers altogether to have some experience with low-level Windows programming and/or experience of a broad range of Windows technologies. After that it’s a matter of matching skills to the teams. 2004-08-04 10:25 am Okay, somehow I missed this from the article before: The developers, who had been studying various ways of improving programmer productivity, will now focus their efforts on improving all phases of Windows development, including design, testing and sustained engineering. Well, that explains what they’re doing anyway. I figure this should produce either shorter development times or better quality software. Both if we’re lucky. It’s a good thing anyway. 2004-08-04 1:45 pm yeh what a laugh thousands of developers working on it. is it that microsoft is scared to death of linux(hardly) i think be saying more people are working on it says that it is a product of many hard labours, wich in itself would increase the stock market of microsoft.clever management microsoft, unfortunately if only linux developers were as ruffless they would have more legs on wich to stand. 2004-08-04 4:46 pm Did ANYONE read the article at all? 2004-08-04 5:27 pm I think that Longhorn will be as special as WindowsXP is compared to Win2000, that is just a theming and some minor tweaks. — I’m talking from an end user point of view of course. 2004-08-04 6:45 pm > the company gears up for Longhorn Wot? Haven’t they been working on Longhorn (AKA ‘cowpies’ around here) for many months?? Or are they just going from 1st gear to 2nd so far… 😀 2004-08-04 9:53 pm MS had about 1000 coders in IE team when MS had goal to beat Netscape and reduce problems created by former MS ignorance (or even arrogance) about Internet at all and TCP/IP particulary. Results? Hmm, anyone has it’s opinion at it, heh.