http://www.amazon.com/exec/obidos/tg/stores/detail/-/books/1932111751/reviews/104-5572603-3056739#19321117514000 Anyone read it? Is it good? Thanks. Yogi Wahyu Prasidha. _______________________________________________ Sweng-gamedev mailing list Sweng-gamedev <at> lists.midnightryder.com http://lists.midnightryder.com/listinfo.cgi/sweng-gamedev-midnightryder.com
Jamie Fristrom offers his opinions on it http://gamedevleague.blogspot.com/ > -----Original Message----- > From: y o g i w p [mailto:yogiwp2 <at> gmx.net] > Sent: Thursday, August 07, 2003 7:50 PM > To: sweng-gamedev <at> midnightryder.com > Subject: [Sweng-gamedev] Book: Game Coding Complete > > > http://www.amazon.com/exec/obidos/tg/stores/detail/-/books/193 > 2111751/reviews/104-5572603-3056739#19321117514000 > Anyone read it? Is it good? > > Thanks. > > Yogi Wahyu Prasidha. _______________________________________________ > Sweng-gamedev mailing list Sweng-gamedev <at> lists.midnightryder.com > http://lists.midnightryder.com/listinfo.cgi/sweng-gamedev-midn ightryder.com _______________________________________________ Sweng-gamedev mailing list Sweng-gamedev <at> lists.midnightryder.com http://lists.midnightryder.com/listinfo.cgi/sweng-gamedev-midnightryder.com
Ok; not having dependencies in our schedule screwed us. Missed milestones, yada yada yada. So I give MS Project a shot. Here's what I discover: If I use Standard levelling, and I have tasks that look like this: Task X Priority 1 Task Y Priority 2 Predecessors 1 Task Z Prioirty 3 Then it does task X first, then task Y, then task Z, instead of Task Z (the highest priority), Task X, Task Y. If I use Priority, Standard, that situation works, but then this situation doesn't work: Task X Priority 1 Task Y Priority 3 Predecessors 1 Task Z Priority 2 This should do Task X, then Task Y, then Task Z, but it goes: Z, X, Y. If I set the priority for task X to 3, accepting that Project doesn't understand that being a predecessor to a high-priority task implies that you're just as high a priority, then it works, but I'm trying to migrate a project from Excel that has about two-thousand tasks in it, and I would need some way to automate the process. Is there some way to automate the process so that priorities are 'correct'? Or is MS Project fundamentally broken? The Outlook / SourceSafe of the Project Management Software world?(Continue reading)
> From: Jamie Fristrom > > Ok; not having dependencies in our schedule screwed us. > Missed milestones, yada yada yada. So I give MS Project a > shot. Here's what I discover: > > If I use Standard levelling, and I have tasks that look > like this: > Task X Priority 1 > Task Y Priority 2 Predecessors 1 > Task Z Prioirty 3 > Then it does task X first, then task Y, then task Z, > instead of Task Z (the highest priority), Task X, Task Y. > [snip] I was under the impression that when dealing with priorities, the greater the magnitude the *lower* the priority. Thus anything of priority 1 must be done first, then 2, then 2+n (where n > 0). It's an arbitrary thing, but one that must be taken into account. So if you said... X, 3 Y, 2 (Pred: X) Z, 1 ...it should sort as Z -> X -> Y. -- Brian(Continue reading)
Project does have its problems. That particular annoyance isn't top of my list (if you've ever tried merging schedules, you'll realise that Project has an almost bottomless pit of annoyances to choose from). I've bitched and moaned about various people in the Project team about my annoyances, but I don't really know if anything happened about them (well, I got some responses which basically indicated that I didn't have the right 'paradigm' for using the product, or something). I figured out how to route around them, and never really dared look to see if they were fixed (since they enraged me so much the first time). My general observation with Project is that it's pretty much a "write once" kind of system. I'm sure it's not *intended* to work like that, but that's how it ends up. What it's good for is building a schedule from scratch. You can shuffle tasks and assignments and dependencies, and plug in stuff like holidays and see where the dates come out. Generally speaking, it all more or less makes sense (provided you don't rely on the priority stuff to be too useful). At that point, I cut and paste it all into Excel, which is much better for tracking stuff in, you can turn on sharing in the document, and everyone understands it. I generally only use Project to plan out a milestone at a time in detail (planning out a single day of work 6 months in advance would be pure bullshit anyway). Which means it's okay to just rebuild the schedule each milestone, because there's only a moderate amount of stuff to rebuild. The schedule always needs readjusting after each milestone(Continue reading)
>I was under the impression that when dealing with priorities, the greater the magnitude the *lower* >the priority. Thus anything of priority 1 must be done first, then 2, then 2+n (where n > 0). It's >an arbitrary thing, but one that must be taken into account. No, I think Project has the greater-numerical-value-is-higher-priority convention. With the added special value of priority 1000 meaning "don't level me". I never really got priority to work very well. I typically just set up dependencies and then drag around tasks until it looks right. Tony Cox - Development Lead, Hockey Microsoft Games Studios - Sports _______________________________________________ Sweng-gamedev mailing list Sweng-gamedev <at> lists.midnightryder.com http://lists.midnightryder.com/listinfo.cgi/sweng-gamedev-midnightryder.com
I just talked to a Microsoft tech representative and she said basically the same thing. "It doesn't do that. Why don't you hook up all your tasks into one giant dependency chain instead of using priorities?" I'm paraphrasing. At least she refunded the $35 for the call. ---- Jamie Fristrom - http://gamedevleague.blogspot.com Treyarch - http://www.treyarch.com -----Original Message----- From: sweng-gamedev-midnightryder.com-admin <at> lists.midnightryder.com [mailto:sweng-gamedev-midnightryder.com-admin <at> lists.midnightryder.com] On Behalf Of Tony Cox Sent: Thursday, August 21, 2003 5:12 PM To: sweng-gamedev <at> midnightryder.com Subject: RE: [Sweng-gamedev] Is MS Project fundamentally broken? >I was under the impression that when dealing with priorities, the greater the magnitude the *lower* >the priority. Thus anything of priority 1 must be done first, then 2, then 2+n (where n > 0). It's >an arbitrary thing, but one that must be taken into account. No, I think Project has the greater-numerical-value-is-higher-priority convention. With the added special value of priority 1000 meaning "don't level me". I never really got priority to work very well. I typically just set up dependencies and then drag around tasks until it looks right.(Continue reading)
This isn't our situation, but a lot of publishers expect your list of deliverables for each milestone at the beginning of production; which would mean that you'd have to at least have a rough plan (wouldn't necessarily have to be day-granularity) for the whole thing at that point, and it would be nice if Project could handle something like that. I guess the best thing is to make sure you have no dependencies. Somebody can't work? Turnaround the placeholder asset for them ASAP. Finish it later. Everybody ends up multitasking. You miss milestones, but everything comes together at the end. Blearg. ---- I generally only use Project to plan out a milestone at a time in detail (planning out a single day of work 6 months in advance would be pure bullshit anyway). Which means it's okay to just rebuild the schedule each milestone, because there's only a moderate amount of stuff to rebuild. The schedule always needs readjusting after each milestone anyway. _______________________________________________ Sweng-gamedev mailing list Sweng-gamedev <at> lists.midnightryder.com http://lists.midnightryder.com/listinfo.cgi/sweng-gamedev-midnightryder.com
I guess in a certain philsophical way, priorities are not required. I mean, why do you have priorities? Why is task A more important than task B? Isn't it often because there really is a depedency chain somewhere, albeit a complex one? Many of the reasons you come up with for priorities are really just dependencies restated. "It's important that we do this first because..." often ends with "...we need it to do X or show to Y, or have enough time to do Z". To be clear: I do think it's broken that Project has such a tough time with dependencies, but I'm just pointing out that you can get quite far if you really analyze *why* you wanted the priorities and discover that some of those needs really are about dependencies. Like I say, the most effective method I've found is to use Project in a 'clean slate' fashion for building an initial schedule. For tracking I use Excel. If you need to make changes as you go along, it's better to start with a clean slate and selectively paste-in and slide around by hand, then try to do complex edits in the middle of a huge Project file (basically because the levelling stuff never does quite what you'd expect). Tony Cox - Development Lead, Hockey Microsoft Games Studios - Sports -----Original Message----- From: sweng-gamedev-midnightryder.com-admin <at> lists.midnightryder.com [mailto:sweng-gamedev-midnightryder.com-admin <at> lists.midnightryder.com] On Behalf Of Jamie Fristrom(Continue reading)
Most of our priorities are based on risk or demos. Stencil shadows - could be a big framerate hit - high priority. Larry Goldberg wants to see this boss in the next greenlight demo - high priority. So for me it's a subtle but important distinction between priority and dependency. This is funny: after talking to the tech support woman, she said she'd send me an e-mail with a link to Microsoft's feature request web page. She sent me a blank e-mail. It's like, "Get a hint, Jamie." ---- Jamie Fristrom - http://gamedevleague.blogspot.com Treyarch - http://www.treyarch.com -----Original Message----- From: sweng-gamedev-midnightryder.com-admin <at> lists.midnightryder.com [mailto:sweng-gamedev-midnightryder.com-admin <at> lists.midnightryder.com] On Behalf Of Tony Cox Sent: Thursday, August 21, 2003 5:49 PM To: sweng-gamedev <at> midnightryder.com Subject: RE: [Sweng-gamedev] Is MS Project fundamentally broken? I guess in a certain philsophical way, priorities are not required. I mean, why do you have priorities? Why is task A more important than task B? Isn't it often because there really is a depedency chain somewhere, albeit a complex one? Many of the reasons you come up with for priorities are really just dependencies restated. "It's important that we do this first because..." often ends with "...we need it to do X or show to Y, or have enough time to do Z".(Continue reading)
RSS Feed46 | |
|---|---|
1 | |
8 | |
6 | |
29 | |
11 | |
19 | |
22 | |
3 | |
4 | |
7 | |
34 | |
11 | |
6 | |
32 | |
249 | |
122 | |
50 | |
21 | |
81 | |
380 | |
149 | |
3 | |
4 | |
131 | |
38 | |
65 | |
131 | |
10 | |
1 | |
25 | |
5 | |
6 | |
7 | |
92 | |
309 | |
200 | |
28 | |
207 | |
264 | |
3 | |
177 | |
358 | |
45 | |
120 | |
247 | |
195 | |
233 | |
7 | |
10 | |
76 | |
67 | |
30 | |
8 | |
31 | |
1 | |
108 | |
124 | |
28 | |
293 | |
390 | |
210 | |
359 | |
106 | |
55 | |
11 | |
123 | |
76 | |
49 | |
32 | |
122 | |
96 | |
83 | |
276 | |
144 | |
57 | |
177 | |
98 | |
80 | |
86 | |
175 | |
141 | |
75 | |
43 | |
18 | |
72 | |
123 | |
104 | |
79 | |
155 | |
194 | |
91 | |
69 | |
155 | |
5 | |
49 | |
74 | |
166 | |
104 | |
56 | |
256 | |
65 | |
127 | |
293 | |
40 |
