(Schoudel had to go quickly, so we did his status first)<br>
[15:06] &lt;Sid&gt; Then Schou what's your status? =)<br>
[15:07] &lt;brians&gt; well, the importer is done<br>
[15:07] &lt;brians&gt; a little buggy i guess<br>
[15:07] &lt;Ross&gt; :)<br>
[15:07] &lt;brians&gt; doesn't seem to handle on cascade in sql<br>
[15:07] &lt;brians&gt; i<br>
[15:07] &lt;brians&gt; i'll need to see if there is an apache version <br>
[15:07] &lt;Sid&gt; Of course, we don't need to do anything with cascades, but it shouldn't die like it does now.<br>
[15:07] &lt;brians&gt; that can handle that<br>
[15:08] &lt;brians&gt; i agree<br>
[15:08] &lt;Sid&gt; Or you can hack the code =)<br>
[15:08] &lt;brians&gt; i'll need to investigate a little further<br>
[15:08] * jrarick has joined #dbviz<br>
[15:08] &lt;Sid&gt; Search on their forums to see if they know of the bug<br>
[15:08] &lt;brians&gt; i do find it strange though, good catch sid<br>
[15:09] &lt;Sid&gt; Hehe, I just wanted to see how well the whole program would handle a big schema since we've never gotten one in there before<br>
[15:09] &lt;brians&gt; where do you want to focus next, or stay with this?<br>
[15:09] &lt;brians&gt; me<br>
[15:09] &lt;Sid&gt; For the week, continue trying to fix the bugs in it.<br>
[15:09] &lt;brians&gt; it should handle a big schema, just no cascades<br>
[15:09] &lt;Sid&gt; We'll see Thursday whether it needs more time or not<br>
[15:10] &lt;Sid&gt; Do we have any big test SQL files without cascades?<br>
[15:10] &lt;brians&gt; i'll come up with one<br>
[15:10] &lt;Sid&gt; Ok.<br>
[15:10] &lt;SandraFaust&gt; you can always delete the cascades<br>
[15:10] &lt;Sid&gt; That might be easiest. =)<br>
[15:10] &lt;brians&gt; yeah, but it should work though or at least not hang<br>
[15:10] &lt;brians&gt; it is part of SQL92<br>
[15:11] &lt;Ross&gt; throw the importException if things fail<br>
[15:11] &lt;Sid&gt; So just keep working on the importer for now, Schoudel, and try to make it bug free.<br>
[15:11] &lt;brians&gt; it doesn't cause an error, just hangs<br>
[15:12] &lt;Ross&gt; right, find what would cause it to hang and rather than hanging throw the exception<br>
[15:12] &lt;brians&gt; ok, i'll investigate and come up with a big file for the week<br>
[15:12] &lt;Ross&gt; that'll get caught and report to the user that file couldn't be used<br>
[15:12] &lt;brians&gt; let me know now if you have anything else<br>
[15:12] &lt;Sid&gt; Well, preferably it should just ignore the cascade statements<br>
[15:12] &lt;Sid&gt; Since they're valid SQL.<br>
[15:12] &lt;Sid&gt; Will do Schoudel<br>
[15:12] &lt;brians&gt; i agree<br>
[15:12] &lt;SandraFaust&gt; how big file are you looking for?<br>
[15:13] &lt;Sid&gt; &gt;50 tables.<br>
[15:13] &lt;brians&gt; i'll dig up something<br>
[15:13] &lt;Sid&gt; Preferably have multiple test files.  One with ~50, ~100, ~200<br>
[15:13] &lt;SandraFaust&gt; I don't have anything such at hand<br>
[15:13] &lt;Sid&gt; I have one I can test with at work, though I cant distribute it out to ya'll.<br>
[15:13] &lt;Ross&gt; come on it's only proprietary<br>
[15:14] &lt;Sid&gt; =)<br>
[15:14] &lt;Sid&gt; Ok, let's get started...<br>
[15:14] &lt;Sid&gt; Meeting 17 Agenda<br>
[15:14] &lt;Sid&gt; 1 Intro<br>
[15:14] &lt;Sid&gt;   - Send status before meetings<br>
[15:14] &lt;Sid&gt; 2 Test schedule update<br>
[15:14] &lt;Sid&gt; 3 Iteration progress<br>
[15:14] &lt;Sid&gt; 4 Status/Tasking<br>
[15:14] &lt;Sid&gt;   - Schoudel, Ross, Sandra, Sid, Larry, Uday, Jim, Sobby<br>
[15:14] &lt;brians&gt; i<br>
[15:14] &lt;Sid&gt; I sent out an email an hour ago with Ross's crazy-good idea<br>
[15:14] &lt;Sid&gt; Go if you need to Schoudel. =)<br>
[15:14] &lt;Ross&gt; just trying to save time<br>
[15:15] &lt;brians&gt; i'll sign off now, and review minutes<br>
[15:15] * brians has quit IRC (Quit: )<br>
[15:15] &lt;Sid&gt; Yea, since most of the time is taken up hearing what everybody's status is<br>
[15:16] &lt;Sid&gt; and if we can trim that down so everyone knows beforehand, then we can shorten the meetings<br>
[15:16] &lt;Sid&gt; and actually discuss things, if anyone feels inclined.<br>
[15:16] &lt;Sid&gt; So everyone, please send out a status message to the group at least an hour before each meeting.<br>
[15:16] &lt;Ross&gt; then review it before the meeting<br>
[15:17] &lt;Ross&gt; review everyone elses I mean<br>
[15:17] &lt;Sid&gt; Well, hopefully people read what's in their email. If you don't it's your fault if you don't know what we're talkinga bout. =)<br>
[15:18] &lt;Sid&gt; BTW, I still hven't heard a peep from Danny since that first meeting we had.<br>
[15:18] &lt;Sid&gt; Seems like one of those big-spurt-of-action talks that get dropped on the floor.<br>
[15:18] &lt;Sid&gt; Course, mebbe he's busy grading exams. =)<br>
[15:18] &lt;Sid&gt; Anyway, #2.  Test Schedule update.<br>
[15:19] &lt;Sid&gt; I've decided that a traditional testing phase won't really work so well with our project.<br>
[15:19] &lt;Sid&gt; 1) We've been lax with Requirements documentation, so we can't really say what exactly is correct behavior.<br>
[15:20] &lt;Sid&gt; 2) The program's so small that having established test cases wouldn't occupy much of anyone's time.<br>
[15:21] &lt;Sid&gt; We still need to do testing, but it will have to be a little informal.<br>
[15:21] &lt;jrarick&gt; Sid, what about code coverage testing<br>
[15:21] &lt;jrarick&gt; Could that be the testing phase<br>
[15:21] &lt;Sid&gt; That's more metrics, in my mind.<br>
[15:21] &lt;Sid&gt; And it's not really a phase. It's a task.<br>
[15:22] &lt;Sid&gt; You run a code coverage test program and it gives you a report.<br>
[15:22] &lt;jrarick&gt; That could be the result of the testing phase<br>
[15:23] &lt;jrarick&gt; Though not many functions have testing routines<br>
[15:23] &lt;Sid&gt; We will need to run metrics as part of tasking, since they said they want to know how we ued metrics.<br>
[15:23] &lt;jrarick&gt; I guess I don't view coverage as testing<br>
[15:23] &lt;jrarick&gt; Metrics is how well the code is organized<br>
[15:23] &lt;jrarick&gt; Coverage is does the stuff work<br>
[15:24] &lt;jrarick&gt; We could have good coverage with poorly organized code<br>
[15:24] &lt;jrarick&gt; and vice versa<br>
[15:24] &lt;Sid&gt; And we could have poor coverage with working code.<br>
[15:24] &lt;Sid&gt; And testing == determining whether the code works.<br>
[15:24] &lt;Sid&gt; Well, regardless of what we call that effort, we'll do code coverage checking.<br>
[15:25] &lt;Sid&gt; #3. Iteration Progress.  Just a reminder, the Construction2 Iteration is scheduled to end next Thursday<br>
[15:26] &lt;Ross&gt; are we having another construciton iteration?<br>
[15:26] &lt;Sid&gt; Yes<br>
[15:26] &lt;Ross&gt; cool<br>
[15:26] &lt;Sid&gt; One more, then a short Transition iteration<br>
[15:26] &lt;Sid&gt; Which is where we'll try to get that Installation Use case working.<br>
[15:26] &lt;Sid&gt; And enhance it<br>
[15:27] &lt;Sid&gt; So Im not so organized jumping topics a bit. =)<br>
[15:27] &lt;Sid&gt; I mentioned the testing stuff and end-of-iteration thing<br>
[15:27] &lt;Sid&gt; because I want everyone to play with dbViZ<br>
[15:27] &lt;Sid&gt; submit bug reports<br>
[15:27] &lt;Sid&gt; And fix bugs if you get em =)<br>
[15:27] &lt;udaykale&gt; we must have some nightly versioning scheme<br>
[15:28] &lt;udaykale&gt; that we can report bugs against<br>
[15:28] &lt;Sid&gt; If you need, you can retrieve a copy of the app at any date using CVS<br>
[15:28] &lt;udaykale&gt; just to better identify when a bug was found and when it was fixed<br>
[15:28] &lt;Sid&gt; The sourceforge bug stuff is all timestamped.<br>
[15:29] &lt;udaykale&gt; thats true, but i was thinking of having a nightly build numbering for our dbViz code<br>
[15:29] &lt;Sid&gt; Hmm, I'm not sure what the advantage would be.<br>
[15:29] &lt;udaykale&gt; so we can say this bug was reported against build no 8 and fixed in build no 10<br>
[15:30] &lt;udaykale&gt; if thats not too difficult<br>
[15:30] &lt;Sid&gt; Why is that better than saying "this bug was reported on Mar 20 and fixed on Mar 23"?<br>
[15:30] &lt;Ross&gt; I don't know if there is enough activity in the project to warrent nightly builds<br>
[15:31] &lt;Sid&gt; I'm fighting it because it doesn't seem like there's much advantage.<br>
[15:31] &lt;Sid&gt; Plus, you'd introduce the ambiguity of "when was Build8 made?"<br>
[15:32] &lt;udaykale&gt; i agree, 'twas just a suggestion<br>
[15:32] &lt;Ross&gt; currently I've been stamping the code on a weekly basis<br>
[15:32] &lt;Ross&gt; effectively giving us time stamped builds<br>
[15:33] &lt;Sid&gt; Yes, there is that, though we don't really use the tags. =)<br>
[15:33] &lt;Ross&gt; true enough<br>
[15:33] &lt;udaykale&gt; thats good enough, but we should be using those weekly tags<br>
[15:33] &lt;Ross&gt; you can in sourceforge<br>
[15:33] &lt;Sid&gt; If I add the tags into SF<br>
[15:34] &lt;Ross&gt; or we could do it informally in the comments section<br>
[15:34] &lt;Sid&gt; Which makes it better than using the timestamps why?<br>
[15:35] &lt;Ross&gt; mmmm.....<br>
[15:35] &lt;Ross&gt; doesn't<br>
[15:35] &lt;Sid&gt; Tagging is good when we all work daily on the project<br>
[15:35] &lt;Sid&gt; and we have management saying "Ok, you deliver Build2 on x date"<br>
[15:36] &lt;Sid&gt; But just because we have all these options available doesn't mean we should make the process more complicated.<br>
[15:36] &lt;Sid&gt; Now if we find we've got problems tracking bugs, then I'll be hapy to admit that we need to change something<br>
[15:37] &lt;Sid&gt; but so far I haven't heard of any problems<br>
[15:37] &lt;udaykale&gt; ok<br>
[15:37] &lt;Sid&gt; Ok, #4, status/tasking stuff<br>
[15:38] &lt;Sid&gt; Did schoudel already<br>
[15:38] &lt;Sid&gt; Ross?<br>
[15:38] &lt;Ross&gt; set up teh basic logger stuff<br>
[15:38] &lt;Ross&gt; fixed some more reported bugs deling with where the project is loaded from<br>
[15:38] &lt;Sid&gt; Didn't you do the basic log4j stuff last week?<br>
[15:39] &lt;Ross&gt; (moved resources into a jarrable location<br>
[15:39] &lt;Ross&gt; I did it since the last meeting<br>
[15:39] &lt;Ross&gt; before my cruise<br>
[15:39] &lt;Sid&gt; oh, ok<br>
[15:39] &lt;Ross&gt; that's it<br>
[15:39] &lt;Sid&gt; Ok, on the bigboard you've got several code reviews scheduled<br>
[15:39] &lt;Sid&gt; as far as I know, all those deadlines were met<br>
[15:39] &lt;Ross&gt; sweet<br>
[15:40] &lt;Ross&gt; I'll get reviewing<br>
[15:40] &lt;Sid&gt; Ok, Sandra, you sent out that status email<br>
[15:40] &lt;SandraFaust&gt; yes<br>
[15:41] &lt;Sid&gt; In the new version of JGraph (v2.0)<br>
[15:41] &lt;Sid&gt; they have line-routing algorithms built in<br>
[15:41] &lt;SandraFaust&gt; oh yes?<br>
[15:41] &lt;Sid&gt; But not object-arrangement algs<br>
[15:41] &lt;Sid&gt; Unfortunately, they broke the API so it'll take some work to changeover to the new version<br>
[15:42] &lt;SandraFaust&gt; ok<br>
[15:42] &lt;Sid&gt; As far as tasking goes, keep working on improving the alg<br>
[15:42] &lt;SandraFaust&gt; i assume you didn't get to see the new version of dbviz?<br>
[15:42] &lt;SandraFaust&gt; ok<br>
[15:42] &lt;Sid&gt; I did<br>
[15:42] &lt;Sid&gt; There are some gui issues to consider too, like<br>
[15:43] &lt;Sid&gt; if it's in autolayout mode and you move a table manualy, does it kick you out of autolayout?<br>
[15:43] &lt;Sid&gt; prevent you from moving items? (bad)<br>
[15:43] &lt;SandraFaust&gt; ok, I'll need Ross to review the code to make sure everything is the way it should be<br>
[15:43] &lt;SandraFaust&gt; no, it doesn't<br>
[15:43] &lt;Ross&gt; will do<br>
[15:43] &lt;SandraFaust&gt; it only rearranges the whole thing after you add new table<br>
[15:43] &lt;Sid&gt; Well, I know it doesn't now. We need to figure out what it should do. =)<br>
[15:44] &lt;SandraFaust&gt; ok<br>
[15:44] &lt;Sid&gt; Ok, my status is sorta sorry.<br>
[15:44] &lt;Sid&gt; I was working on that fancy line drawing stuff.<br>
[15:45] &lt;SandraFaust&gt; :)<br>
[15:45] &lt;Sid&gt; But JGraph is designed so "well" using all sortsa design patterns<br>
[15:45] &lt;Sid&gt; that I spent most of the time pondering how to add new stuff to it.<br>
[15:45] &lt;Sid&gt; Its probably worth the effort to have a better design, but it sure doesn't feel like it right now.<br>
[15:46] &lt;Sid&gt; I'm hoping to get that done by tomorrow.<br>
[15:46] &lt;Sid&gt; Though one important issue is that we don't have an Oracle db that everyone can hit when they need to<br>
[15:47] &lt;Ross&gt; speaking of... where's sobby?<br>
[15:47] &lt;Sid&gt; Since my machine is available 24-7, I'm going to install Oracle on it. Or maybe I'll try resurrecting one of the computers that went rhough the fire.<br>
[15:47] &lt;jrarick&gt; Sobby dropped the course<br>
[15:47] &lt;Sid&gt; Ahh, nice of him to inform us.<br>
[15:47] &lt;Ross&gt; oh...<br>
[15:47] &lt;Ross&gt; wow<br>
[15:47] &lt;SandraFaust&gt; did he?<br>
[15:47] &lt;jrarick&gt; I got an email from him last week<br>
[15:48] &lt;jrarick&gt; I tried to work with him and he replied<br>
[15:48] &lt;Sid&gt; Sid is a bastard and I hate him?<br>
[15:48] &lt;Ross&gt; haha<br>
[15:48] &lt;jrarick&gt; No<br>
[15:48] &lt;Ross&gt; Ross is an ass?<br>
[15:48] &lt;jrarick&gt; He just wanted to join a project from the beginning<br>
[15:48] &lt;SandraFaust&gt; you guys were nice to him, weren't you?<br>
[15:49] &lt;Sid&gt; Hmm, took the wrong class for that.<br>
[15:49] &lt;Sid&gt; I started nice. And then I wasn't so nice when he never did what he said he would, or he did it very poorly.<br>
[15:49] &lt;Ross&gt; like the homework<br>
[15:49] &lt;Sid&gt; I always wrote civilized emails to him?<br>
[15:49] &lt;SandraFaust&gt; yes, that's the only thing I know (the homework) and you were quite patient<br>
[15:50] &lt;Sid&gt; *shrugs* ok, that changes things slightly.<br>
[15:50] &lt;Sid&gt; Thanks for the info Jim<br>
[15:50] &lt;jrarick&gt; np<br>
[15:50] &lt;SandraFaust&gt; I wonder if Danny knows<br>
[15:50] &lt;Sid&gt; Anyway, since Sobby was going to be the one setting up the Oracle DB<br>
[15:50] &lt;Sid&gt; I guess I'll definitely spend some time this week on it.<br>
[15:51] &lt;Sid&gt; I'll also need to write up the Iteration Plan C3 and Iteration Assessment C2.<br>
[15:51] &lt;Sid&gt; Larry, I know you wrote the use case up for New Diagram.<br>
[15:51] &lt;larknox&gt; yes<br>
[15:52] &lt;larknox&gt; I revised it again just before the meeting.<br>
[15:52] &lt;larknox&gt; Added references to the import schema and Load saved diagram use cases.<br>
[15:52] &lt;Sid&gt; Hmm, in your last email, you said under the "New Diagram" menu item, we'll have submenus "Import Schema" and "Load schema"<br>
[15:53] &lt;larknox&gt; If that makes sense I'm ready to start trying to implement it.<br>
[15:53] &lt;Sid&gt; I think a better approach would be to have the "New Diagram..." menu item and have it open a dialog box<br>
[15:53] &lt;larknox&gt; I think a couple of main menu changes will be in order, yes?<br>
[15:53] &lt;Sid&gt; that lets you choose whether to import or load.<br>
[15:53] &lt;Sid&gt; We'll have to replace Import with New, yes.<br>
[15:54] &lt;larknox&gt; That's what I had in mind<br>
[15:54] &lt;SandraFaust&gt; Isn't that what the current dialog already ask you for?<br>
[15:54] &lt;Sid&gt; The current dialog is for importing a schema from external representations.<br>
[15:54] &lt;Sid&gt; "Load schema" in this context means that we have an existin Diagram file saved to disk, and we want to copy the schema from that diagram into a new diagram.<br>
[15:54] &lt;SandraFaust&gt; never mind...this is for loading saved schema?<br>
[15:55] &lt;SandraFaust&gt; wouldn't it make sense to call it Open Diagram... ?<br>
[15:55] &lt;Sid&gt; No, this is "New" (diagram). =)<br>
[15:55] &lt;SandraFaust&gt; or open Schema?<br>
[15:55] &lt;Sid&gt; One thing we were pondering last semester<br>
[15:55] &lt;larknox&gt; Can you load the schema from a saved diagram and then build a different diagram?  That's what I was thinking.<br>
[15:56] &lt;Sid&gt; was what sort of model to present to users with schemas and diagrams.<br>
[15:56] &lt;SandraFaust&gt; yes, it seems like it is still confusing. :)<br>
[15:56] &lt;Sid&gt; If you have a diagram, and you save it, and load it again, is the schema saved with it or not<br>
[15:56] &lt;Sid&gt; Yes, the current way is not good<br>
[15:56] &lt;Sid&gt; I want to move to a diagram-centric approach, where the user concentrates on making a diagram.<br>
[15:57] &lt;Sid&gt; It still needs the backing schema, but when they create a diagram, they can grab a schema from multiple sources<br>
[15:57] &lt;SandraFaust&gt; I still think that Open Diagram is better...In Word you open saved document and it comes up with a new window with your documnet. Just my 2c.<br>
[15:57] &lt;Sid&gt; like a database, a SQL file, a currntly=open diagram, or a diagram saved on disk<br>
[15:57] &lt;Sid&gt; Thats a different Use case, Sandra. =)<br>
[15:57] &lt;Sid&gt; This is the New Diagram use case, not Save/Load Diagram.<br>
[15:57] &lt;SandraFaust&gt; ok, I am shutting up :)<br>
[15:58] &lt;larknox&gt; If you choose new diagram, can you use the schema from a saved diagram and create a diffferent diagram with it?<br>
[15:58] &lt;Sid&gt; Yes<br>
[15:59] &lt;Sid&gt; It needs to be as easy as possible to grab that schema so users aren't puzzled about why they need to bother with em<br>
[15:59] &lt;larknox&gt; So that's the one option.  the other is to create a new diagram with schema imported form anywhere else that you can import a schema from.<br>
[15:59] &lt;Sid&gt; Yea, so that makes the Import Schema use case a branch of the New Diagram use case.<br>
[16:00] &lt;Sid&gt; Hmm, we'll need to make some modifications to the use case model...<br>
[16:00] &lt;Sid&gt; Anything else on this topic?<br>
[16:01] &lt;larknox&gt; Will there still be the import Schema coice on the menu?<br>
[16:01] &lt;Sid&gt; No.<br>
[16:01] &lt;larknox&gt; choice<br>
[16:01] &lt;Sid&gt; We'll replace it with "New..."<br>
[16:01] &lt;Sid&gt; The dots are important. They mean a window will appear with more options.<br>
[16:01] &lt;larknox&gt; Ah.  Then is import schema use case just part of new diagram? <br>
[16:01] &lt;larknox&gt; use casde<br>
[16:02] &lt;larknox&gt; case<br>
[16:02] &lt;Sid&gt; Yes, though we'll keep it a separate doc to keep things simpler<br>
[16:02] &lt;larknox&gt; And I'll just efer to it.<br>
[16:02] &lt;Sid&gt; Ya<br>
[16:02] &lt;larknox&gt; refer<br>
[16:02] &lt;larknox&gt; good grief!<br>
[16:02] &lt;Sid&gt; =)<br>
[16:03] &lt;Sid&gt; We'll keep talking offline, to tweak the use case, Larry<br>
[16:03] &lt;Sid&gt; Ok?<br>
[16:03] &lt;larknox&gt; sure<br>
[16:03] &lt;larknox&gt; I noticed something sitting hee playing with dbviz.<br>
[16:04] &lt;Sid&gt; Hmm?<br>
[16:04] &lt;larknox&gt; It failed similarly to the installer version, and I have some error messages.<br>
[16:04] &lt;larknox&gt; I think it involves import.properties.<br>
[16:04] &lt;larknox&gt; I'll talk to Ross later.<br>
[16:04] &lt;Sid&gt; Sounds like you didn't do a "cvs update -dP"<br>
[16:05] &lt;Ross&gt; I think I know what it is<br>
[16:05] &lt;Ross&gt; that's right sid<br>
[16:05] &lt;Sid&gt; Ok, next... Uday<br>
[16:05] &lt;larknox&gt; Oh, then I rbuilt and it worked.<br>
[16:05] &lt;udaykale&gt; well, the database importer that imports schema from oracle database is done.<br>
[16:05] &lt;udaykale&gt; thanks for submitting the bug, sid.<br>
[16:05] &lt;Sid&gt; How much testing have you done with your importer?<br>
[16:05] &lt;udaykale&gt; i also changed the gui part a little - is it beter now<br>
[16:05] &lt;Sid&gt; Yes, it is<br>
[16:06] &lt;udaykale&gt; well enough - i think it works well<br>
[16:06] &lt;udaykale&gt; just manual though<br>
[16:06] &lt;Sid&gt; Though I don't think it'll be appropriate if we allow them to import from Informix using the same options<br>
[16:06] &lt;udaykale&gt; ya, will work on it as per your priority suggestions ;)<br>
[16:07] &lt;Sid&gt; Ok. I sent you some stuff on how we might possibly stick a diagram on disk. Any progress on that?<br>
[16:07] &lt;Sid&gt; This is for the Save/Load Diagram use case, everyone. =)<br>
[16:07] &lt;udaykale&gt; i am just browsing through the code to get a feel of what it would take to save/load a dgm<br>
[16:07] &lt;udaykale&gt; i havent yet gone thru the urls<br>
[16:07] &lt;Sid&gt; Oh, I found out a couple days ago, looking thru the JGraph site<br>
[16:08] &lt;Sid&gt; that JGraph itself doesn't directly support saving a diagram to disk<br>
[16:08] &lt;Sid&gt; It needs to be done by the application<br>
[16:08] &lt;Sid&gt; so just using Java's XML-serialization stuff won't work.<br>
[16:08] &lt;Sid&gt; JGraphPad, a sort of "reference implementation" of how to use JGraph in a real app<br>
[16:08] &lt;Sid&gt; does have a good diagram-save-load function that you might want to look at<br>
[16:09] &lt;udaykale&gt; sure, send me the url, if there is any<br>
[16:09] &lt;Sid&gt; It's off the main jgraph site:  http://jgraph.sourceforge.net<br>
[16:10] &lt;udaykale&gt; ok<br>
[16:10] &lt;Sid&gt; So keep working on learning how it can be implemented, and keep me informed<br>
[16:10] &lt;udaykale&gt; sure<br>
[16:11] &lt;udaykale&gt; I also updated the Use Case on time<br>
[16:11] &lt;udaykale&gt; on 16th i mean :)<br>
[16:11] &lt;Sid&gt; Ahh, okies<br>
[16:11] &lt;Ross&gt; guys I gotto run, sorry<br>
[16:11] &lt;Ross&gt; I'll be in touch<br>
[16:11] * Ross has quit IRC (Quit: Client exiting)<br>
[16:12] &lt;Sid&gt; Ok, Jim, metrics and testing. 'sup?<br>
[16:12] &lt;jrarick&gt; Well, I have been working on code coverage<br>
[16:12] &lt;jrarick&gt; Some functions are easy to unit test, some not so<br>
[16:12] &lt;Sid&gt; The unit tests themselves should be written by the developers.<br>
[16:12] &lt;jrarick&gt; I agree, but many are not<br>
[16:12] &lt;Sid&gt; Well, the ones that can be automated.<br>
[16:13] &lt;jrarick&gt; In the IDE I am using<br>
[16:13] &lt;Sid&gt; Most testing needs to be done just by playing with the app.<br>
[16:13] &lt;jrarick&gt; Okay<br>
[16:13] &lt;jrarick&gt; I was wondering if I should work with individual developers<br>
[16:13] &lt;Sid&gt; And saying "I don't think it should do that!" and posting bugs. =)<br>
[16:13] &lt;jrarick&gt; to make up some unit tests<br>
[16:13] &lt;jrarick&gt; If you think it is worthwhile<br>
[16:14] &lt;Sid&gt; What did you have in mind as far as "make up some unit tests"?<br>
[16:14] &lt;Sid&gt; Writing docs up?<br>
[16:14] &lt;jrarick&gt; No<br>
[16:14] &lt;jrarick&gt; Test that are run in a JUnit suite for a class or package<br>
[16:15] &lt;Sid&gt; Ok, and those're the ones where it'd be very tough to get good functionality coverage.<br>
[16:15] &lt;jrarick&gt; Right<br>
[16:15] &lt;Sid&gt; And most of the time for dbViZ it won't be worth it.<br>
[16:15] &lt;Sid&gt; Things like the importers and the schema itself are good<br>
[16:15] &lt;Sid&gt; but the gui stuff isnt.<br>
[16:15] &lt;jrarick&gt; Okay<br>
[16:15] &lt;jrarick&gt; I have not worked on GUI testing<br>
[16:16] &lt;jrarick&gt; though I can devote my time to that<br>
[16:16] &lt;jrarick&gt; Metrics is straightforward<br>
[16:16] &lt;Sid&gt; Just do GUI testing as far as playin with it<br>
[16:16] &lt;jrarick&gt; No I think we need to find a more systematic way to do it<br>
[16:16] &lt;jrarick&gt; I have not researched it at all though other than the papers<br>
[16:17] &lt;Sid&gt; The main problem with adding "systematic" is that we haven't written down what the exact behavior for the app is.<br>
[16:17] &lt;jrarick&gt; I was wondering if that should be addressed<br>
[16:17] &lt;jrarick&gt; We could take a week to do that<br>
[16:17] &lt;Sid&gt; Itd take more than a week.<br>
[16:17] &lt;SandraFaust&gt; isn't that what the use cases should have covered?<br>
[16:17] &lt;Sid&gt; We don't have use cases for every bit of functionality.<br>
[16:18] &lt;SandraFaust&gt; oh ok<br>
[16:18] &lt;jrarick&gt; It will be hard to say that we have done what we said we would do then, don't you think?<br>
[16:18] &lt;Sid&gt; We can say what's said in the Use Cases has been accomplished.<br>
[16:18] &lt;jrarick&gt; That's right<br>
[16:18] &lt;jrarick&gt; I guess I am looking for direction then<br>
[16:19] &lt;jrarick&gt; I can keep an ongoing metrics report<br>
[16:19] &lt;jrarick&gt; But that is not very much time<br>
[16:19] &lt;Sid&gt; Hmm<br>
[16:19] &lt;Sid&gt; Well, writing the first version will take a lot of time, right?<br>
[16:20] &lt;Sid&gt; If you're going to do, like, at least 5 of the metrics<br>
[16:20] &lt;jrarick&gt; The IDE I am using does 7 metrics in about 10 seconds for all our classes<br>
[16:20] &lt;jrarick&gt; I have imported it into that<br>
[16:20] &lt;Sid&gt; But that's not an analysis of the metrics.<br>
[16:20] &lt;Sid&gt; Its just the raw numbers.<br>
[16:21] &lt;jrarick&gt; That is true<br>
[16:21] &lt;jrarick&gt; Analysis will take more time.<br>
[16:21] &lt;jrarick&gt; I can do that pretty easily<br>
[16:21] &lt;jrarick&gt; One problem I had was finding an Excel <br>
[16:21] &lt;Sid&gt; How fast is your computer, Jim?<br>
[16:22] &lt;jrarick&gt; 700 MHz<br>
[16:22] &lt;jrarick&gt; My work is 2.0 GHz<br>
[16:22] &lt;jrarick&gt; But I never use that (sure)<br>
[16:22] &lt;Sid&gt; Hmm. I think we said this thing'd work reasonably fast on a 400MHz system<br>
[16:23] &lt;SandraFaust&gt; i am running dbviz on 400mhz<br>
[16:23] &lt;Sid&gt; How is it? =)<br>
[16:23] &lt;SandraFaust&gt; it's not bad<br>
[16:23] &lt;jrarick&gt; It will depend on the database size, right<br>
[16:23] &lt;SandraFaust&gt; sometimes it gets sluggish but normally is fine<br>
[16:23] &lt;Sid&gt; Well, there are a lot of things that need to be "fast"<br>
[16:24] &lt;Sid&gt; Import from SQL file depends on general system speed<br>
[16:24] &lt;SandraFaust&gt; i'll need to install it to the other comp to see how it runs on 2ghz :)<br>
[16:24] &lt;Sid&gt; Oracle import depends on the speeds of the 2 systems and network<br>
[16:24] &lt;Sid&gt; lol<br>
[16:24] &lt;Sid&gt; But we need big test schemas to be able to actually measure anything<br>
[16:25] &lt;Sid&gt; Jim, can you work on the Metrics report for a few days<br>
[16:25] &lt;jrarick&gt; Sure<br>
[16:25] &lt;Sid&gt; and we'll revisit this issue then?<br>
[16:25] &lt;SandraFaust&gt; well, if you want me to test things in terms of speed, that's no problem<br>
[16:25] &lt;jrarick&gt; Absolutely, and we will keep in touch<br>
[16:25] &lt;Sid&gt; That sounds good, Sandra<br>
[16:25] &lt;jrarick&gt; I have a question about Excel<br>
[16:25] &lt;SandraFaust&gt; sure, no problem<br>
[16:26] &lt;jrarick&gt; Is there a way to count the number of values in a range <br>
[16:26] &lt;Sid&gt; Histogram?<br>
[16:26] &lt;jrarick&gt; The Excel histogram I used had some strange division going on<br>
[16:26] &lt;jrarick&gt; I will look at it again<br>
[16:26] &lt;jrarick&gt; It could have been how my axis were setup<br>
[16:27] &lt;jrarick&gt; Anyway, I'll figure it out<br>
[16:27] &lt;jrarick&gt;  and get some nice graphs<br>
[16:27] &lt;Sid&gt; I'm not sure otherwise.<br>
[16:27] &lt;Sid&gt; =)<br>
[16:27] &lt;Sid&gt; yay, pretty graphs<br>
[16:27] &lt;Sid&gt; Ok, anything else anyone?<br>
[16:27] &lt;SandraFaust&gt; is this better time for meeting? I think I like it better, than late Thursday nights<br>
[16:28] &lt;jrarick&gt; I like it better<br>
[16:28] &lt;Sid&gt; Ok, quick vote, who likes Sat aft better?<br>
[16:28] &lt;udaykale&gt; fine with me<br>
[16:28] &lt;Sid&gt; Larry?<br>
[16:28] &lt;SandraFaust&gt; we'll need to check with abentiees too :)<br>
[16:28] * Sid slaps larknox around a bit with a large trout<br>
[16:28] &lt;larknox&gt; It worked today, not sure about always.  If you want to try it I can let you know wahead when I can't<br>
[16:28] &lt;SandraFaust&gt; absentees<br>
[16:29] &lt;Sid&gt; I'll check with Ross and Schoudel<br>
[16:29] &lt;Sid&gt; I'll send an email out<br>
[16:29] &lt;SandraFaust&gt; ok<br>
[16:29] &lt;SandraFaust&gt; it's not a big deal<br>
[16:29] &lt;Sid&gt; Everyone, don't forget to try to play with dbViZ and try to break it!<br>
[16:30] &lt;SandraFaust&gt; :)<br>
[16:30] &lt;Sid&gt; Meetings' done. Thanks everyone<br>
[16:30] &lt;udaykale&gt; bye<br>
[16:30] &lt;jrarick&gt; see ya all<br>
[16:30] &lt;larknox&gt; Thnak you!<br>
[16:30] * udaykale has quit IRC (Quit: )<br>
[16:30] &lt;SandraFaust&gt; bye bye, have a good weekend<br>
[16:30] * jrarick has quit IRC (Quit: )<br>
[16:30] &lt;larknox&gt; bye<br>
[16:30] * SandraFaust has quit IRC (Quit: )<br>
[16:30] * larknox has quit IRC (Quit: )<br>
[16:30] * Disconnected<br>
Session Close: Sat Mar 22 16:30:54 2003<br>