[67cat] Project scan issues.
|
View:
New views
7 Messages
—
Rating Filter:
Alert me
|
| < Prev | 1 - 2 | Next > |
 |
Re: [67cat] Project scan issues.
by Anuradha G
::
Rate this Message:
+1
On Tue, Jun 23, 2009 at 8:45 PM, Scott Palmer <Scott@...> wrote: +1 -- Anuradha G |
|
|
Re: [67cat] Project scan issues.
by HakSun Kim
::
Rate this Message:
+1
2009/6/24 Anuradha G <theanuradha@...> +1 -- Thanks and best regards, Sun Kim |
|
|
Re: [67cat] Project scan issues.
by rohitbrai
::
Rate this Message:
+1
As I said, we should have a release dedicated to performance improvements. Eclipse has done it in early days once, I remember around 3-4 year back. The diamond cannot be polished without friction, nor man perfected without trials http://mytechrantings.blogspot.com On Wed, Jun 24, 2009 at 10:13 AM, HakSun Kim <glarian@...> wrote: +1 |
|
|
Re: [67cat] Project scan issues.
by marianm
::
Rate this Message:
Hi all,
thanks for your feedback. I would like to ask you to report as many of cases mentioned by you in this thread as possible into IssueZilla. This will help us to address/investigate all problems case by case in the next NB release. FYI ... for NB 6.8 we plan to focus on Quality & Performance in Editor : http://wiki.netbeans.org/EditorPlan68 Thanks, *Marian Petr Dvorak wrote: > Hi everyone, > > you can help us by submitting a bug reports on this - I have already > posted this blog post many times but I guess it is never enough: > > http://blogs.sun.com/joshis/entry/netbeans_ide_is_scanning_parsing > > It describe what specific things you need to do in order to submit a > useful bug report on scanning... > > With regards, > > Petr Dvorak > > Rohit Rai wrote: >> +1 >> >> As I said, we should have a release dedicated to performance >> improvements. Eclipse has done it in early days once, I remember >> around 3-4 year back. >> >> >> The diamond cannot be polished without friction, nor man perfected >> without trials >> http://mytechrantings.blogspot.com >> >> >> On Wed, Jun 24, 2009 at 10:13 AM, HakSun Kim <glarian@... >> <mailto:glarian@...>> wrote: >> >> +1 >> >> 2009/6/24 Anuradha G <theanuradha@... >> <mailto:theanuradha@...>> >> >> +1 >> >> >> On Tue, Jun 23, 2009 at 8:45 PM, Scott Palmer >> <Scott@... <mailto:Scott@...>> >> wrote: >> >> +1 >> - Performance ("Scanning...", "Please wait", "what >> happened to the >> Navigation window?") >> - Stability (how many NPEs have encountered?) >> - Correctness (error badges, debugger display values) >> ... are the main "problems" with NetBeans. >> >> Fluff like Kenai integration may make for a neat demo, but >> it is 100% >> useless to me in my day job doing closed-source projects >> that actually >> make money ;-) >> It's not that I don't appreciate many of the new >> features.. but >> performance has been significantly behind the competition >> for years. >> It's time to catch up. >> >> Scott >> >> >> > -----Original Message----- >> > From: Matthies, Niklas >> [mailto:niklas.matthies@... >> <mailto:niklas.matthies@...>] >> > Sent: June 23, 2009 10:33 AM >> > To: netcat@... <mailto:netcat@...> >> > Subject: RE: [67cat] Project scan issues. >> > >> > I agree, but I'm not holding my breath. >> > >> > Getting up to Eclipse performance levels for project >> scanning and code >> > editing/navigation/completion should really be priority >> #1 for 6.8, >> > though. In the current state it's almost impossible to >> get someone >> used >> > to Eclipse to use NetBeans, and NetBeans users are prone >> to jump ship >> > due to these performance issues with NB 6.x that are >> getting worse >> from >> > release to release. >> > >> > -- Niklas Matthies >> > >> > >> > On Tue, 23. Jun 2009 01:56, nleck wrote: >> > > Please see bug#167399 >> > > (http://www.netbeans.org/issues/show_bug.cgi?id=167399) >> > > >> > > >> > > >> > > I really don't think we should release with know >> performance >> > > regressions. 6.5 performance was bad when compared to >> Eclipse >> > > and now 6.7 will be worst. >> >> >> >> >> -- Anuradha G >> >> >> >> >> -- Thanks and best regards, >> Sun Kim >> >> > |
|
|
Re: [67cat] Project scan issues.
by Glenn Holmer
::
Rate this Message:
On Wed, 2009-06-24 at 11:54 +0530, Rohit Rai wrote:
> As I said, we should have a release dedicated to performance improvements. Laudable, to be sure, but the real world gets in the way sometimes. 6.8 is going to give us JEE6 support (which is absolutely vital), so it makes no sense to suggest that for the next release. In addition, of course, Sun has been in a bit of a stressful situation recently as we all know, and their personnel cuts extended to the NetBeans team. Let's help them out any way we can by doing as Petr suggested. -- ____________________________________________________________ Glenn Holmer gholmer@... Software Engineer phone: 414-908-1809 Weyco Group, Inc. fax: 414-908-1601 |
|
|
RE: [67cat] Project scan issues.
by Scott Palmer-2
::
Rate this Message:
I think that the daily builds moving in a helpful direction - though I
haven't tried one recently. The auto detection of a slow operation is a good start, but it can go further. In Petr's blog there are a few suggestions that could be automated. "How long did scanning take?", "Was it the first scan, or whenever you open the project?" I think those things could be logged automatically. Surely something as simple as timing how long the scan took can be automated and the data could be available to automatically report, along with simple metrics like how many source files & lines of code there are in the open projects. I think if a developer ever notices a scanning delay after the initial scan then there better be a really good reason for it -- like you just did a massive wide spread refactoring or something. Once the database is built from the initial scan I would think that the computer should be able to keep up with my coding fast enough that a scanning delay is simply never long enough to notice. Things like getting the Call hierarchy, Go To, and Find Usages should be blazingly fast because any non-private identifiers would be indexed and that index could be kept up-to-date on the fly through incremental updates as I edit. And of course those incremental updates should never block what I'm doing. It's easy to say, I don't know how easy it is to implement given what is already in place... but that is the direction I envision things heading. Scott > -----Original Message----- > From: Glenn Holmer [mailto:gholmer@...] > Sent: June 24, 2009 9:22 AM > To: netcat@... > Subject: Re: [67cat] Project scan issues. > > On Wed, 2009-06-24 at 11:54 +0530, Rohit Rai wrote: > > > As I said, we should have a release dedicated to performance > improvements. > > Laudable, to be sure, but the real world gets in the way sometimes. > 6.8 > is going to give us JEE6 support (which is absolutely vital), so it > makes no sense to suggest that for the next release. > > In addition, of course, Sun has been in a bit of a stressful situation > recently as we all know, and their personnel cuts extended to the > NetBeans team. Let's help them out any way we can by doing as Petr > suggested. > > -- > ____________________________________________________________ > Glenn Holmer gholmer@... > Software Engineer phone: 414-908-1809 > Weyco Group, Inc. fax: 414-908-1601 > > |
|
|
Re: [67cat] Project scan issues.
by esmithbss
::
Rate this Message:
I agree with Scott on how much detecting can be automated. Many of the long operations I've experienced (project scanning, code-completion list creation, Inter-panel updates/communications) should be discrete enough within the code base to bracket with a timing block.
I think, but I'm not sure, that at least 60 or 70% of my speed issues could be fixed if the keystroke monitor (where the timing for auto-deployment of code-completion, and any other "automatic" actions occurs) was cleaned up. This could be as simple as adding additional wait time between keystrokes before invoking automatic actions or as complex as re-threading the dispatcher. I don't know the details on how to fix it. I just know that from observed behavior, if I pause for any period of time between keystokes, my system is effectively dead until the code-completion window appears. Eric M. Smith Burning Sun Enterprises http://typicalisoverrated.com On Wed, Jun 24, 2009 at 9:00 AM, Scott Palmer <Scott@...> wrote: I think that the daily builds moving in a helpful direction - though I |
| < Prev | 1 - 2 | Next > |
| Free embeddable forum powered by Nabble | Forum Help |
