Thread: Problem with web TODO list
Why is the web TODO list in ASCII?
http://www.postgresql.org/docs/faqs.TODO.html
It should be pulling from my home directory with ~momjian/TODO.html.
--
Bruce Momjian | http://candle.pha.pa.us
pgman@candle.pha.pa.us | (610) 359-1001
+ If your life is a hard drive, | 13 Roberts Road
+ Christ can be your backup. | Newtown Square, Pennsylvania 19073
Because all the FAQs are pulled from the main CVS (we don't pull in anything automatically that isn't in CVS) - and the onlyversion there is plain text http://developer.postgresql.org/cvsweb.cgi/pgsql/doc/ Regards, Dave -----Original Message----- From: pgsql-www-owner@postgresql.org on behalf of Bruce Momjian Sent: Sun 4/10/2005 5:24 AM To: PostgreSQL www Subject: [pgsql-www] Problem with web TODO list Why is the web TODO list in ASCII? http://www.postgresql.org/docs/faqs.TODO.html It should be pulling from my home directory with ~momjian/TODO.html. -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup. | Newtown Square, Pennsylvania 19073 ---------------------------(end of broadcast)--------------------------- TIP 6: Have you searched our list archives? http://archives.postgresql.org
Dave Page wrote: > Because all the FAQs are pulled from the main CVS (we don't pull in anything automatically that isn't in CVS) - and theonly version there is plain text > > http://developer.postgresql.org/cvsweb.cgi/pgsql/doc/ OK, so you want the HTML in CVS? I can do that. -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup. | Newtown Square, Pennsylvania 19073
>> Because all the FAQs are pulled from the main CVS (we don't >pull in anything automatically that isn't in CVS) - and the >only version there is plain text >> >> http://developer.postgresql.org/cvsweb.cgi/pgsql/doc/ > >OK, so you want the HTML in CVS? I can do that. Yup, if it's going to go on the site in HTML format at least ;-). If you put it where the other HTML format FAQs are (doc/src/FAQ/), it will automatically be picked up on the site after the next cvs pull (once / hour from the anoncvs mirror). Please make sure you make the same kind of cleanups done to the other HTML files (missing tags and invalid attributes) before you put it in :-) //Magnus
>> >> Because all the FAQs are pulled from the main CVS (we don't >> >pull in anything automatically that isn't in CVS) - and the >> >only version there is plain text >> >> >> >> http://developer.postgresql.org/cvsweb.cgi/pgsql/doc/ >> > >> >OK, so you want the HTML in CVS? I can do that. >> >> Yup, if it's going to go on the site in HTML format at least >;-). If you >> put it where the other HTML format FAQs are (doc/src/FAQ/), it will >> automatically be picked up on the site after the next cvs >pull (once / >> hour from the anoncvs mirror). >> >> Please make sure you make the same kind of cleanups done to the other >> HTML files (missing tags and invalid attributes) before you put it in >> :-) > >In this case the HTML is derived from the text file, which is the >reverse of the FAQs. Yes, that is because it's not present in CVS in HTML format. The site code first searches the HTML versions and if it's not found there, it searches the plaintext versions. But it will bring anything it finds in the HTML version, so the source that goes into cvs needs to be clean. //Magnus
Magnus Hagander wrote:
> >> HTML files (missing tags and invalid attributes) before you put it in
> >> :-)
> >
> >In this case the HTML is derived from the text file, which is the
> >reverse of the FAQs.
>
> Yes, that is because it's not present in CVS in HTML format. The site
> code first searches the HTML versions and if it's not found there, it
> searches the plaintext versions.
>
> But it will bring anything it finds in the HTML version, so the source
> that goes into cvs needs to be clean.
Please look on my current HTML TODO version:
http://developer.postgresql.org/~momjian/TODO.html
Because it is generated from text, I can't clean it up. What are our
options?
--
Bruce Momjian | http://candle.pha.pa.us
pgman@candle.pha.pa.us | (610) 359-1001
+ If your life is a hard drive, | 13 Roberts Road
+ Christ can be your backup. | Newtown Square, Pennsylvania 19073
>> >> HTML files (missing tags and invalid attributes) before >you put it in >> >> :-) >> > >> >In this case the HTML is derived from the text file, which is the >> >reverse of the FAQs. >> >> Yes, that is because it's not present in CVS in HTML format. The site >> code first searches the HTML versions and if it's not found there, it >> searches the plaintext versions. >> >> But it will bring anything it finds in the HTML version, so >the source >> that goes into cvs needs to be clean. > >Please look on my current HTML TODO version: > > http://developer.postgresql.org/~momjian/TODO.html > >Because it is generated from text, I can't clean it up. What are our >options? Hmm. one main problem is that it doesn't close the <P> or <LI> tags properly. What do you use to generate it from text? //Magnus
Magnus Hagander wrote: > >> Because all the FAQs are pulled from the main CVS (we don't > >pull in anything automatically that isn't in CVS) - and the > >only version there is plain text > >> > >> http://developer.postgresql.org/cvsweb.cgi/pgsql/doc/ > > > >OK, so you want the HTML in CVS? I can do that. > > Yup, if it's going to go on the site in HTML format at least ;-). If you > put it where the other HTML format FAQs are (doc/src/FAQ/), it will > automatically be picked up on the site after the next cvs pull (once / > hour from the anoncvs mirror). > > Please make sure you make the same kind of cleanups done to the other > HTML files (missing tags and invalid attributes) before you put it in > :-) In this case the HTML is derived from the text file, which is the reverse of the FAQs. I will put what I have into CVS in the doc/src/FAQ directory. -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup. | Newtown Square, Pennsylvania 19073
Magnus Hagander wrote: > >> >> HTML files (missing tags and invalid attributes) before > >you put it in > >> >> :-) > >> > > >> >In this case the HTML is derived from the text file, which is the > >> >reverse of the FAQs. > >> > >> Yes, that is because it's not present in CVS in HTML format. The site > >> code first searches the HTML versions and if it's not found there, it > >> searches the plaintext versions. > >> > >> But it will bring anything it finds in the HTML version, so > >the source > >> that goes into cvs needs to be clean. > > > >Please look on my current HTML TODO version: > > > > http://developer.postgresql.org/~momjian/TODO.html > > > >Because it is generated from text, I can't clean it up. What are our > >options? > > Hmm. one main problem is that it doesn't close the <P> or <LI> tags > properly. What do you use to generate it from text? I use txt2html at http://txt2html.sourceforge.net/. I just found an --xhtml flag to the utility. I added that and regenerated the HTML at the above URL. WOuld you check it again? Thanks. -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup. | Newtown Square, Pennsylvania 19073
Bruce, > I use txt2html at http://txt2html.sourceforge.net/. > > I just found an --xhtml flag to the utility. I added that and > regenerated the HTML at the above URL. WOuld you check it again? > Thanks. Are we trying to fix this so it will be able to be pasted into the main website? If so, I have some tweaks to txt2html I did for PWN. -- Josh Berkus Aglio Database Solutions San Francisco
Josh Berkus wrote: > Bruce, > > > I use txt2html at http://txt2html.sourceforge.net/. > > > > I just found an --xhtml flag to the utility. ?I added that and > > regenerated the HTML at the above URL. ?WOuld you check it again? > > Thanks. > > Are we trying to fix this so it will be able to be pasted into the main > website? Yes. > If so, I have some tweaks to txt2html I did for PWN. Oh, what were they? -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup. | Newtown Square, Pennsylvania 19073
Bruce, > Yes. > > > If so, I have some tweaks to txt2html I did for PWN. > > Oh, what were they? A bunch of switches and a shell script to fix some formatting problems which txt2html doesn't do. If you can link me the text version, I'll prepare a package for you that takes care of it. -- Josh Berkus Aglio Database Solutions San Francisco
Josh Berkus wrote: > Bruce, > > > Yes. > > > > > If so, I have some tweaks to txt2html I did for PWN. > > > > Oh, what were they? > > A bunch of switches and a shell script to fix some formatting problems which > txt2html doesn't do. If you can link me the text version, I'll prepare a > package for you that takes care of it. The text version is already in CVS. -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup. | Newtown Square, Pennsylvania 19073
> > >> >> HTML files (missing tags and invalid attributes) before > > >you put it in > > >> >> :-) > > >> > > > >> >In this case the HTML is derived from the text file, > which is the > > >> >reverse of the FAQs. > > >> > > >> Yes, that is because it's not present in CVS in HTML format. The > > >> site code first searches the HTML versions and if it's not found > > >> there, it searches the plaintext versions. > > >> > > >> But it will bring anything it finds in the HTML version, so > > >the source > > >> that goes into cvs needs to be clean. > > > > > >Please look on my current HTML TODO version: > > > > > > http://developer.postgresql.org/~momjian/TODO.html > > > > > >Because it is generated from text, I can't clean it up. > What are our > > >options? > > > > Hmm. one main problem is that it doesn't close the <P> or <LI> tags > > properly. What do you use to generate it from text? > > I use txt2html at http://txt2html.sourceforge.net/. > > I just found an --xhtml flag to the utility. I added that > and regenerated the HTML at the above URL. WOuld you check it again? In general that should work much better. but that page has *two* <body> tags! One within the <head> section. Either you missed a cut/paste somewhere, or that package is *very* broken. And it's not valid XHTML, which really should happen with --xhtml, no? Even if I fix that, I get some invalid uses of <LI> in the output. http://validator.w3.org/check?uri=http%3A%2F%2Fdeveloper.postgresql.org% 2F%7Emomjian%2FTODO.html&charset=%28detect+automatically%29&doctype=%28d etect+automatically%29 (it really should validate as XHTML strict, but it doesn't even manage XHTML Transitional) //Magnus
> > I use txt2html at http://txt2html.sourceforge.net/. > > > > I just found an --xhtml flag to the utility. I added that and > > regenerated the HTML at the above URL. WOuld you check it again? > > Thanks. > > Are we trying to fix this so it will be able to be pasted > into the main > website? > > If so, I have some tweaks to txt2html I did for PWN. Not sure they all apply. The PWN stuff has to render natively, whereas the FAQs are post-processed by the site code (becausethey have to work both inside the siet as a wrapper and as standalone HTML documents) //Magnus
Magnus, > Not sure they all apply. The PWN stuff has to render natively, whereas the > FAQs are post-processed by the site code (because they have to work both > inside the siet as a wrapper and as standalone HTML documents) I thought we were talking about the TODO list, not the FAQs? And let me pitch Bruce again to database the TODO list .... -- Josh Berkus Aglio Database Solutions San Francisco
> > Not sure they all apply. The PWN stuff has to render > natively, whereas > > the FAQs are post-processed by the site code (because they have to > > work both inside the siet as a wrapper and as standalone HTML > > documents) > > I thought we were talking about the TODO list, not the FAQs? We are. But the TODO list is in the FAQ directory, and is handled just like the FAQs. > And let me pitch Bruce again to database the TODO list .... Um. Let me second that ;-) But I've also tried before :-) //Magnus
Magnus Hagander wrote: > > > Not sure they all apply. The PWN stuff has to render > > natively, whereas > > > the FAQs are post-processed by the site code (because they have to > > > work both inside the siet as a wrapper and as standalone HTML > > > documents) > > > > I thought we were talking about the TODO list, not the FAQs? > > We are. But the TODO list is in the FAQ directory, and is handled just > like the FAQs. > > > > And let me pitch Bruce again to database the TODO list .... > > Um. Let me second that ;-) But I've also tried before :-) Go ahead. I find a database is more work but if someone else wants to try it, they are welcome. I will continue maintaining the TODO for a while just in case the new idea doesn't work out, like it didn't last time. :-) -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup. | Newtown Square, Pennsylvania 19073
> > > > Not sure they all apply. The PWN stuff has to render > > > natively, whereas > > > > the FAQs are post-processed by the site code (because > they have to > > > > work both inside the siet as a wrapper and as standalone HTML > > > > documents) > > > > > > I thought we were talking about the TODO list, not the FAQs? > > > > We are. But the TODO list is in the FAQ directory, and is > handled just > > like the FAQs. > > > > > > > And let me pitch Bruce again to database the TODO list .... > > > > Um. Let me second that ;-) But I've also tried before :-) > > Go ahead. I find a database is more work but if someone else > wants to try it, they are welcome. I will continue > maintaining the TODO for a while just in case the new idea > doesn't work out, like it didn't last time. :-) Actually, the even more important things about this I beleive are: 1) Unless you, who are actually going to maintain it, think it's good, it's never going to work. And I'm certainly no saying someone else (expecially not me! :P) should maintain it... (while it could perhaps be shared, but that can be done either way) 2) I *personally* think it should also be combined with being some kind of bug tracker (more along the line of a general issue tracker), and that's *definitly* going to fly unless the people who would use it most (-core and other main developers) want it. And the archives are full of discussinos about that ;-) If you're interested in giving at least (1) a try, let me know. But since you're the guy who maintains it (and has been for a long time), you're the one who knows what the specs will have to be so it fullfills your requiremetns :-) //Magnus
Magnus Hagander wrote: > > > >Please look on my current HTML TODO version: > > > > > > > > http://developer.postgresql.org/~momjian/TODO.html > > > > > > > >Because it is generated from text, I can't clean it up. > > What are our > > > >options? > > > > > > Hmm. one main problem is that it doesn't close the <P> or <LI> tags > > > properly. What do you use to generate it from text? > > > > I use txt2html at http://txt2html.sourceforge.net/. > > > > I just found an --xhtml flag to the utility. I added that > > and regenerated the HTML at the above URL. WOuld you check it again? > > In general that should work much better. but that page has *two* <body> > tags! One within the <head> section. Either you missed a cut/paste > somewhere, or that package is *very* broken. > > And it's not valid XHTML, which really should happen with --xhtml, no? > Even if I fix that, I get some invalid uses of <LI> in the output. > > http://validator.w3.org/check?uri=http%3A%2F%2Fdeveloper.postgresql.org% > 2F%7Emomjian%2FTODO.html&charset=%28detect+automatically%29&doctype=%28d > etect+automatically%29 > > (it really should validate as XHTML strict, but it doesn't even manage > XHTML Transitional) Would you please check the above URL again? I think I fixed all the problems with the TODO list HTML. It validated with XHTML transitional, and failed with XHTML strict only because of the BODY color tags. If it is OK I will add the TODO.html to CVS in doc/src/FAQ/. -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup. | Newtown Square, Pennsylvania 19073
Magnus Hagander wrote: > > Go ahead. I find a database is more work but if someone else > > wants to try it, they are welcome. I will continue > > maintaining the TODO for a while just in case the new idea > > doesn't work out, like it didn't last time. :-) > > Actually, the even more important things about this I beleive are: > > 1) Unless you, who are actually going to maintain it, think it's good, > it's never going to work. And I'm certainly no saying someone else > (expecially not me! :P) should maintain it... (while it could perhaps be > shared, but that can be done either way) > > 2) I *personally* think it should also be combined with being some kind > of bug tracker (more along the line of a general issue tracker), and > that's *definitly* going to fly unless the people who would use it most > (-core and other main developers) want it. And the archives are full of > discussinos about that ;-) > > > If you're interested in giving at least (1) a try, let me know. But > since you're the guy who maintains it (and has been for a long time), > you're the one who knows what the specs will have to be so it fullfills > your requiremetns :-) The big issue here is time investment vs. payback. On the one extreme, you have what we have now --- a flat file that I can edit in 2 seconds and push into CVS and the web with one command. On the other end, you have the bugzilla monster that is very hard to understand let alone use. The issue is what do you want to track, and how much time is it worth expending to track it. Personally, I am just fine with the current system. I don't have enough time to do the important things, let alone time to track who requested certain features and why, and I don't think it is useful for me to change that. If others want to expend the time, I will support them in hopes that they will expend the time to track it and I will not have to be as involved. -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup. | Newtown Square, Pennsylvania 19073
Bruce, > The issue is what do you want to track, and how much time is it worth > expending to track it. Personally, I am just fine with the current > system. I don't have enough time to do the important things, let alone > time to track who requested certain features and why, and I don't think > it is useful for me to change that. If others want to expend the time, > I will support them in hopes that they will expend the time to track it > and I will not have to be as involved. Agreed, bugzilla isn't going to help. Frankly, I haven't see the software that *would* help. I don't think it's been written yet. What we need is software that: 1) let us track various TODOs 2) kept draft specifications for those TODOs 3) kept partial code when contributed 4) linked to e-mail discussions 5) tracked who'd volunteered to implement and/or test features 6) tracked which features were dependant on which other features being implemented. 7) was searchable That, IMHO, would make it worth a little extra effort to enter stuff. And you could draft help. -- Josh Berkus Aglio Database Solutions San Francisco
> > > > >Please look on my current HTML TODO version: > > > > > > > > > > http://developer.postgresql.org/~momjian/TODO.html > > > > > > > > > >Because it is generated from text, I can't clean it up. > > > What are our > > > > >options? > > > > > > > > Hmm. one main problem is that it doesn't close the <P> or <LI> > > > > tags properly. What do you use to generate it from text? > > > > > > I use txt2html at http://txt2html.sourceforge.net/. > > > > > > I just found an --xhtml flag to the utility. I added that and > > > regenerated the HTML at the above URL. WOuld you check it again? > > > > In general that should work much better. but that page has *two* > > <body> tags! One within the <head> section. Either you missed a > > cut/paste somewhere, or that package is *very* broken. > > > > And it's not valid XHTML, which really should happen with > --xhtml, no? > > Even if I fix that, I get some invalid uses of <LI> in the output. > > > > > http://validator.w3.org/check?uri=http%3A%2F%2Fdeveloper.postgresql.or > > g% > > > 2F%7Emomjian%2FTODO.html&charset=%28detect+automatically%29&doctype=%2 > > 8d > > etect+automatically%29 > > > > (it really should validate as XHTML strict, but it doesn't > even manage > > XHTML Transitional) > > Would you please check the above URL again? I think I fixed > all the problems with the TODO list HTML. It validated with > XHTML transitional, and failed with XHTML strict only because > of the BODY color tags. > > If it is OK I will add the TODO.html to CVS in doc/src/FAQ/. Yup, that looks good. Validates as XHTML strict when loaded in the framework. Go for it. //Magnus
On Sun, 2005-04-17 at 21:51 -0700, Josh Berkus wrote: > Agreed, bugzilla isn't going to help. Frankly, I haven't see the software > that *would* help. I don't think it's been written yet. What we need is > software that: > > 1) let us track various TODOs > 2) kept draft specifications for those TODOs > 3) kept partial code when contributed > 4) linked to e-mail discussions > 5) tracked who'd volunteered to implement and/or test features > 6) tracked which features were dependant on which other features being > implemented. > 7) was searchable Sounds a bit like the RFE tracker on a PGFoundry project... Yours, Tom
OK, the TODO on the web site is updated to HTML. It looks great! Thanks Magnus. --------------------------------------------------------------------------- Magnus Hagander wrote: > >> Because all the FAQs are pulled from the main CVS (we don't > >pull in anything automatically that isn't in CVS) - and the > >only version there is plain text > >> > >> http://developer.postgresql.org/cvsweb.cgi/pgsql/doc/ > > > >OK, so you want the HTML in CVS? I can do that. > > Yup, if it's going to go on the site in HTML format at least ;-). If you > put it where the other HTML format FAQs are (doc/src/FAQ/), it will > automatically be picked up on the site after the next cvs pull (once / > hour from the anoncvs mirror). > > Please make sure you make the same kind of cleanups done to the other > HTML files (missing tags and invalid attributes) before you put it in > :-) > > //Magnus > -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 359-1001 + If your life is a hard drive, | 13 Roberts Road + Christ can be your backup. | Newtown Square, Pennsylvania 19073
>> Agreed, bugzilla isn't going to help. Frankly, I haven't >see the software >> that *would* help. I don't think it's been written yet. >What we need is >> software that: >> >> 1) let us track various TODOs >> 2) kept draft specifications for those TODOs >> 3) kept partial code when contributed >> 4) linked to e-mail discussions >> 5) tracked who'd volunteered to implement and/or test features >> 6) tracked which features were dependant on which other >features being >> implemented. >> 7) was searchable > >Sounds a bit like the RFE tracker on a PGFoundry project... Really? In my experience, those trackers certainly make me understand Bruce's not wanting to use them ;) AFAIK, they're missing a lot of the requirements Josh put forward, and IMHO (but that's MHO only, of course) are not very flexible to work with ;-) (for example, the fact that it remails the entire item on any change at all annoys me very much :P) It may be closer than Bugzilla, but I think it's pretty far off still... //Magnus
On Mon, 2005-04-18 at 19:03 +0200, Magnus Hagander wrote: > >> 1) let us track various TODOs > >> 2) kept draft specifications for those TODOs > >> 3) kept partial code when contributed > >> 4) linked to e-mail discussions > >> 5) tracked who'd volunteered to implement and/or test features > >> 6) tracked which features were dependant on which other > >features being > >> implemented. > >> 7) was searchable > > > >Sounds a bit like the RFE tracker on a PGFoundry project... > > It may be closer than Bugzilla, but I think it's pretty far off still... Yup, the tracker doesn't do some of that stuff... like #4 and #6. Maybe the tasks thingy is better; it does dependencies: http://tinyurl.com/6v32z Its primary virtue may be that it currently exists :-) Yours, Tom
>> >> 1) let us track various TODOs >> >> 2) kept draft specifications for those TODOs >> >> 3) kept partial code when contributed >> >> 4) linked to e-mail discussions >> >> 5) tracked who'd volunteered to implement and/or test features >> >> 6) tracked which features were dependant on which other >> >features being >> >> implemented. >> >> 7) was searchable >> > >> >Sounds a bit like the RFE tracker on a PGFoundry project... >> >> It may be closer than Bugzilla, but I think it's pretty far >off still... > >Yup, the tracker doesn't do some of that stuff... like #4 and >#6. It does #7 now?! tell me how :D I've been longing for this! (I'd say that is one of the most important features!) > Maybe >the tasks thingy is better; it does dependencies: > >http://tinyurl.com/6v32z Hmm. Could be. I still don'tt hink it'll cut it all the wya, though. >Its primary virtue may be that it currently exists :-) Oh, most definitly :-) But I think we'll need something that's closer to "a perfect match". You need a tool that's adapted to the way the development works, not the other way around... Also, I'm having a hard time seeing this one scale to the size of postgresql (not technically - I'm sure it can do that, though it's dog slow on pgFoundry I'm sure it can be made fast on different hardware - I'm talking about usability-wise. The TODO list currently has somewhere around 250 items, and if you eventually want to bring in all bugs from pgsql-bugs etc.. //Magnus
On Mon, 2005-04-18 at 19:33 +0200, Magnus Hagander wrote: > It does #7 now?! tell me how :D I've been longing for this! (I'd say > that is one of the most important features!) If you go to, say, the support project bug tracker: http://pgfoundry.org/tracker/?atid=146&group_id=1000013&func=browse there's a dropdown list at the top of the page and a textfield to put a search keyword in. This dropdown list is context-sensitive, so if you're viewing a forum, the "bugs" option will go away and "this forum" or something to that effect will appear. > >http://tinyurl.com/6v32z > > Hmm. Could be. I still don'tt hink it'll cut it all the wya, though. Yup, true, probably not. > Oh, most definitly :-) But I think we'll need something that's closer to > "a perfect match". You need a tool that's adapted to the way the > development works, not the other way around... Also, I'm having a hard > time seeing this one scale to the size of postgresql (not technically - > I'm sure it can do that, though it's dog slow on pgFoundry I'm sure it > can be made fast on different hardware - I'm talking about > usability-wise. The TODO list currently has somewhere around 250 items, > and if you eventually want to bring in all bugs from pgsql-bugs etc.. Yup, PGFoundry is pretty glacial right now. It's getting put on new hardware, though, which should help. FWIW, RubyForge (another GForge instance) has some projects that have a fair number of items in the trackers: http://rubyforge.org/tracker/?atid=202&group_id=31&func=browse and it seems to hold up fairly well. But I agree wholeheartedly that it's better to tailor the tool to the dev process than vice versa. Yours, Tom
>> It does #7 now?! tell me how :D I've been longing for this! (I'd say >> that is one of the most important features!) > >If you go to, say, the support project bug tracker: > >http://pgfoundry.org/tracker/?atid=146&group_id=1000013&func=browse > >there's a dropdown list at the top of the page and a textfield to put a >search keyword in. This dropdown list is context-sensitive, so if >you're viewing a forum, the "bugs" option will go away and "this forum" >or something to that effect will appear. Oh, there it is! It was just kind of weird that it wasn't along witth the tracker, that looks more like a general search box :-) Now, if I could combine this with filtering, I would be completely satisfied :-) >> Oh, most definitly :-) But I think we'll need something >that's closer to >> "a perfect match". You need a tool that's adapted to the way the >> development works, not the other way around... Also, I'm >having a hard >> time seeing this one scale to the size of postgresql (not >technically - >> I'm sure it can do that, though it's dog slow on pgFoundry >I'm sure it >> can be made fast on different hardware - I'm talking about >> usability-wise. The TODO list currently has somewhere around >250 items, >> and if you eventually want to bring in all bugs from pgsql-bugs etc.. > >Yup, PGFoundry is pretty glacial right now. It's getting put on new >hardware, though, which should help. FWIW, RubyForge (another GForge >instance) has some projects that have a fair number of items in the >trackers: > >http://rubyforge.org/tracker/?atid=202&group_id=31&func=browse > >and it seems to hold up fairly well. Well, that's still only 37 on that page :) But a working search function takes care of a lot of the issues I've seen with large number of items on the lists. It just gets impossible to find. I even had that problem with pginstaller, which certainly doesn't have a lot of entries in its trackers. //Magnus //Magnus
On Mon, 2005-04-18 at 20:40 +0200, Magnus Hagander wrote: > Oh, there it is! It was just kind of weird that it wasn't along witth > the tracker, that looks more like a general search box :-) I agree, it is in a weird location... I don't know, my eyes just kind of skip over it. This comes up on the GForge forums occasionally, same sort of thing. > Well, that's still only 37 on that page :) Touche :-) > But a working search function takes care of a lot of the issues I've > seen with large number of items on the lists. It just gets impossible to > find. I even had that problem with pginstaller, which certainly doesn't > have a lot of entries in its trackers. Cool, Tom
On Monday 18 April 2005 14:52, Tom Copeland wrote: > On Mon, 2005-04-18 at 20:40 +0200, Magnus Hagander wrote: > > Oh, there it is! It was just kind of weird that it wasn't along witth > > the tracker, that looks more like a general search box :-) > > I agree, it is in a weird location... I don't know, my eyes just kind of > skip over it. This comes up on the GForge forums occasionally, same > sort of thing. > > > Well, that's still only 37 on that page :) > I think the largest use of a gforge-esque tracking system is probably sourceforge themselves: http://sourceforge.net/tracker/?group_id=1 -- Robert Treat Build A Brighter Lamp :: Linux Apache {middleware} PostgreSQL