> I think the two questions are largely independent. Having a clear mission for > what you get when you go to http://www.mozilla.org/ doesn't mean we need to > change the URLs of all the other content that lives on the site.
Therefore I'm splitting the discussion. This thread is for arguing about old content. I will post another thread about www.mozilla.org's mission statement. Follow-ups to m.d.mozilla-org.
For context, most comments after #20 (except for #29-34) are about which content to remove and what to do with it afterward. Here are some highlights:
5) How do I through old documentation in the bin? Looking at what we have in MLP and RDF, not all of that documentation is of historic value for the project, some of it is just deprecated and wrong now. If we could have project-wise 404 pages, throwing those away may be less painfull for broken bookmarks than anything. Setting up redirects for each and every page that doesn't speak the truth anymore is just going to keep us looking funny, and is may lead to confusing results in searches.
Policy should be a) If it can be updated to be useful 1. Migrate it to the right place (e.g. MDC) 2. Set up a 301 Moved Permanently redirect b) If it's a historical document (e.g. press release, slideshow for a past developer conference, release notes, status notes, announcement) 1. Move+redirect it to an archive, if it's not there already 2. Get it indexed appropriately c) If it's completely useless 1. cvs remove it 2. Set up a 410 Gone
And maybe we could customize the 410 message to point to a Web Archive copy of the missing document. (Epiphany does that automatically for failed domain lookups, iirc.)
There are very good reasons to believe that outdated (updatable or not; worth keeping or not; worth archiving or not) documentation represents right now over 6,000 webpages. My reasoning is that no one other than project manager (or project webpages maintainer) should address its own webpages under his project, under his directory. You can not and you should not expect a single person to address/deal with 6,000+ webpages. A single person having to address 100 or 200 webpages is a manageable workload/task.
Here's what I"m planning to do with the developer documentation on the mozilla.org web site:
If there's already equivalent content on MDC, delete it.
If there's not already equivalent content on MDC, and the content is reasonably accurate and current, migrate it.
If there's not already equivalent content on MDC, and the content is obsolete or so far out of date that updating it is more work than rewriting from scratch, delete it.
From talking to documentation contributors on two projects (rdf and l10n), I find three competing lines of thought:
- that's old bitrotten cruft, let's remove it - we need that documentation to unbitrot it or to resurrect the project - please don't purge my contribution to the mozilla project from the world
I probably ignored a few others, but it seemed to boil down to this.
As an example, http://www.mozilla.org/projects/l10n/mlp_status.html links to builds such as 1.8alpha4 or 1.3, or even 0.9. Not that we want anybody to use those builds, yet, if someone would want to contribute a firefox localization in that language, he or she could still reuse quite a bit of that work.
Most of the outdated stuff there has been rottening there for years and since nobody has taken any steps on resurrecting most of the outdated stuff, I don't think that people will suddenly step up and do it now that stuff is even more rotten than it was 2 years ago.
As for archival, that's why the cvs attic exists IMO. Anybody can resurrect stuff there if he wants to.
Just because something hasn't been modified for a while doesn't mean it's not useful.
As a user, I hate it when a page that I have bookmarked and use occasionally disappears.
That pages are no longer useful is a judgment call that needs to be made by people who know something about their content, not something that you can presume from lack of recent modification.
I also disagree that errors in a document due to underlying changes in the technology mean that the document is inherently obsolete. For example, it's useful to have design documents explaining why parts of Mozilla were designed the way they were -- they explain the rationale for many of the old decisions, which is useful to know -- even if some parts of the design have since been changed.
Understanding why something is designed the way it is is some of the most important documentation we have. It's easy to get so caught up in the disadvantages of a current design that we forget about the advantages and redesign them away. I don't think that should be out of the main flow of documentation.
majken@gmail writes:
This is where the audience has to be kept in mind. Keeping old documents somewhere that regular users are going to find them doesn't strike me as a good idea. ...
I'm not sure how many people would read an old document and be able to determine the decision path from it. Given, I don't know how the documents you're talking about are structured, but if they come right out and say "we made this decision because of x" then that's definitely something migratable. If the explanations are relevant to developing those tools/sections today then some sort of history should be included in, or near, the current documents.
I'm not sure what "archive a page" means -- but I'm ok with it as long as it means that: a) there are redirects so URLs don't break, and b) we can still edit the pages at least as easily (if not more) c) we can still redirect the archived URLs to pages on developer.mozilla.org or wiki.mozilla.org if we choose to migrate them That said, I'm not sure what the point of such archival would be.
~fantasai please don't attribute all the above to me when quoting, thanks~
I'm not sure obsoleting things that don't fit into the vision is completely orthogonal on the vision itself, IMHO it largely depends on what the vision is. Still a separate thread is good for it, the vision discussion may grow large enough by itself.
That proposal sounds quite good to me, but I'm not completely sure what falls in the "completely useless" category, I don't think I have met a lot of content there that I believe to fall in there.
> Just because something hasn't been modified for a while doesn't mean > it's not > useful.
Full ACK. I'm also with David on his other statements on this topic - we need to be quite careful there. I often found interesting things by digging through mutli-year-old historic documents, e.g. when I looked through the old docs we didn't remove from the projects/seamonkey dir (even if they don't belong to the current SeaMonkey project)...
> That proposal sounds quite good to me, but I'm not completely > sure what falls in the "completely useless" category, I don't > think I have met a lot of content there that I believe to fall > in there.
If that's the case then you probably haven't looked very thoroughly.
Most of this stuff is 6 years or older, some stuff even dates back to 1998 (the pre-XUL days). The only thing, that this obsolete stuff does, is that it diverts valuable time from various contributors, who have worked hard in the last 12-18 months to make the HTML code of these pages valid HTML.
All their valuable time could have been channeled to much more useful things.
>> Just because something hasn't been modified for a while doesn't >> mean it's not useful.
> Full ACK.
Let me say it this way: It does not mean that it's not useful in *all* cases. It's only not useful in 99% of all cases.
> I'm also with David on his other statements on this topic - we > need to be quite careful there. I often found interesting things > by digging through mutli-year-old historic documents, e.g. when > I looked through the old docs we didn't remove from the > projects/seamonkey dir (even if they don't belong to the current > SeaMonkey project)...
The SeaMonkey project pages are well-owned. If you want to retain historic documents that is of course your choice.
I think we should only deal with stuff at first, that isn't well-owned anymore or never has been.
I think this archiving process mentioned in fantasai's earlier post makes a lot of sense:
Policy should be a) If it can be updated to be useful 1. Migrate it to the right place (e.g. MDC) 2. Set up a 301 Moved Permanently redirect b) If it's a historical document (e.g. press release, slideshow for a past developer conference, release notes, status notes, announcement) 1. Move+redirect it to an archive, if it's not there already 2. Get it indexed appropriately c) If it's completely useless 1. cvs remove it 2. Set up a 410 Gone
I also like the Mozilla history page that is attached to bug 345664:
Having a curated History section or History site would make this sort of information much easier to find than if we removed all prominent links to obsolete pages just to get them out of the flow of the main information. Since the 10th anniversary of the www.mozilla.org site launch is coming up, maybe it would make sense to tie a History archive in with that anniversary?
My earlier comment about archiving in bug 345664 was mainly meant to address the issue about how to avoid getting into a debate over the fate of every page on the site. A clear mission for www.mozilla.org will help us figure out what should be archived and what shouldn't (I think splitting the discussion on the bug into an archive thread and a vision thread has made this point more clear). The next step would be to find people to do that archiving work.
Lastly, I think the point of bothering with archiving is that the www.mozilla.org site has 10 years worth of content on it and I would be surprised if every page matches up with the vision (whatever that vision ends up being). We would need to work out the mechanics of how archiving would work, but we should be able to do it without breaking URLS and still allow people to edit the content.
Simon Paquet wrote: > Most of this stuff is 6 years or older, some stuff > even dates back to 1998 (the pre-XUL days). The only thing, that > this obsolete stuff does, is that it diverts valuable time from > various contributors, who have worked hard in the last 12-18 months > to make the HTML code of these pages valid HTML.
Doing that work at all has been a futile effort anyways, IMHO (and caused some needless stepping on toes of people in addition). Oh, and it made the pages look quite up-to-date as they all were updated in 2006 or such now - so the content must be current, right? :P
A page that has not been touched in 6 years doe not mean at all that it's useless. Ideas from vixen might still be interesting for Mozpad people thinking about a Mozilla IDE now, some color management support just went in, not sure how it realtes to colorsync, etc.
Moving this content to a archive or obsolete area/domain might be a good idea, but purging our heritage of old ideas and reinventing those wheels once again in the future is not the best option IMHO.
Updating it with any changes, even for standard compliance is and was a futile and useless effort (in German I'd call it "vergebene Liebesmüh"), no matter where it lives, esp. in times where noone know what should be done with www.m.o at all.
> I think we should only deal with stuff at first, that isn't > well-owned anymore or never has been.
You claim that vixen or anya were never (well-)owned? Pav and Ben might disagree heavily with that.
>> Most of this stuff is 6 years or older, some stuff even dates >> back to 1998 (the pre-XUL days). The only thing, that this >> obsolete stuff does, is that it diverts valuable time from >> various contributors, who have worked hard in the last 12-18 >> months to make the HTML code of these pages valid HTML.
> Doing that work at all has been a futile effort anyways, IMHO > (and caused some needless stepping on toes of people in addition).
Exactly.
> Oh, and it made the pages look quite up-to-date as they all were > updated in 2006 or such now - so the content must be current, > right? :P
Yes, it takes some digging to find out what the state of that content *really* is like.
> A page that has not been touched in 6 years does not mean at all > that it's useless.
In at least 99% of all cases it does. In the Internet where 6 months is already a very long time, technical information that hasn't been touched that long is like giving physics student a book from 1850.
> Ideas from vixen might still be interesting for Mozpad people > thinking about a Mozilla IDE now, some color management support > just went in, not sure how it relates to colorsync, etc.
There's a difference between what *might* be interesting and what *is* interesting. If that information hasn't been touched or been reused in such a long time, then I doubt that it is still of real value.
And even if it is, CVS has this thing called Attic where all the information that was removed from a repository still resides and can be ressurected if the need arises.
What I'm essentially prposing is as follows: - we should remove all outdated information as long nobody steps up with a valid reason to keep it and then contributes to updating it - such a removal notice should be posted to all relevant newsgroups and the Mozilla Web Development Blog - if nobody steps up within four weeks, the content will be removed - if it turns out that the removal of some information was an error a bug should be filed and the information can be resurrected from the CVS Attic - it might be a good idea to create www-archive.mozilla.org with all the content of www.mozilla.org as of a certain point in time (say the 1st of September or the 1st of October). After the creation of that archive, the actual removal of outdated content on www.mozilla.org can proceed
>> I think we should only deal with stuff at first, that isn't >> well-owned anymore or never has been.
> You claim that vixen or anya were never (well-)owned? Pav and Ben > might disagree heavily with that.
They aren't well-owned now. I'm not sure if they were ever well-owned. I have only the (rather poor) project pages as a guide for such a
Robert Kaiser wrote: > A page that has not been touched in 6 years doe not mean at all that > it's useless. Ideas from vixen might still be interesting for Mozpad > people thinking about a Mozilla IDE now, some color management support > just went in, not sure how it realtes to colorsync, etc.
I don't think our criterion should be "it might be useful to somebody". Old and inaccurate pages have a negative impact on our total documentation effort: they pollute search results and confuse new members of the community.
I think the criterion should be "would migrating this content to MDC help the project overall"? If not, we should spend as little time on it as possible and just remove it from CVS outright.
Simon Paquet wrote: > Robert Kaiser wrote on 26. Aug 2007: >> A page that has not been touched in 6 years does not mean at all >> that it's useless.
> In at least 99% of all cases it does. In the Internet where 6 months > is already a very long time, technical information that hasn't been > touched that long is like giving physics student a book from 1850.
Oh, that's the reason why all the old ideas come up again and are tried again every couple of years by reinventing the wheels, just because we're in a technology sector that thinks everything older than 6 months is prehistoric? I guess we should stop producing browsers because they were invented more than 10 years ago so they must be long obsolete.
The comparison with physics is bogus, because scientists often refer back to the works of older studies and would never throw away a book even if is was 150 years old, just because it's amazing how well the old theories still fit as (generalized or special) cases of the old theories and the old experiments still need to hold true with current theories or the new theory is bogus (or one has to do a new in-deep study to find out if the old experiment might have been flawed). Physics tries to describes natural laws that are true time-independently, so old information can never be thrown away but needs to be kept for reference indefinitely.
>> Ideas from vixen might still be interesting for Mozpad people thinking >> about a Mozilla IDE now, some color management support just went in, >> not sure how it relates to colorsync, etc.
> There's a difference between what *might* be interesting and what *is* > interesting. If that information hasn't been touched or been reused in > such a long time, then I doubt that it is still of real value.
Oh, so you know who has read the documents and used their contents for their own ideas? Interesting. Just because someone isn't interesting for you does not mean it "is not interesting".
> And even if it is, CVS has this thing called Attic where
...nobody looking for information ever looks.
So, I'm all for cleaning up www.mozilla.org itself, but not without keeping the old information around as real HTML (not something where one needs to jump through lots of bonsai hoops and Attic to see some content that once may have been somewhere) in some place. This place can very well be a archive site or so, that's OK, this need (and probably should) not be www.m.o itself, but there should be a good way to find such pages in that archive with a search.
There are cases where you want to find such things because you know there has been something or want to know if it has been there. E.g. when thinking about menu structures in the browser, one might want to find the UI doc we had on doing previous big restructurings and reasons for how it has been done the way it has been. I often find myself digging around in old Bugzilla reports for such info, and I was digging through old www.m.o content just recently to find if we have design guideline docs for the Modern theme. I guess other developers run into similar issues.
>>> A page that has not been touched in 6 years does not mean at all >>> that it's useless.
>> In at least 99% of all cases it does. In the Internet where 6 months >> is already a very long time, technical information that hasn't been >> touched that long is like giving physics student a book from 1850.
> Oh, that's the reason why all the old ideas come up again and are > tried again every couple of years by reinventing the wheels, just > because we're in a technology sector that thinks everything older > than 6 months is prehistoric?
No. The reason for that is mainly that people often do not read documentation and the fact that innovation in some sectors makes people try things again that did not work years ago.
None of that really relates to the current question.
> I guess we should stop producing browsers because they were > invented more than 10 years ago so they must be long obsolete.
Your attempt at sarcasm could definitely be improved :)
The reason that browsers are still relevant and aren't obsolete is exactly that constant innovation has happened in that space and Netscape did not finish its efforts with their 1.0 release.
Browsers became more and more irrelevant, when Microsoft (as the amrket leader) stopped innovation after IE6 and it needed the success of Firefox (and Safari and Opera) to bring Microsoft back into the game.
> The comparison with physics is bogus, because scientists often > refer back to the works of older studies and would never throw > away a book even if is was 150 years old
I doubt that a physics student would be given a book for learning without all the stuff in it that Einstein, Planck, Heisenberg and others researched after 1850.
The same goes for mozilla.org. Outdated content pollutes search results and gives people false hopes, when they use outdated sample code or try to contribute to a project which has long been stalled or been forgotten by its former owners.
>>> Ideas from vixen might still be interesting for Mozpad people >>> thinking about a Mozilla IDE now, some color management support >>> just went in, not sure how it relates to colorsync, etc.
>> There's a difference between what *might* be interesting and what >> *is* interesting. If that information hasn't been touched or been >> reused in such a long time, then I doubt that it is still of real >> value.
> Oh, so you know who has read the documents and used their contents > for their own ideas? Interesting. Just because someone isn't > interesting for you does not mean it "is not interesting".
Have you actually looked at the content on the samples that I mentioned? There's really nothing there, that one can use as a source for new ideas, new developments, etc.
>> And even if it is, CVS has this thing called Attic where
> ....nobody looking for information ever looks.
I can't say that I'm sorry for the two people (at a maximum) on this planet, that this would apply to.
> So, I'm all for cleaning up www.mozilla.org itself, but not without > keeping the old information around as real HTML (not something > where one needs to jump through lots of bonsai hoops and Attic to > see some content that once may have been somewhere) in some place.
It really comes down to the question of what this whole effort should achieve?
Do you want www.mozilla.org to be relevant again, with content that is current and well-owned? If yes, you will have to make it easy for people to reach that goal. It won't be perfect. It may be ugly in some cases, but in the end something will happen.
Or do you see the preservation of all content on www.mozilla.org as a pre-requisite to all other actions? That is surely the way for a perfect solution, it just might take you 10 years to reach your goal, which you will most likely never reach.
> This place can very well be a archive site or so, that's OK, this > need (and probably should) not be www.m.o itself, but there should > be a good way to find such pages in that archive with a search.
> There are cases where you want to find such things because [...]
I don't doubt that you and other can come up with lots of cases, where one might take a look at all the old stuff or has done so in the past.
My question is: How relevant is this use-case? How often does or did this happen for how many people? How relevant are those people to the community?
I don't think that 5-10 people that rarely use this information should be able to forestall this effort. If it turns out that this is a common use-case for about a hundred people (many of those valuable contributors to the community), then this is of course a different matter.
I seriously doubt that this is the case, though. I even doubt that the first case is true, but I might be mistaken, of course.
Cya Simon
PS: I have said it before, but want to say it again. My plan is, not to touch any content in www.mozilla.org/projects/*.* that resides in a project space that is well-owned. So if you have some old documentation lying around under projects/seamonkey, that you want to keep for whatever reasons, I will not touch it (ever) unless you ask me to.
>> A page that has not been touched in 6 years doe not mean at all >> that it's useless. Ideas from vixen might still be interesting >> for Mozpad people thinking about a Mozilla IDE now, some color >> management support just went in, not sure how it realtes to >> colorsync, etc.
> I don't think our criterion should be "it might be useful to > somebody". Old and inaccurate pages have a negative impact on > our total documentation effort: they pollute search results and > confuse new members of the community.
Exactly.
> I think the criterion should be "would migrating this content to > MDC help the project overall"? If not, we should spend as little > time on it as possible and just remove it from CVS outright.
Exactly.
Just FYI, my list in <1187896069...@user.newsoffice.de> already lists some content which might be valuable to migrate. The following projects "might" still have content that Sheppy and others could migrate to MDC:
Simon Paquet wrote: > PS: I have said it before, but want to say it again. My plan is, not > to touch any content in www.mozilla.org/projects/*.* that resides > in a project space that is well-owned. So if you have some old > documentation lying around under projects/seamonkey, that you want > to keep for whatever reasons, I will not touch it (ever) unless > you ask me to.
All documents in that space that I haven't edited are historic and should be subject to the same treatment as other historic content. There are a few docs there about how to prepare releases, which still apply to a certain extent, there are some guidelines there that are an interesting read and parts of which never lose a state of being current - but they are all barely linked, barely (if at all) owned content and as said above should be subject to the same process as they pollute the current SeaMonkey project's web space (even though I'm working hard on moving that away from www.m.o again). I would want to have those docs thrown away completely though, as I don't think they're worthless - they just belong to a (searchable) archive.
> > This place can very well be a archive site or so, that's OK, this > > need (and probably should) not be www.m.o itself, but there should > > be a good way to find such pages in that archive with a search.
I am happy to lead or help with the archiving process (if anyone else wants to lead or help with this, let me know or add yourself to the mozilla.org planning wiki page). I wanted to see what happened with this discussion before updating that planning document with an owner and any action items though.
To see if we have a consensus, let me make this proposal based on the conversation so far:
- First, we take a snapshot of the site and make it available somewhere (archive.mozilla.org or wherever). At some point, we may put together a curated history page or section that makes the historically relevant information easier to find on the archive site. - Second, we post a page on the wiki where we list all of the pages currently under consideration for archiving. To make things easy to manage, we should start by posting only a small batch of pages (5 or 10?). Once we have the process working more smoothly we can post larger batches of pages. - Third, after a certain period of public discussion (2 weeks maybe?) these pages will be removed from www.mozilla.org's CVS and a redirect set up to the archive site. We may need to set up some sort of archive landing page or header so that people are aware that the content they are seeing is no longer considered current.
Does that sound reasonable to everyone? If not, please post suggestions for changes to this process. It goes without saying that we can't get started with this until the vision is sorted out, but I think we can at least agree on the archiving process now.
> It really comes down to the question of what this whole effort should > achieve?
> Do you want www.mozilla.org to be relevant again, with content that > is current and well-owned? If yes, you will have to make it easy for > people to reach that goal. It won't be perfect. It may be ugly in > some cases, but in the end something will happen.
For me, it's about having a clear definition of the purpose of each site. When mozilla.com was created, it was clear to me, that www.mozilla.com was for the commercial content, more specifically, the user-targeted content, and www.mozilla.org was about the open source project.
I'm having much more trouble knowing the difference in purpose, between www.mozilla.org and MDC, and especially wiki.mozilla.org.
If a person wants to search for content, they should know where to find it. If a person wants to contribute content, they should where to put it. -- Chris Ilias <http://ilias.ca> List-owner: support-firefox, support-thunderbird, test-multimedia
Chris Ilias wrote: > I'm having much more trouble knowing the difference in purpose, between > www.mozilla.org and MDC, and especially wiki.mozilla.org.
> If a person wants to search for content, they should know where to find > it. If a person wants to contribute content, they should where to put it.
I think the link descriptions on the future www.m.o site should make this clear (yes, I imply that www.mozilla.org will prominently link them, but from what I hear that's a pretty accepted target).
>If a person wants to search for content, they should know where to find >it. If a person wants to contribute content, they should where to put it.
If a person wants to search for content he should use Google or another search engine of his choice. That may not be the answer that you want to hear, but search engine do a much better job on finding the stuff that you're looking for than any content reorganization that I or other can conceive.
Our first goal should be to weed out all the outdated and obsolete content. Once that's done, we can think about better content presentation mechanisms.
For the proposed archiving process mentioned earlier in this thread, we have one vote of sounds reasonable and no one objecting. Many people were away on vacation last week though, so it might take some more time to get comments. If there aren't any objections by the end of the week, I'll update the planning page on the wiki and open an archiving bug to get things moving.
|
