What's the plan for fixing MDC?
Peter Weilbacher
after the conversion (posting "Can't figure out the new MDC",
MID bcCdneKvwYxOVV7V...@mozilla.org).
I was just looking again through that list, after trying to figure out the change
history for a few pages. All the problems are still unresolved. So, what's the
plan on finally getting that stuff fixed? Are there bug reports that one can track?
After trying to use it for the past few months, I still have the feeling that you
exchanged a nicely working system that attracted at least those contributors
that were familiar with the Wikipedia editing style with something that is only
half baked and was (is?) unfamiliar to all people who have contributed before.
Do you at least have stats that show that now many more people are
contributing something? Or contributed something important, i.e. more than
typo changes? (I remember that this was the original reason why the change
was done.)
Peter.
Brett Zamir
the new system.
I also would suspect also that the stats would bear out that this just
isn't as popular for significant editing as the old site.
Can't Mozilla provide some kind of democratic processes (not expecting
or wanting a pure democracy, mind you), to make some kind of poll among
those of us who are chipping in our free time to help develop this site?
Mozilla takes surveys (which I regrettably missed last time), and seems
at some level at least to be interested in adapting itself, so I am
hopeful that if this new software can be gauged to indeed be unpopular,
that adjustments could be made to return to the old one.
Mediawiki can be adapted. There are already extensions for it which do a
number of things not present at Wikipedia, and no doubt programmers
could be paid to build any further extensions needed (maybe even helping
Mediawiki in the process). The Dekiwiki software does have a few
advantages, but they really seem dwarfed by all of the problems.
While I think we should ultimately abide by and support the decision
that is made, and not continue wrangling over it, but I also feel that
consultation with all involved should be a part of the process that
leads to that being possible. If this were a decision that the majority
of contributors supported (or if we could know that to be the case),
then I think it would put to rest such grumblings as some of us have.
Brett
> _______________________________________________
> dev-mdc mailing list
> dev...@lists.mozilla.org
> https://lists.mozilla.org/listinfo/dev-mdc
>
>
John J. Barton
> I strongly agree with this, even though I've been trying to make do with
> the new system.
+1
Sheppy
> I posted here in September, i.e. more than 3 months ago, with a list of bugs
> after the conversion (posting "Can't figure out the new MDC",
> I was just looking again through that list, after trying to figure out the change
> history for a few pages. All the problems are still unresolved. So, what's the
> plan on finally getting that stuff fixed? Are there bug reports that one can track?
The next release of MindTouch Deki, slated for Q1, is going to address
a large chunk of our outstanding issues. You can find my list of top-
priority issues that I'm pushing them to fix, and what the status of
those items is, here:
http://developer.mozilla.org/User:Sheppy/MindTouch_Deki_bug_priorities
You can see the plans and design documentation for the next release of
Deki (called Lyons) here:
http://wiki.developer.mindtouch.com/MindTouch_Deki/Release/Lyons
Eric Shepherd
Developer Documentation Lead
Mozilla Corporation
http://www.bitstampede.com/
Peter Weilbacher
"Q1" could still be a long way off, but thanks for letting me (us) know. The list
of planned improvements looks good. :-)
Peter.
Jean-Marc Desperrier
> [...]
> The next release of MindTouch Deki, slated for Q1, is going to address
Back in mid-september, my experience with the switch to Deki was that it
was simply a "Miserable Failure" (c).
I'll try to find the time to share my frustrations at trying to do some
simple editing of the "Using_the_Mozilla_symbol_server" page, but it was
*awfully* slow (like waiting 10 minute for a page to load), and I just
couldn't make sense of the system to insert images. I ended up reading
the doc, going to the MinTouch site to try to find it, and still not
really understanding the behaviour I was getting.
I regret there was no revolt against this new system.
Axel Hecht
Bugs on the list of things that make me throw stuff against the wall
instead of editing docs:
- two competing vertical scrollbars, both wrong.
- weirdest scrolling behaviour on trying to edit an element
- link editor is useless to me, and source-editing links is cumbersome, too
I don't recall those being that awkward initially, the link editor aside.
Axel
Brett Zamir
have yet to receive any response...
1) Where is the dump of the site contents? Any site which is transparent
and is serious about wanting to engage the community is not going to
hoard things in closed-source-business style, the least of all, I would
hope, Mozilla.
2) How many people would (or will) it take to get a change in course of
action here? Is there any democratic aspect to community involvement or
feedback here? Seriously! There have been plenty of unhappy people here
(actually, only unhappy people from all the comments I've read over
time), and short of anything really ugly, I think there has been a
"revolt" with people expressing unhappiness with this from the
beginning. No doubt there has been some hard work to fix things.
The emperor has no clothes, but I just don't know who else here has the
ear of the emperor to tell him (unless the emperor himself is just
content to be naked). I also can't help wonder if there's something
else going on with this...
Brett
Sheppy
> I've asked two basic questions which I think are important for this and
> have yet to receive any response...
>
> 1) Where is the dump of the site contents? Any site which is transparent
> and is serious about wanting to engage the community is not going to
> hoard things in closed-source-business style, the least of all, I would
> hope, Mozilla.
Do you mean the old site, or the current one? I'm not sure what
you're asking for here. The old site is actually still available if
you change your /etc/hosts file (or equivalent) to direct
developer.mozilla.org to the old IP address, but we're not providing
this globally because we don't want to encourage people to read now
out of date content. However, we can provide it to people that want
to check the current site against the old one for integrity and the
like.
> 2) How many people would (or will) it take to get a change in course of
> action here? Is there any democratic aspect to community involvement or
> feedback here? Seriously! There have been plenty of unhappy people here
> (actually, only unhappy people from all the comments I've read over
> time), and short of anything really ugly, I think there has been a
> "revolt" with people expressing unhappiness with this from the
> beginning. No doubt there has been some hard work to fix things.
What do you want to see happen here? We're working like crazy to fix
things up. We went out of our way to solicit feedback before the
transition, and nobody had any significant complaints. Once we
committed to the change, going back is practically impossible (unless
someone wants to volunteer to copy all the changes back into
MediaWiki), which is why we gave the community many, many months to
comment on not only the proposal but on test versions of the redone
web site.
The fact that nobody complained until it was too late to reverse
course is very frustrating for me, as the input of the community is
extremely important to me.
All that aside, at this point, I think we're closer to fixing up the
current site than we are to being able to revert back to MediaWiki, so
I think our best bet is to push forward and get things done. I
continue to ask for your patience while we get this done, please.
Sheppy
see what's going on here:
http://www.bitstampede.com/2009/01/07/mdc-stuff-and-notes/
John J. Barton
> The fact that nobody complained until it was too late to reverse
> course is very frustrating for me, as the input of the community is
> extremely important to me.
>
It seemed to me that the conversion from MediaWiki was a done deal, so I
complained a lot about the new version but not about the conversion
since there was no point. Blaming us is not a great way to encourage
community.
I don't think any one was asking for a non-markup input UI. (and the
current tool is worse because you have to use the HTML view to get
anything done). I think the community was asking for more and better
docs not a new editor.
But it's done. Let's try to fix it.
jjb
John J. Barton
Maybe there is something different between your browser and the one Deki
devs use?
Firefox version? extensions? Do you have Firebug installed?
I think all of the pages need to be changed to make the edit sections
much smaller. Deki seems to have real problems with large edits.
(I'm not make excuses, just trying to help given the current state of
affairs).
jjb
Robert Kaiser
> Is there any democratic aspect to community involvement or
> feedback here?
No actually working open source community works really democratically,
most that really do work are benevolent dictatorships (like Linux or
Wine) or meritocracies (like Mozilla).
That said, from what I know, the MDC team tries to address feedback they
get as far as possible - it's just not possible in a timely fashion
everywhere.
Robert Kaiser
Ran Ari-Gur
issues" that would supposedly[1] be fixed either before the
conversion, or soon after. There was no real stable version or code
freeze before the conversion, so there was no opportunity to complain
about what we have now, until we actually had it. (Also, I for one
never felt like it was asked "should we convert?" or even "should we
convert soon, or should we wait till more bugs are fixed?"; it always
seemed more like "we're converting, send me input." Which really
limits the range of input you can expect to get. But, maybe those
questions were asked before I joined the list.)
But this sort of conversion is always messy and painful, and I don't
think this specific one was too bad as these things go.
Once the basic functionalities are fixed, I really look forward to the
new features MindTouch will create for us.
-Ran
[1] In all fairness, I think you and MindTouch were pretty up-front
about which bugs were pre-conversion, which somewhat-post, and which
long-post, but it was hard for us -- or at least for me, I don't speak
for anyone else -- to keep them all straight. And some of the bugs
that were fixed *right* before conversion were serious blockers that
discouraged people from testing things out and discovering other
problems until after conversion.
On Thu, Jan 8, 2009 at 11:23 AM, Sheppy <the.s...@gmail.com> wrote:
> [...]
>
> What do you want to see happen here? We're working like crazy to fix
> things up. We went out of our way to solicit feedback before the
> transition, and nobody had any significant complaints. Once we
> committed to the change, going back is practically impossible (unless
> someone wants to volunteer to copy all the changes back into
> MediaWiki), which is why we gave the community many, many months to
> comment on not only the proposal but on test versions of the redone
> web site.
>
> The fact that nobody complained until it was too late to reverse
> course is very frustrating for me, as the input of the community is
> extremely important to me.
>
> [...]
>
> Eric Shepherd
> Developer Documentation Lead
> Mozilla Corporation
> http://www.bitstampede.com/
Sheppy
just before launch. In retrospect, waiting a little longer before
throwing the switch would have been a good idea, but it didn't seem at
the time like it was necessary.
As for slow load times, we think it's because since switching to SSL
for the site, nothing is being cached locally by your browser
anymore. We're working on addressing that by adding the necessary
cache control lines. I don't know when that will be done, but it's in
the pipeline and should help.
I personally was not interested in switching us to use SSL but enough
people wanted it that we went for it anyway. See, democracy in
action! :)
Peter Weilbacher
> There have been plenty of unhappy people here (actually, only unhappy
> people from all the comments I've read over time)
That's normal for newsgroups and mailing lists. Everybody who is happy
doesn't have a reason to post. I would actually say that the traffic
can only be so low in this group if most people are happy. ;-)
Peter.
P.S.: Sheppy, thanks for the update in your blog!
Sheppy
our needs. I've been chatting with them today about their work on the
diff and history features; they've definitely made some good progress
and I'm nudging them to do a few more things to it before they ship
their next release.
Also, it looks like some of the issues we have are caused by setup
problems with our install. I've already mentioned the slowness and it
probably being caching related. There appears to be a configuration
problem causing the double scrollbar and jumpiness in the editor.
MindTouch is investigating what we might have set up wrong that's
causing that, and we'll fix it as soon as we can.
John J Barton
> On 08.01.09 15:50, Brett Zamir wrote:
>> There have been plenty of unhappy people here (actually, only unhappy
>> people from all the comments I've read over time)
>
> That's normal for newsgroups and mailing lists. Everybody who is happy
> doesn't have a reason to post. I would actually say that the traffic
> can only be so low in this group if most people are happy. ;-)
I agree with this idea, so how many people contribute content to MDC?
Let's say there are 5 unhappy people. Are their 500 contributors? 50? or 6?
To Eric's credit I think part of the reason for Deki was the idea that
more folks would contribute if the UI was less geeky. Sure that irks us
geeks, but could be a win in the long run.
jjb
Benoit Leseul
> We went out of our way to solicit feedback before the
> transition, and nobody had any significant complaints. Once we
> committed to the change, going back is practically impossible (unless
> someone wants to volunteer to copy all the changes back into
> MediaWiki), which is why we gave the community many, many months to
> comment on not only the proposal but on test versions of the redone
> web site.
Sorry but I have to "respectfully disagree".
The problems we are seeing were exactly the ones we were fearing when
the switch was announced: loss of the wiki syntax, disruption of our
habits, availability of statistical tools, unusable diffs, slowness.
The very fact that going back is impossible was one of our main
concerns. I remember being ridiculed when I said the pseudo-XHTML
syntax was awful compared to the MediaWiki one because "XHTML is the
future" and "you will be using wysiwyg anyway". Both are so untrue.
The net result is that the French MDC is not maintained anymore. All
our contributors just left one after another.
I can be mistaken but looking at the "Recent changes", I think the
situation is more or less the same for the other localized versions. I
don't see anything like the healthy competition we had one year ago
when we were just trying to grow faster than Japanese or Polish with
dozens of edits per week for each locale.
To end on a more positive note, I will try to restart our effort once
more after the announced update which seems to fix at least some of
the worse showstoppers. I hope we will just be one year late.
--
Benoit
Sheppy
Yes, that was the thinking. I just hope you guys can keep trying to
be patient while we work out the issues that bug you the most. Your
input is definitely being heard -- it just takes a while to get all
this stuff done. :)
L. David Baron
> about what we have now, until we actually had it. (Also, I for one
> never felt like it was asked "should we convert?" or even "should we
> convert soon, or should we wait till more bugs are fixed?"; it always
> seemed more like "we're converting, send me input." Which really
As an occasional observer, the posts I saw were mostly of the form:
Here's another import; please have a look at pages you care
about and see that they were imported correctly. But don't
edit, since edits will get lost.
I did in fact check that a bunch of pages I cared about were
imported correctly.
But I actually didn't know that the editing mechanism wasn't
wikitext anymore until after the conversion happened and I tried
editing a page for the first time in the new system.
Since the conversion I think I've been editing MDC pages much less,
I think since I'm much less comfortable editing things when I feel
like I can't tell what I've changed.
-David
--
L. David Baron http://dbaron.org/
Mozilla Corporation http://www.mozilla.com/
Brett Zamir
> On Jan 8, 9:50 am, Brett Zamir <bret...@yahoo.com> wrote:
>
>
>
>> I've asked two basic questions which I think are important for this and
>> have yet to receive any response...
>>
>> 1) Where is the dump of the site contents? Any site which is transparent
>> and is serious about wanting to engage the community is not going to
>> hoard things in closed-source-business style, the least of all, I would
>> hope, Mozilla.
>>
>
> Do you mean the old site, or the current one?
> I'm not sure what
> you're asking for here. The old site is actually still available if
> you change your /etc/hosts file (or equivalent) to direct
> developer.mozilla.org to the old IP address, but we're not providing
> this globally because we don't want to encourage people to read now
> out of date content. However, we can provide it to people that want
> to check the current site against the old one for integrity and the
> like.
>
>> 2) How many people would (or will) it take to get a change in course of
>> action here? Is there any democratic aspect to community involvement or
>> feedback here? Seriously! There have been plenty of unhappy people here
>> (actually, only unhappy people from all the comments I've read over
>> time), and short of anything really ugly, I think there has been a
>> "revolt" with people expressing unhappiness with this from the
>> beginning. No doubt there has been some hard work to fix things.
>>
>
> What do you want to see happen here? We're working like crazy to fix
> things up. We went out of our way to solicit feedback before the
> transition, and nobody had any significant complaints. Once we
> committed to the change, going back is practically impossible (unless
> someone wants to volunteer to copy all the changes back into
> MediaWiki)
I would personally be most happy to work with others in doing so. I
wouldn't even hesitate to completely lose all of my edits since the
changeover (which came at some cost to my wrists) if that would be
necessary.
The fact that there even is a problem to revert/convert the current
site's contents gives I think some serious reason for concern and even
more reason to revert back now rather than later (unless XHTML importers
can be made by relying on a base URL perhaps, or use something like
this:
http://search.cpan.org/~diberri/HTML-WikiConverter-0.55/lib/HTML/WikiConverter.pm
). Mediawiki is not going anywhere, maybe...like...ever, and there are
different kinds of converters for and to it. I'd also hope it would be
possible for the licenses (as well as pseudo-code) to be made compatible
somehow, e.g., if someone wanted to spin off and make a wiki book.
Mediawiki has recently made some license changes which might make that
possible.
Admittedly, Mediawiki received a large grant recently to make the site
more user-friendly to editing
(http://wikimediafoundation.org/wiki/Press_releases/Wikipedia_to_become_more_user-friendly_for_new_volunteer_writers
). Of course they are not a /developer/ center, but I agree it's still a
good aim to try to give extra ways of editing--with an emphasis on
"extra". That brings up another advantage with going back to
Mediawiki--it's guaranteed to continue evolving. And again, there are
already plugins such as FCKeditor which can give a WYSIWYG option to
Mediawiki. There's a demo here which produces Mediawiki syntax directly:
http://mediawiki.fckeditor.net/index.php/Sandbox .
I think the above reasons alone would make it very worthwhile to go back
to Mediawiki, despite the work for those of us involved. It's already
limiting productivity, so getting it over with should, I feel, pay off
in the long run (and even not so long run).
Brett
Brett Zamir
> Brett Zamir wrote:
>> Is there any democratic aspect to community involvement or
>> feedback here?
>
> No actually working open source community works really democratically,
> most that really do work are benevolent dictatorships (like Linux or
> Wine) or meritocracies (like Mozilla).
here before)--that would indeed be a nightmare, just as it is in the
U.S. where actors are disproportionately elected to high office,
uneducated juries make life-turning decisions, and mass media and polls
disproportionately affect policy which should be determined by competent
experts. I'm just suggesting to have at least a democratic aspect in
which people are consulted before major changes that affect them are
made (not to mention those involved in making regular contributions). If
Linus decided to do something like replace their version control with a
wiki because it might invite more newcomers, there'd certainly be a
revolt there! (Given the manners of some Linux die-hards I've met, that
could be really scary...)
Actually, some kind of elected community board like Mediawiki has might
not be a bad idea (and give a venting outlet too).
Brett
Brett Zamir
> On 08.01.09 15:50, Brett Zamir wrote:
>> There have been plenty of unhappy people here (actually, only unhappy
>> people from all the comments I've read over time)
>
> That's normal for newsgroups and mailing lists. Everybody who is happy
> doesn't have a reason to post. I would actually say that the traffic
> can only be so low in this group if most people are happy. ;-)
>
changes, and now people complaining shows that people do care about what
happens here... :) You also have to account for people dropping off
silently, as Benoit suggested (something which happens when people feel
they don't have a say in the matter, as John J. and Ran also suggested).
At least two other people I ran into were motivated to start and build
their own wiki due to the lack of openness here (in license and
decision-making). And if people were happy, I think they also might be
taking the time to say so now...
I'm not going on about this to be counter-productive, but rather in
hopes that both the means of consultation is improved and perhaps more
formalized for the future, and that this current issue can be fully
discussed, rather than swallowing a consensus that doesn't seem to be a
consensus... If this is just inevitably going to be a dictatorship here,
I want the old dump to start a new (sustainable) version with anyone
else here who's interested... But I hope it doesn't need to come to that...
Brett
Sheppy
> Sheppy wrote:
> I would personally be most happy to work with others in doing so. I
> wouldn't even hesitate to completely lose all of my edits since the
> changeover (which came at some cost to my wrists) if that would be
> necessary.
Given that all the work on Firefox 3.1 documentation has happened
since the transition, I'd sooner shoot myself in the head than lose
those changes and have to plug them all back in again by hand. :)
> I think the above reasons alone would make it very worthwhile to go back
> to Mediawiki, despite the work for those of us involved. It's already
> limiting productivity, so getting it over with should, I feel, pay off
> in the long run (and even not so long run).
I disagree that it's limiting productivity. I'm easily among the
biggest (in terms of sheer volume) contributors to the site at present
(not that this is some kind of contest, I'm just saying that as a
component to this argument), and I find the current environment vastly
easier to work in than MediaWiki. Aside from a few annoyances, some
of them admittedly significant -- but in the process of being fixed --
I'm much happier now than I was in MediaWiki.
That being said, it's certainly not only my opinion that matters here.
What I'd really like to propose is that we table any discussion about
reverting to MediaWiki and give Deki's next release a solid try when
it comes out next month. Give ourselves through the spring to adapt
to it and to do the customizations we might like to attempt, then re-
evaluate things in the summer. I think most people will find
themselves much happier with the Deki system once we get that stuff
done, and we're close enough to it happening that IMHO it makes sense
to try it instead of going through all the work of reverting right
now.
I don't personally want to switch back, because, as I said, I find
myself much more productive now than I was before. And I still get
the distinct impression that only a fairly small percentage of our
users are so frustrated with the current site that they seriously want
to look at switching back to MediaWiki.
It saddens me to think there are people who seem to feel that this is
some sort of authoritarian setup, because it really, really isn't. I
recognize that folks have their reasons for not having offered
feedback before the final decision to switch to Deki was made, but
that doesn't change the fact that feedback was solicited time and
again. I'd really appreciate it if folks could stop questioning my
motives here. I'm really just trying to make sure the site moves
forward in the best way for the entire community (since it's my job to
do that).
John J. Barton
>
> I don't personally want to switch back, because, as I said, I find
> myself much more productive now than I was before. And I still get
You are have a completely different experience that at least I am. There
has to be some explanation.
> the distinct impression that only a fairly small percentage of our
> users are so frustrated with the current site that they seriously want
> to look at switching back to MediaWiki.
>
But the question is are the so frustrated that they don't contribute. I
haven't read any rave reviews about how well the editor works.
jjb
Robert Kaiser
> I disagree that it's limiting productivity. I'm easily among the
> biggest (in terms of sheer volume) contributors to the site at present
> (not that this is some kind of contest, I'm just saying that as a
> component to this argument), and I find the current environment vastly
> easier to work in than MediaWiki. Aside from a few annoyances, some
> of them admittedly significant -- but in the process of being fixed --
> I'm much happier now than I was in MediaWiki.
And actually, the argument that counts is the one of those people who
contribute most of the work.
I've been silent on liking MediaWiki better as with my very few
contributions, it was easy for me to stop contributing at all and going
unnoticed (which I did because I just can't manage to do useful work
with the Deki editor, the simplicity of MediaWiki is hard to beat for me
there).
I don't have the urge to do a lot of editing there most of the time
anyway (not to mention that I lack the time for doing substantial stuff
there), my largest problem recently with MDC was that it lacks some
content that still is available from xulplanet only, which I saw that
with nsITreeBoxObject - but that's something that's probably not
connected with what wiki system is in use.
Robert Kaiser
Robert Kaiser
> I didn't say I wanted it to be mostly democratic (as I've also argued
> here before)--that would indeed be a nightmare, just as it is in the
Sorry, but I lost you right with going to US politics, which is
controversial by itself to not be an argument in either direction (yes,
I apparently have completely different views than you on that one).
> Actually, some kind of elected community board like Mediawiki has might
> not be a bad idea (and give a venting outlet too).
I don't even believe in elected boards to be beneficial. If you want to
see how well "democratic" votes work in open source communities, see the
recent Debian vote (<http://lwn.net/Articles/311695/> et al.)
Robert Kaiser
John J. Barton
> Sheppy wrote:
>>
>> I don't personally want to switch back, because, as I said, I find
>> myself much more productive now than I was before. And I still get
>
> You are have a completely different experience that at least I am. There
> has to be some explanation.
>
I started to investigate by running FF with a clean profile and
re-editing a page I touched this week.
But I failed. Turns out there are two *different* copies of the page!
https://developer.mozilla.org/En/Installing_Mercurial
https://developer.mozilla.org/en/Installing_Mercurial
Just for grins I opened the editor on one of them. It did come up a bit
faster, but still the crazy jumpy thing. And when it ready to edit I see
stuff like this:
----------
Installing Mercurial
{{template.Note("If you have not yet read the <a
href=\"en/Mercurial_basics\">Mercurial basics</a> do so now, or see <a
href=\"en/Mercurial\">Mercurial</a> for other resources.")}}
Installing ...
----------
Now I really don't get it. I thought the point of Deki was to avoid
geeky wiki markup.
Not encouraged,
jjb
Atsushi Shimono
On Fri, Jan 9, 2009 at 07:45, Benoit Leseul <benoit...@gmail.com> wrote:
> On Thu, Jan 8, 2009 at 17:23, Sheppy <the.s...@gmail.com> wrote:
>> We went out of our way to solicit feedback before the
>> transition, and nobody had any significant complaints. Once we
>> committed to the change, going back is practically impossible (unless
>> someone wants to volunteer to copy all the changes back into
>> MediaWiki), which is why we gave the community many, many months to
>> comment on not only the proposal but on test versions of the redone
>> web site.
>
> Sorry but I have to "respectfully disagree".
>
> The problems we are seeing were exactly the ones we were fearing when
> the switch was announced: loss of the wiki syntax, disruption of our
> habits, availability of statistical tools, unusable diffs, slowness.
And, data storage (xml is ok for the backend?), useless? wysiwyg...
> The net result is that the French MDC is not maintained anymore. All
> our contributors just left one after another.
> I can be mistaken but looking at the "Recent changes", I think the
> situation is more or less the same for the other localized versions.
I think I should agree with you. Might be the same for Japanese.
Thanks to a few, yeah, a few heavy contributors, some (and might be
main?) documents are continuing updated now.
Ah, yeah, i know my last contribution to the wiki document was at
about half year ago,,, might be.
> I don't see anything like the healthy competition we had one year ago
> when we were just trying to grow faster than Japanese or Polish with
> dozens of edits per week for each locale.
Yeah, our graph system was,, yes *was*, not the main product which
I wanted to lunch. We had,, yes this is also *had*, some useful tool to
track the changes between English wiki and Japanese wiki, which can
view and search the last-updated time with linking via the inter-language
links. What we want to lunch was, was the wide-spread status tools like
above for all languages.
As I didn't think seriously that the something easy-to-view is so
important for the healthy competition, I didn't take to discuss about
the status tools (in deki). Also for the separating the template pages
language-by-language. (Yeah, as you know, we have heavy contamination
on each language's RSS by the other languages' templates or user
pages.)
> To end on a more positive note, I will try to restart our effort once
> more after the announced update which seems to fix at least some of
> the worse showstoppers. I hope we will just be one year late.
I hope so, too.
As my last status update of the MDC Japan Project, we already
discontinued some of our serivces, and marked as 'no plan' for the others.
Of cource, I think I cannot make the same things, b/c some key features
are not impremented in Deki, as noted at my last status update. But I'll
try to figure out how can I do. :D
# refer http://www.mozilla.gr.jp/~shimono/blog/?p=57
and sheppy wrote:
> The fact that nobody complained until it was too late to reverse
> course is very frustrating for me, as the input of the community is
> extremely important to me.
I think the first time that we heard about MW=>Deki switch plan was
on your blog post
> MDC: Big changes ahead
> http://www.bitstampede.com/2007/11/16/mdc-big-changes-ahead/
and you said in this post,
> After a lot of research, it looks likely that we'll be moving the Mozilla
> Developer Center away from MediaWiki, which we've been using for
> some time now, to Deki Wiki. There are a number of reasons we're
> probably going to be making this move
Where do we have 'a lot of research'?
I might think, and thought at that time, that this is the decision of the
Mozilla, but not the devmo community. :-p
Of cource, I think
"we were just beginning to understand how to handle DekiWiki"
borrowing from Benoit's comment at the beginning of the deki
discussions.
regards,
--
Atsushi Shimono - shi...@mozilla.gr.jp
http://www.mozilla.gr.jp/~shimono/blog/
Mozilla-Gumi : Japanese Mozilla Users Group / Server Administrator
Bugzilla Localization Working Group Member
Justin Wood
visual editor and my dislike for having 3 (or is it 4 now) wiki variants
hosted by Mozilla.
That and I have always found the MediaWiki syntax useful and easy to
remember, and being forced into either "full XHTML" or "WYSIWYG" to be
disheartening. (I do know XHTML but would rather not have to use that
complex to edit an article).
I had presented this concern in the past when I first heard about the switch
to Deki, but I do acknowledge that Deki seems to be better in *general* from
"target audience" for a Developer Wiki, even though I have greatly reduced
my contributions due to my concerns here.
I look forward to the day that editing is even easier, and I also recognize
that WYSIWYG is preferred by a large number of people, even a good number of
developers, and my preference may just stem from always writing HTML from
scratch for many years (never using WYSIWYG even when it was available to
me).
--
~Justin Wood (Callek)
potappo
> I continue to talk regularly to the folks at MindTouch to advocate for
> our needs. I've been chatting with them today about their work on the
> diff and history features; they've definitely made some good progress
> and I'm nudging them to do a few more things to it before they ship
> their next release.
>
On Fri, Jan 9, 2009 at 07:45, Benoit Leseul <benoit...@gmail.com> wrote:
> The net result is that the French MDC is not maintained anymore. All
> our contributors just left one after another.
> I can be mistaken but looking at the "Recent changes", I think the
> situation is more or less the same for the other localized versions.
Diff and history features are most important features for localizers.
Less features like now DekiWiki, localizers need more time than
MediaWiki.
So, l expect the next release to call back localizers.
2009/1/10 Atsushi Shimono <shi...@mozilla.gr.jp>:
> Of cource, I think
> "we were just beginning to understand how to handle DekiWiki"
> borrowing from Benoit's comment at the beginning of the deki
> discussions.
I think so, too.
--
potappo
MDC Japanese Translation Leader
mail:pot...@gmail.com
Sheppy
wrote:
> John J. Barton wrote:
> > Sheppy wrote:
>
> >> I don't personally want to switch back, because, as I said, I find
> >> myself much more productive now than I was before. And I still get
>
> > You are have a completely different experience that at least I am. There
> > has to be some explanation.
>
> I started to investigate by running FF with a clean profile and
> re-editing a page I touched this week.
>
>
> Just for grins I opened the editor on one of them. It did come up a bit
> faster, but still the crazy jumpy thing. And when it ready to edit I see
> stuff like this:
> ----------
> Installing Mercurial
>
> {{template.Note("If you have not yet read the <a
> href=\"en/Mercurial_basics\">Mercurial basics</a> do so now, or see <a
> href=\"en/Mercurial\">Mercurial</a> for other resources.")}}
That's not geeky markup, that's a template. Not much way to get
around that kind of thing if you want to use templates.
What makes you think those two pages are different? They're not; the
capitalization of the language identifier doesn't matter.
The jumpy editor thing is a configuration on our end, and we're trying
to figure out why it's happening. We suspect it's an unfortunate
interaction between a floating element in our skin and FCEditor. Work
is ongoing there.
John J. Barton
>> {{template.Note("If you have not yet read the <a
>> href=\"en/Mercurial_basics\">Mercurial basics</a> do so now, or see <a
>> href=\"en/Mercurial\">Mercurial</a> for other resources.")}}
>
> That's not geeky markup, that's a template. Not much way to get
> around that kind of thing if you want to use templates.
Sure I know that, but I'm a geek. My point is if we have this kind of
stuff anyway we didn't gain that much by leaving wiki markup (which
wikipedia has shown is accessible by any one who might be interested in
reading MDC anyway).
jjb
Sheppy
wrote:
> > That's not geeky markup, that's a template. Not much way to get
> > around that kind of thing if you want to use templates.
>
> Sure I know that, but I'm a geek. My point is if we have this kind of
> stuff anyway we didn't gain that much by leaving wiki markup (which
> wikipedia has shown is accessible by any one who might be interested in
> reading MDC anyway).
The point, though, is that now the vast majority of work can be done
without needing to know any markup, or even know that templates exist.
Benoit Leseul
> As my last status update of the MDC Japan Project, we already
> discontinued some of our serivces, and marked as 'no plan' for the others.
> Of cource, I think I cannot make the same things, b/c some key features
> are not impremented in Deki, as noted at my last status update. But I'll
> try to figure out how can I do. :D
> # refer http://www.mozilla.gr.jp/~shimono/blog/?p=57
You don't have to do it alone. I think many of us care about such
things and would be willing to help if we have a good idea of where
we're going.
Let me know if I can help to (re)implement some of the wanted
features. I'm mostly fluent in PHP but can learn any other system
quickly if needed.
For example, what is missing in order to make a basic interlang bot
(just notification of the missing links, I'm not even talking about
automatic tagging yet)? Is there something available in the Deki API
in order to do this or should we hack some kind of pseudo-HTML parser?
--
Benoit
seven...@gmail.com
> I disagree that it's limiting productivity. I'm easily among the
> biggest (in terms of sheer volume) contributors to the site at present
> (not that this is some kind of contest, I'm just saying that as a
> component to this argument), and I find the current environment vastly
> easier to work in than MediaWiki. Aside from a few annoyances, some
> of them admittedly significant -- but in the process of being fixed --
> I'm much happier now than I was in MediaWiki.
On Jan 9, 12:30 am, "John J. Barton" <johnjbar...@johnjbarton.com>
wrote:
> You are have a completely different experience that at least I am. There
> has to be some explanation.
I don't have the time to write up a response to many of the points and
comments
made about the transition, but I think it's important to note this
aspect of
the experiences post-changeover. I believe, quite confidently, that
the
disparity here is that Sheppy is primarily a content generator,
writing
documentation on that which didn't have exisiting material to work
with. On the
other hand, you have those spotting changes that need to be made to
existing
content and formatting. This part encompasses a fair share, moreso now
as a
result of the changeover.
Depending on how you look at it, this kind of stuff can be worse. On
the face
of things, you'll see a decent, albeit sparse page, or maybe even a
lengthy
document. If you don't quite grasp the subject of the article you can
move on
thinking that need is covered. However, a surprising amount of
documentation
still on Devmo includes stuff from the DevEdge days, inadequately
covering
important subjects and mentioning now-obsolete topics. This is the
tedium. Add
to that the general wiki maintenance--the stuff that provided the
impetus for
the tools and features developed as part of MediaWiki to make this
work more
manageable.
Like I said, there are a lot of things brought up here, and if I had
the time,
I'd like to give my input on the whole thing, although it would be
quite
lengthy, and I'd hate to do a poor and incomplete job in doing so. But
I feel I
can say this with it being quite self-contained: the changeover to the
not-yet-
mature product that is Deki Wiki has made it painful for anything
other than
basic article creation and minor additions of content. As for non-
trivial
maintenance, the really tedious stuff that no one really wants to do
anyway,
its difficulty has increased severalfold.
Sheppy
editing as well (except during these late in product development cycle
periods where I'm heavily focused on new content for an impending
release).
That said, we're working hard with the guys at MindTouch to ensure
that the capabilities needed to make editing and other routine chores
easier get added to the software, and many of those are coming in the
next release.
I recognize that we may have jumped the gun and switched to Deki too
soon. I do continue to believe it was the right move for us, but it's
certainly proven to be a rougher transition than I expected it to be.
If we had realized sooner just how many frustrations we'd run into due
to features not yet implemented per our needs, we'd have delayed at
least a little longer.
That said, it's time to look forward instead of back and do everything
we can to smooth things out going forward. I honestly feel that the
Lyons release of Deki due next month is going to make most of you much
happier, and we'll soon be in a position to implement extensions for
Deki that will let us do some very cool stuff.
Atsushi Shimono
Hash: SHA512
hi, and sorry for late responce due to my univ work (incl. business trips)
# and i'll be offline from 2/1-22, 3/1-15 due to another business trips. :D
Benoit Leseul wrote:
> On Fri, Jan 9, 2009 at 18:49, Atsushi Shimono <shi...@mozilla.gr.jp> wrote:
>> As my last status update of the MDC Japan Project, we already
>> discontinued some of our serivces, and marked as 'no plan' for the others.
>> Of cource, I think I cannot make the same things, b/c some key features
>> are not impremented in Deki, as noted at my last status update. But I'll
>> try to figure out how can I do. :D
>> # refer http://www.mozilla.gr.jp/~shimono/blog/?p=57
>
> You don't have to do it alone. I think many of us care about such
> things and would be willing to help if we have a good idea of where
> we're going.
Yeah, I know.
> Let me know if I can help to (re)implement some of the wanted
> features. I'm mostly fluent in PHP but can learn any other system
> quickly if needed.
Statuses and overviews of the past? system is here (2008/Aug's)
http://www.mozilla.gr.jp/~shimono/blog/?p=57
I think it might be easy (than the others) to log bot like
http://bugzilla.mozilla.gr.jp/show_bug.cgi?id=5298
Of cource, some lack of features on multi-language of Deki, we should
write some another filter, but basically we can use the same algorithm
as before. (number 5 and 8 in the blog post, written in php/5 or perl/8)
For some security reason, like we've hardcoaded passwords or login cookies
of Mgjbot in the source :-D, we can never open the source repository, but I
can give the current source archive of the tools.
> For example, what is missing in order to make a basic interlang bot
> (just notification of the missing links, I'm not even talking about
> automatic tagging yet)? Is there something available in the Deki API
> in order to do this or should we hack some kind of pseudo-HTML parser?
For the interlang bot (number 4), we've first customized pywikipedia bot.
After that, to make the document conveter or something, I've wrote some
basic tools in perl (number 1).
Yeah, I tought, but not done yet, we should search something available in
the Deki API to do such basic manipulations. (or, if no API, we can trace
and analyze the network stream to make them :D)
regards,
- --
Atsushi Shimono - shi...@mozilla.gr.jp
http://www.mozilla.gr.jp/~shimono/blog/
Mozilla-Gumi : Japanese Mozilla Users Group / System Administrator
MDC Japan Project Leader, Bugzilla Localization Working Group Member
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (MingW32)
iEYEAREKAAYFAkl/37YACgkQBygbhFUCKWZqNgCdF4YAeDneDXMR1oKy4hzwaIfY
6dsAnA+7/celvKbgsUo20YcSbBTMfi1L
=WjaP
-----END PGP SIGNATURE-----