Misplaced Pages

Revision as of 20:53, 14 October 2006

Benchmark link

Benchmarks link was changed to a test on a forum that compared Swiftfox without extensions to other Browsers with extensions. This is an apples to oranges comparison. I reverted it back to a benchmark of freshly installed applications.Kilz 18:48, 7 October 2006 (UTC)

The ubuntuforums benchmark link is valid as it compares Swiftfox without extensions to Konqueror and Epiphany (both without extensions). Just because that particular benchmark includes Firefox with extensions doesn't invalidate the comparison it provides between Swiftfox, Konqueror and Epiphany. —The preceding unsigned comment was added by 74.71.112.182 (talk • contribs) 20:34, 8 October 2006 (UTC)

The benchmark would need to be done by a reliable source. Otherwise it is just the say-so of some unknown forum poster and does not warrant an external link on Misplaced Pages. —Centrx→talk • 21:32, 8 October 2006 (UTC)

Since Swiftfox, on the Swiftfox site says its only a compile or build of the Firefox code, is it correct to call it an offshoot? The site doesnt say there are changes, but that the build is simply optimised to enable or disable things.Kilz 02:22, 12 October 2006 (UTC)

I've changed this to the correct term - build. Widefox 15:13, 12 October 2006 (UTC)

Free download

I've added back the fact that this is a free download. Please do not delete this again. I checked the Firefox article and the download and official site is listed multiply, so this is in keeping, especially as the box to the right does not reference Swiftfox, but Firefox.

But no other Custom Build or fork has this section. I don't think its right to compare this short article to the main firefox one. It is better to compare it to the other custom builds that have the same length. Either they all need to have download sections or this one needs to be removed. This whole page is starting to look like an advertisement for Swiftfox. That is not neutral, positive bias is still bias. This needs to be addressed. I will give you time to resopnd or edit. If not I feel so strongly that I will remove it.Kilz 14:42, 14 October 2006 (UTC)

Optimisation

the build info indicates swiftfox is built with optimisation -O3, where normally -O2 is used. I added this back, and if you don't agree, please do not delete, but just mark with , as right now I can't see the build info on the redesigned Swiftfox site, or in the discussion group. Widefox 15:13, 12 October 2006 (UTC)

IMHO the multiple download references take away the neutral point of view. It appears we are selling this version by pointing out that multiple versions are available in the opening, adding a download section that is in no other Firefox custom distribution article. Then adding multiple links to the Swiftfox main site. Yes that it can be downloaded is a fact. But is it necessary to constantly repeat it throughout the article. None of the other versions say they are avilable for download, the main Firefox page has information on how many times Firefox has been downloaded as part of over all web browser usage. Its also 7 times as big, and none of the other articles have links to the main site as references and external links. Maybe if it was only in one place it wouldn't look as bad in such a short article. How about just adding it can be downloaded to the top and remove the download section. Maybe we should also choose what section the main site link is in.Kilz 20:58, 12 October 2006 (UTC)

I've already covered multiple refs above, in comparison with Firefox. Firefox mentions the word "free" very early on, and swiftfox not, so I think some balancing could be done, but for now it seems OK to me. I'd be more concerned about the claim that it is any faster, as that is the purpose! Certainly removing facts about free download, URLs, build info is not acceptable and I will revert immediately. Additionally, the only reason swiftfox is of interest is because it can be downloaded (as with any build), and there's a disambiguation to be done, due to the multiple versions multiplied by multiple processor types, which is again more pertinent to swiftfox than firefox. Whole point about this article is that it is a 3rd party build that can be downloaded. This article is a stub. Please continue your efforts to enlarge this article so that it reaches a more mature state. Widefox 02:15, 13 October 2006 (UTC)

I didn't say remove the fact it can be downloaded, I said maybe we should think about editing out the multiple download references and duplicate information because of the size of this article. Firefox says it a few times but its 7 times the size. I placed that it can be downloaded in the beginning and removed the duplicate information. I'm still not sure we should have a link to the Swiftfox in References and External links. It just doesn't look right. Kilz 02:24, 13 October 2006 (UTC)

Did some editing, I noticed that you are asking for citations on Font rendering. Added links to pages that explain exactly what the engines improve, and that pango improves non English fonts. That freetype2 is a default of enabling gtk2 toolkit on a mozilla page. Also a link Swiftfox to a build summary that shows they are disabled. But I do think we need citations that Swiftfox starts faster from a neutral party, and that 03 is in fact a higher compiler optimization and what that means, all I can find is that its different. Kilz 03:11, 13 October 2006 (UTC)

please discuss and reach consensus with these controversial edits. I include some pointers and discussion below. Widefox 14:17, 13 October 2006 (UTC)

I have given citations for this point. I removed the citation request. Widefox 15:37, 13 October 2006 (UTC)

additionally, please could I ask you to keep to the topic indicated in the section heading, or start a new one, makes it a lot easier to discuss individual points, for more help on this, checkout the guidelines. thanks Widefox 15:41, 13 October 2006 (UTC)

controversial edits

I had to revert the article, and I have justified that in the edit comment, and in more detail here...

I'm pasting in some out-takes to save them.

A faster startup time, together with a small speedup (1.7%) has been reported.

this is not needed - it is already referenced -> removed

• by removing pango layout and text rendering support. This can result in lower quality text with some fonts. It can also cause problems with fonts used in languages other than English.
• by removing Freetype2 font rendering support, a default of the GTK2 toolkit that is enabled. This can result in lower quality text with some fonts. + * by removing Freetype2 font rendering support, a default

please use {{cite web}} - for style see the reference already using cite web in the article.

Also, the changes to the article did not qualify the criticism "quality", or detail which aspects are affected, or indicate where that information is.

licensing - please do not revert my edit about licensing - you wording highlights what the licensing is NOT, not what it is, so is less clear. This is the second time I have said this. Please justify your reasoning here before re-inserting.

number of references to downloads - I don't understand what you mean about there being too many - you have just added one more at the top, while removing the clear section at the bottom. Also, just because the article is a stub, doesn't mean that multiple references to download should be deleted. I have already justified and compared with Firefox. Disambiguating is important due to the total number of downloads available (more than Firefox), so your overall aim does not seem in keeping with the needs of the article. Widefox 14:17, 13 October 2006 (UTC)

In order to prove that it is not a free license one has to understand what a free license is. The link explains that if some things are removed the license is non free. Swiftfox has taken a free application and licensed it under a non free license. Why is it conciderd non free is the reason its linked. Second, I gave more than enough info on the fact that the font rendering engines were removed and what they do. Also what types of fonts they affect. You changed all that information. I then removed duplicate information by adding download to the statment above. IMHO we dont need that things are avilable, then they can be downloaded. This line "Several versions of Swiftfox are available for specific types of microprocessor architecture." and this line"It is available as a free download for several architectures, covering both Intel and AMD processors, and several versions of Firefox." are the same thing. This is in no other firefox custom build page on wiikipedia. You keep changing my edits. It does not appear we can reach agreement. Next, there is no proof that Swiftfox starts faster. The benchmarks deal with rendering. Not start up times. There is no proof it starts faster. The references to Swiftfox forums and its site should be backed up by an Independent secondary source reliable source

It you intend to use Firefox benchmarks for Swiftfox, expect the free software license to be replaced. I discussed why I removed the duplicate line above.

It is conventional practice to sign your comments. I do not fully understand what you mean here. Widefox 19:28, 13 October 2006 (UTC)

If you expect me to discuss why something is changed. I expect the same in return. Simply changing my edits will force me to revert them. Kilz 18:57, 13 October 2006 (UTC)

If I understand you correctly, you have an objection to Swiftfox - in particular it's licensing. If others feel that way, so that you can find a reference, you might want to add that into the article. Currently, I consider that your changes just detract from the article with your point of view (POV). Find some references and add it in, the article is small! Please could you look at wikipedia guidance on POV and original research though.

Please refrain from repeatedly changing a referenced and valid licensing statement, without discussing it here, as I have done. Also, you have changed spelling from British to American (licence, license), which according to guidelines is questionable in itself - given that I am the main author now (see guidelines on spelling), but the thing is, it's only used inside markup, so it doesn't really matter, does it? Please could you provide more edit summary (as per guidelines) - 2 of the last few were blank. I would prefer if you made some attempt to discuss before going any further, as is considered best practice in this circumstance. Please consider what I am saying here, as I will only revert your edit 1 more time before before I consider that I've come up against the 3 revert limit, and will be forced to escalate. Widefox 19:28, 13 October 2006 (UTC)

Then escalate. I was working on this article before you got here. How did you become the main author? You have changed things without taking one thing I have said into concideration or discussing it with me. Deleted and changed what I added without one discussion. But you expect me to clear everything with you. My point of view is the truth. The license is non free, I am suprised you dont see it as you use Debian. Maybe you dont think that informattion is important, but I do. You are pusing your point of view that its free as in cost, more than free as in freedom. I linked to a article on wikipedia (Free Software License) explaining why its non free as in freedom and you remove it. I added information to show the license is different than every other Firefox fork or build. Even the sorce of the code (Firefox) and you remove it. But add references to speed that dont even apply because they didnt benchmark Swiftfox. You have made it into an advertisment, not reference. You are more interested in showing its avilable for download and linking to its download page. Not just once but mutiple times in a very short article. That in my opinion is selling it. You have a POV problem. Kilz 23:13, 13 October 2006 (UTC)

I'm currently merging your refs, with the article. Just for your information, I fully understand your enthusiasm for free software. I've been using Linux over 10 years, and as you might have seen on my userpage, I work on open-source projects, so I do understand where you are coming from. It's just that you appear to be new around here, and I've found that things go much smoother on wikipedia if you talk through controversial edits to reach consensus beforehand, and not assume the worst. I think your point about licensing is valid, and I've already said, find a ref and add it in. That doesn't mean that the whole article should be a licensing issue. Another question - do you know how to get to your talk page? You didn't reply to any of the points there, that I made to discuss things. Widefox 23:41, 13 October 2006 (UTC)

Using Firefox benchmarks

I have steped back from this article. I have only eddited things you have asked (which?) to. But I do have some concerns on other areas. 1. I still see no independant reference that Swiftfox starts faster than Firefox. I feel we need some indepandant conformation of this from a site other than Swiftfox, or its forum. I cited reliable source, specificly the Independent secondary source section, above. Please respond to this or add another site with benchmarks as to start up. 2. I think Free in the download section needs to be qualified. It is directly over the license section that says its non-free. Maybe "No cost" would fit better. At present I am gathering citations as to the licenses that the other forks are released under. So far no other fork is restricted. Kilz 03:47, 14 October 2006 (UTC)

This section is faulty reasoning to me.
" A different comparison between Firefox and other browsers has also been done,allowing an indication of the absolute speed of Swiftfox:"
We cant compare Swiftfox to other browsers based on what Firefox has done. For all we know it could be slower in those areas than Firefox. Swiftfox does not use the Firefox binaries in any way. This is why the benchmarks that list Firefox against other browsers is inappropriate to use when suggesting Swiftfox is just as fast. So what basis are you using the above quote and information? I feel it needs to be removed as it has no Swiftfox information. I will give you time to respond or remove it. Kilz 13:55, 14 October 2006 (UTC)

many points. General status comment - this article is far better reff'ed than the other 3rd party builds, and insertion of more licensing issues starts to look like political POV (these issues are under Political and not Licensing in the Firefox article) as is the use of "non-free" generally associated with Debian politics (and politics are best kept to a minimum - this is no place for political activism according to wikipedia guidelines). You do realise that prominently, Firefox has "free" listed. IMHO, that is more questionable than the (now removed) statements here, due to prominence. OK, more licensing issues are great (IMHO) as long as there are references - as I encouraged before. Currently I have no evidence that the authors claim is not to be taken at face value. My opinion is that 3rd party builds are a trust issue, not a licensing issue, but that's my opinion, that if I felt more strongly, I might go and find refs of browser build abuse (as is popular with toolbars). I noticed he's put up Debian binaries BTW. OK, now speed test reference - my recollection is that it mentions faster startup. The way to go is just to compare binary sizes, this being primarily the reason. Binary size should go in, especially in light of lack of hard benchmarking. A ref for that would be best, but can be done by reffing the two download locations. 1.5.0.7 is the one to go for. The faulty reasoning, is not faulty, and the topic is central to the reason to download swiftfox. It was however badly worded in the early hours! I've changed it now, with disambiguation to the larger issues and more performance issues, and benchmarking. As you say, we cannot draw a direct comparison for swiftfox vs. opera etc, but for ballpark reasoning FF vs SF, then FF vs. Opera it has limited merit. Note that at the big speedtest reference, they test an early FF 2.0, and I believe a pre 1.5, so it's not definitive at all either. The author of that has said he will update for FF 2.0. removal of these indications of absolute rank in terms of speed would seem counter-productive, given this is a speed issue. Widefox 16:05, 14 October 2006 (UTC)

If you would perfer proprietary software instead, maybe we should change it. I strongly believe that if speed claims cant be proved by direct links and evidance that it should be removed untill we find such evidance. It apears we are writing an advertisment not information. "ballpark reasoning" cant be proved. Kilz 16:31, 14 October 2006 (UTC)

Info Box

Swiftfox is not an extension it is a build. It is wrong to list it as a extension. I corrected this and it was reverted. I will correct this again. Respond here before reverting it again. Kilz 14:48, 14 October 2006 (UTC)

just typo. Fixed.Widefox 16:05, 14 October 2006 (UTC)

was merge issue, not a revert due to editing at same time - my fault though. only one to slip through. Widefox 16:11, 14 October 2006 (UTC)

Use of Swiftfox site as a reference, Positive bias problem

As I understand the reliable sourcepage. We should use Independent secondary source if we use any citations from the site/project we are describing/writing about. I notice we have a lot of references that go directly back to the Swiftfox site, and to its forum. Without any neutral source listed. As I was asked to discuss things I see as problems before editing them, I am posting this before removing the references. Bias, even a positive bias is to be avoided as I understand it. Please provide a link that says otherwise, or provide the neutral citations. Kilz 16:24, 14 October 2006 (UTC)

please be explicit - which bias do you suspect? also which refs do you object to. citing technical docs is normal. Knowing that you might object to providing a link to Swiftfox, I've already refrained from adding a ref to the "lack of a claim" of "speedup". I'm already over the line in your direction. I detect from your comments that you have a strong opposition to the Swiftfox licensing terms. Pushing that harder will only force me to consider the article Debian "non-free" political POV. Have you tried emailing the author to complain? Is there an article where non-free software can be incorporated. I just don't agree this is a place for activism. I've provided an alternative section now, as I thought that might alleviate this angle you are coming from. Widefox 16:48, 14 October 2006 (UTC)

for example see Intel. You kind of giving me a hard time here. I have wide interests - just look at my edits. Nobody has ever claimed that I am pushing some POV, so I would like you to be explicit. Just for the record, I have no affiliation with Swiftfox, Fasterfox (although I have contributed code/ideas, so I am an expert in browser speedup. I also have no affiliate with any company, and need the speedup due to not having enough CPU!). Widefox 16:57, 14 October 2006 (UTC)

You may not have a personal bias, but the Swiftfox website and Forums will defiantly be pro Swiftfox. Anything they claim should be backed up by a 3rd party imho. Specifically in the optimizations we have 2 links to the Swiftfox forums. Taking them at face value that they improve things is "iffy". I'm not trying to give you a hard time, I just want the page to be as accurate as possible. Especially in speed claims.
If you prefer proprietary over non-free so it isn't political that's ok with me. But Jason, the builder of Swiftfox has created .deb files (for Debian, spcificly listed on the main Swiftfox page)then restricted freedom by a license in the file. By doing so he has invited the free/non free issue. This isn't like he was making rpm's and didn't expect Debian to point out it isn't free. Yes I have written him and he just got more crazy, by adding the tainted versions line to the license. But users of Debian (and derivatives)deserve to know that they may be installing non-free or proprietary software since its targeted to them. Kilz 18:01, 14 October 2006 (UTC)

I'm glad you've clarified that, thankyou.

If I understand which links you mean, it's only technical details - build info. I already detailed how to check that yourself to back it up (BTW installing MR Tech About:about 2.2 will give you a menu item for easy access to those things). In which way do we have to go further?

Erm, just as comparison, Madfox doesn't even have a single ref. According to the rules, it could be deleted! We've got 17 now! We do not follow the rules to the letter, else wikipedia would get small! I think a sitting back and congratulating ourselves is in order! This is the first article you've worked on right? could you do an A-B comparison at this point with some others of similar stature. The main advertising aspect comes from that darn Firefox TOC, which will send people there, especially with a name like Swiftfox!....

• IF* I was going to play devils advocate (which I do to ensure POV), and take your anti "non-free" stance, the TOC would be the key to it all, in terms of eyeballs, and not refs backing up easily verifiable tech documents. As you can see, I've removed the TOC where I can for legal issues (it's come back 2x already). For that legal issue, I don't think any of the 3rd party builds should be listing in the same box as Firefox. It is not up to me to pursue my licensing beliefs here, even though I have worked on the Linux kernel, and am an extensive open source contributor. That is separate. It is up to the reader of wikipedia, and the user of the software. In the end, all our words will be edited away!

In an ideal world, we should all run Gentoo, as it is true to Unix / open-source tradition. No hidden gotchas there. That's what I learnt in the 80s. But that's just my POV. I'm a Kanotix user myself, but there's no userbox for it, so I list it as Knoppix. I guess you run Debian (Kanotix is 100% Debian).

Speed claim - they don't make one! Of course, if they called it Slowfox, nobody would download it. Interesting, as I've put in the article. They just claim "optimization". FYI I've heard of significant MMX/SSE speedups by using a patch set by mmoy (It's only for Windows). In my experience, manual optimisation like that is way more effective (typically >10-20% faster) than relying on the compiler (~same unless trivial case, sometimes slower), like swiftfox folk do. I'd be interesting to see a build with an Intel compiler though. It produces significantly faster binaries.

I cannot and will not comment on Jason as I have never met him, emailed him (as far as I can remember), although I have nudged him for new builds, due to using FF 2.0 nightly builds. I also use FF official nightly builds, and so remain agnostic when it comes to browsers. They both have imperfect licences. Widefox 19:06, 14 October 2006 (UTC)

Well I don't run Debian, I run Ubuntu. But its a Debian derivative that strives to using free and open source software. There are a few of us on the Ubuntu forums who oppose non free software. I also fully agree with the 4 basic freedoms as defined by the fsf. Jason has tried to stop me from telling people Swiftfox is non-free. I'm also the person he changed the license from MPL to proprietary to stop me making 64bit deb files for the amd64 version of Ubuntu. So that it wouldn't lower the clicks he gets from his google adds. I think I know how unethical he is, and know he will lie to win. I have tried to limit my opinions of him, and just stick to the facts. That's also why I stepped back. Personaly I use Iceweasel :)

Firefox TOC

I see the Firefox TOC is back! Please, before adding this back in, please see the discussion at the Template_talk:Firefox_TOC. Please, someone just create a 3rd party build TOC as someone else suggested so that 1. we get no legal issues 2. the Firefox TOC does not come back. Please direct all talk to that page, unless specifically Swiftfox related. Widefox 17:21, 14 October 2006 (UTC)

• Yeah, I did that, the TOC is included in the rest of the forks with no problem, see Template talk:Firefox TOC for relevant discussion. To prevent edit wars, I'm not reinserting TOC. Feureau 19:16, 14 October 2006 (UTC)

1. "Benchmark at APC Magazine".