I was building Crowdwide, a social media platform I started for new indie devs, smaller communities, and smaller creators to raise their voices higher.
Your voice shouldn't be judged based on how big a platform you join. Everyone's voice should be equal.
And today is Day 21 of building Crowdwide!!!
YESSSSS!! π₯π₯
Crowdwide is in a mostly stable state now! Its core features, main algorithm, mailing system, and recommendation system, Widewave, are finally DONE!
So now, it's time to polish the UI.
On Day 21, I'm just polishing the UI of Crowdwide.
The things I thought needed the most improvement were the background color combination, theme adaptability, and sidebar (THIS NEEDS HEAVY IMPROVEMENT!).
The background used to have gradients, but when users scrolled a lot, the background stretched and created a weird shade at the top.
In some themes, a few text colors weren't looking good and were making things difficult to read.
And the sidebar... I was SO confused about how it should look!
Like, should it be on the left or right?
When it opens, should it compress the page?
Or should it stay over the page without compressing anything, making the rest of the screen a bit darker to keep the focus on it?
SO MANY QUESTIONS π
FIXING...FIXING...FIXING...FIXING...FIXING...FIXING...FIXING...FIXING...
AND
DEPLOYING...DEPLOYING...DEPLOYING...DEPLOYING...DEPLOYING...DEPLOYING...
AND IT'S DONE!!!
I've fixed all those things, and now the UI looks better. I THINK THE UI LOOKS BETTER NOW!! ππ₯
Also, a shoutout to Harun!
My brother Harun is organizing #HackerStage on October 25th! π₯ If you love coding, building cool projects, or connecting with fellow builders, this is your chance to jump in, challenge yourself, and have fun with the community!
Check it out and join the jam π
https://hackerstage.netlify.app
Show some love, guys! π―π€
And now, I think the things left are just handling basic updates based on requirements and user requests, and checking moderation.
THAT'S IT!!
21 DAYS OF BUILDING CROWWIDE! π
LINKS π
π Crowdwide: https://www.crowdwide.run.place/
π» GitHub Repository: https://github.com/DeveloperPuneet/Crowdwide
See you in the next update! π
Top comments (239)
@akhourianmolkumar @koda2026 and @technogamerz joined, so me too joined! π₯³
Bro i just joined now and bro when i use '@' and type the lazy girl i cant find her and i tried it with ππ‘π πππ³π² ππ’π«π₯ still i cant find it and i cant find her with technogamerz how you find her lol π€£
Me too cannot find her so I just type it and then verify by clicking the link after commenting (it takes extra work to mention her lol)
Wait i need to test it @technogamerz
Yeah it works hahahaha π€£ bro actually lets talk in my latest post please
And my username bro????
It works perfectly @star123
Oh... Okay! Thanks!!
@star123 yeah it works now bro can we chat in my latest post
Me?
you online now. Where have you been all these days?
@star123 I had gone to Kerala for some work.
Kerala!? π± Where?!
WHATTTTT!??????? π€―π€―π€―π€―π€―π€―π€―π€―π€―
You came to my place?! FOR REAL? Or just joking?
For real, Iβm not even joking. I didnβt know that was your place.π
BUT WHY WOULD YOU COME TO ALAPPUZHA?!!! What kind of work is that?!
Because my friends live there, I went to meet them.
What a coincidence is that?! Where are you now?
Interesting thread... meanwhile me: READING π
Bro i can't believe she came to Alappuzha. I am right in the center of Alappuzha. @technogamerz were you near the Alappuzha railway station, the Alappuzha beach, or something?
But bro, youβre shy, arenβt you?
π
@puneetkumar2010 ππ»Huge Respect Puneet superstar ππ»
@effessdev
I mostly love traveling around South India, especially Kerala and Tamil Nadu. It was my first time visiting your place, and honestly, it was so beautiful, truly stunning! If we ever get the chance, we should all meet someday, offline or online, haha. Next time I visit your place again! Iβll definitely come to your college. I donβt even know which college it is yet, hahaha! π€£
Haha, superstar? ππ ππ You're making me famous already! Thank you so much for the love and support! π And yesss, we should definitely meet someday, online or offline! π
@technogamerz I was about to ask you what you think about our place but you already said?
My college is not in Alappuzha and also it's on top of a hill lol. It's very hard to travel to there. So not college but anywhere else... π€£ (I don't like schools & colleges)
@akhourianmolkumar bro you said that I'm shy right??
Do you play hill climbing IRL on your way to college? :D
heheheheh π
bro come online please
Aha I got it!
@akhourianmolkumar bro I want to come online but I have a stupid assignment (I mean literally a STUPID assignment)
coming or not? bro my 2 days school holiday tomorrow because of severe weather π
I'll come when I finish bro...
congrats broπ
Bro how to stop getting notifications from the post
π idk... i think unsubs to thread
okay so can you be online in dev.to for some reasons (remember that??????)
Guys why this thread messages are coming to my notifications?
THESE MEMES ππππ€£ MOODI JI ON PEAK
Online?
?
yupp
What?
rizzzler.work.gd/puneet2010
Song name? π€£
Link not working :(((
Link not working:(((((( why???
My favourite Phonk song!
casually didi, flexed that he can solve all physics and didi really???????????
π IITian listening phonks.
π π i got this from youtube...
and... one more things... Actually, it was my old project lol. It was a link-in-bio platform that kinda helps you build your portfolio in one place! π π€£π€£
rizzzler.work.gd/puneet2010 try this one...
woow!! Nice taste ππ₯
@effessdev for you.....
India is changing, world is changing
IITians are sending shorts in dev.to π
didi now you are in league of own rules people
in this league currently 3 people only fall before you
Me, FS BRO, and puneet we 3 were the people who break the law of professionalism of dev.to but now we all 3 are considering you too in the league
congrats
Yeah now it's so easy to sound professional because of AI which makes improfessional sound better π
Coincidence?
Anyway thank you I liked it π
Why is it like it whenever I reply to your comment it gets deleted >_<
yeah but my messages are pure hand written from heart
π
also bro just saying
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
"We Sleep Early π€£"
Dev.to to your comments π
I don't want to sleep early but I have to wake up early so I have to sleep early π’
I am the kind of person who becomes active during the night bro... I am the most productive at night.
bruh then why you say we sleep early lol?
I laugh so much reading you twoβs comments/chats π€£
Good night guys!
because we sleep early so I can wake up early π₯²
yeah in FS bro's language: We sleep early
peak communism
I don't even know what communism is π
yeah, you should sleep early because you've to wake up early and wake up early is only possible by sleeping early and sleeping early results in waking up early, and waking up early is direct proportional to sleeping early.
FS BRO BE LIKE: MY BRAIN IS NOT BRAINING
I mean I do not sleep early because I want to sleep early I sleep early because I have to sleep early because then only I can wake up early bro... now makes sense?
ah and also I don't want to wake up early I have to wake up early because my bus comes early
MY BRAIN IS NOT BRAINING
AKHOURI SYSTEMS FOUNDER AND SLY GUY FAILED TO PROCESS THIS
6:00 π€
6:00?
what time do you wake up 6 o'clock?π€
Let me be translator between you two leaders:
so didi means that is your bus time 6:00am
6:20... and I leave home at 7:20
bro that's my school time π
Ok you two gone? Good night π₯±π΄
School time what?
Good night π€π
@akhourianmolkumar is your school time 6:20 or 7:20?
6:20 my bus arrives
I woke up at 4 o'clock in the morning!
:(
π€£ OKAY so I am in a much better position... π€£π€£π€£
peak IITian
Bro I never thought I had more sleep I am happy :)
bro just a random thought came in my mind if we forgot our names and represent it as age
then you probably 23-25, didi probably 27-30 and me 13
means 24, 27, 13 talking what a ages
Ok so bye good night π
hmm, gd night we sleep early and please try to find upgrades for apic v2
22!!!π€£π€£π€£π€£
whatt? didi how can a IITian be 22?
like how? did you skipped any class in school?
Yeah π€£π€£π€£π€£
woahhh, I also skipped UKG class so I'm 13 but in 8th class I should be in 7 I guess
but it means FS bro is old?
BRO HOW?? I also skipped UKG!!!!
bro was sleeping but after listening I skipped UKG he came lol
Yeah I skipped UKG so I am in 8th class
1 november 2004(not official but real)
Wow! Coincidence? Again?
But I had to sit 2 times in 3th standard because my father's stupid friend told him that CBSE is hard so I should study 3rd standard again because I moved from state to CBSE lol otherwise I'd have been 2nd year by now :(
WHAT??? AGAIN COINCIDENCE? @technogamerz me too November but 2nd!!! π±π±π±
So @technogamerz @effessdev and @akhourianmolkumar skipped classes, @effessdev and @akhourianmolkumar skipped the SAME classes, and @technogamerz & @effessdev were born in the same month?? All coincidences?!!!
whattt you are 2004 but a IITian?
my math is not mathing I thought you are like 2000s or 1998s because IIT takes long time right?
Another coincidence: my mom's and your birthday is the same day!!!! didi
What a coincidence!!!π―π―π―
Yeah why sooo many coincidences?? So I'll tell you happy birthday when it's my mom's birthday!
coincidences ππππ
Ok good bye everyone this time I am truly gone. πββοΈ
lol
Good morning! These are all the coincidences we discovered in this thread:
@technogamerz @effessdev and @akhourianmolkumar skipped classes, @effessdev and @akhourianmolkumar skipped the SAME classes (UKG), and @technogamerz & @effessdev were born in the same month, @effessdev 's mom and @technogamerz have the same birthday, @technogamerz came to visit my district!
Did i miss anything?
β οΈπ
@effessdev I'm back
I send you message on discord!
I replied too.
yes that didi award of dev 2026 is left!
hmm, I'm happy that I skipped UKG because now in less age I'm in 8 means same like Didi If I ever became a IITian then I will be a young one or if I came in forbes list then maybe a young person in forbes.
My DOB is 10-10-2010 π
Wow coincidence again?!!!
π€£ DEV Developers corner β DEV coincidence corner β
My DOB is 31-08-13
in aadhar card it is 2014-31august π
lamooo!! π€£π€£
UKG attendance day by day ππππ
I'm in 12th!
bro in 15th age 12 how is that possible
I have ID card... going to give exam of 12th in Feb ...
Hence Im in 12th
maths
hence, proved.
12th?????
Just Puneet things
Yes 12th
ok. by the way I am 18 & I remember @technogamerz telling me she's 2004 (so 21 or 22 years is that correct didi?)
DEV.TO is NOT DEV.TOING
puneet being puneet
β οΈπ₯π«‘
in linkedin me: π«‘
in dev.to me: π€£
π€£π€£ i have an account on LinkedIn, but i never used it!
puneet be like let's see linkedin
next day putin be like if he had to say that he is eating roti: I'm glad to announce that I'm consuming carbo hydrate rich circular shaped thing called as roti and I'm really thrilled to notify my connections
ππ€£π€£π€£π€£π€£π€£
My frnds also call me putin instead of Puneet!
π Full real-life scenario
Meanwhile, Anmol is announcing the next 5000000th milestone achieved
vladimir putin
β οΈπ
π Yea im putin
this is my unoff. account!
No you are superstar!
sir, please do not nuke
π Ohhh... yes yes im Puneet Superstar
Well, why does everyone call me βdidiβ? Like, almost everyone does ππ» I mean, itβs nice, but itβs kind of funny too!π
because you're didi
π€£
Didi of dev.to award 2026 & forever goes to mrs. Divyanshi didi also known as (@technogamerz)
Please use , Ms not βMrs.βπ
bcoz u r didi... β
when puneet enters the chat, chat became π
@effessdev bro I'm publishing post hope you will redirect ;)
Ok redirecting..........
Why and what does ms means??
Lol π π€£ only Mr and Mrs are there Didi is ms is for more educated person like IITian?
Guys can you stop this thread is lagging the whole article page π€£
let them continue... ππ€£
oh sorry lol I mistakenly typed putin β οΈ
β β β
ππ warm hearted welcome didi @technogamerz
ππ»β€οΈ
π Before the build in public series of crowdwide, I used AI to write my articles...
I'm here because 2 days holiday of my school so I'm gonna sleep late
but my messages are pure hand-typed (or it is finger-typed? π€)
"think about it bro"
π€£ bruh where is option of gif in dev!!
Lol i dont want the thread
ππ
ππππ THIS GIF!!
Yep. And not only on our way to college but inside college too (if one department is on top of the hill, the other is on the bottom lol).
Your college is really based on real-life ups and downs π
YOOO LET'S GOOO!! π₯ Welcome aboard, brooo!! π―π€ Hope you enjoy being part of Crowdwide!! β€οΈ
Thank you so much! π₯³π
βπ
The sequencing here is the thing I would highlight back to you: core algorithm and mailing system done before UI polish is exactly the order that keeps an MVP alive. I have seen the reverse too many times β founders polishing gradients while the activation path is broken, and the pretty UI becomes expensive camouflage for a product nobody gets to use. Making the mailing system part of the core feature set is the interesting call to me; most social MVPs defer notifications and then wonder why nobody comes back for a second visit. And overlaying the sidebar instead of compressing the page is the right instinct β reflow shifts on every toggle would have been a paper-cut complaint from every user, every session. One concrete question: for the theme contrast issues you fixed, are you checking those ratios by eye, or against a standard like WCAG? A systematic pass with a contrast checker before the next theme update would catch the bad pairs before your users do.
that's some really valuable feedback!! π€
Honestly, I was mostly checking the theme contrast visually, but you're right, using WCAG contrast ratios would be a much more systematic way to catch issues. I'll look into running a proper contrast check before the next theme update!
And yeah, I really wanted to get the core features stable before spending too much time polishing the UI Thanks for the detailed feedback, really appreciate it!!
That sequencing instinct is worth protecting. The cheapest version of a systematic check I have used: run your palette tokens through a contrast checker once (axe DevTools has a quick theme-level pass) and pin the minimum ratios as lint rules. I once caught a failing secondary-button state that way that looked perfectly fine to my eyes.
Core-first really is the order that keeps momentum. What part of the polish pass are you most looking forward to once it is stable?
that secondary-button example makes so much sense! π₯ I'll definitely check out axe DevTools and try to make the contrast checking more systematic.
and honestly, I'm most looking forward to making the whole UI feel consistent across themes and getting the sidebar and feed interactions really smooth! Once that's stable, I can focus more on user requests and moderation.
Visual contrast checks lie in exactly the way WCAG ratios don't β I once shipped a 'looks fine' secondary button that failed at 2.1:1 on a projector during a demo. axe DevTools will also catch focus-order issues that contrast tools miss, which matter more for keyboard users. Core-features-first was the right call, by the way β polish on an unstable layout is just expensive repainting. Which page are you running axe on first?
that 2.1:1 example is wild π I'm gonna start with the high-traffic pages first: landing page, dashboard, posts/articles, polls, Widewave, and community pages!
Also, I'm curious about something β how can a page look perfectly fine visually but still fail the WCAG contrast check? π€ Is it mostly because our eyes perceive colors differently, or are there other factors too?
It's mostly the second thing β the check measures math your eyes are bad at. The ratio comes from relative luminance with weights tuned to human vision (green counts far more than blue), and your eyes do local adaptation: mid-grey looks brighter against a dark background than the formula says it is. Thin strokes and large text cheat it further β they read fine at 20px even at 2.1:1, and the ratio treats all text the same regardless of size. So 'looks fine' is your brain's renderer and the ratio is the spec. Trust the spec β the projector in the demo room won't do local adaptation for you. Which page surprised you most when you ran axe?
I was genuinely shocked when I ran axe and found that almost every page had at least a few accessibility issues!! I thought things were looking fine, but apparently there's a lot more to polish thanks for pointing me toward this tool, Odinaka!
That 'almost every page' result is exactly why I start with the high-traffic pages you named β landing, dashboard, and posts β instead of trying to fix everything at once. In my experience the same few failures repeat everywhere: low-contrast secondary text, buttons without names, and form labels that are only placeholders. Fixing those three patterns once usually clears most of the list. Which page gave you the longest axe list β the dashboard or a form-heavy page?
THE DASHBOARD!! It had the most issues out of all the pages I checked! Looks like I've got quite a bit of polishing to do there. Thanks again for all the guidance, Odinaka!! π€
The dashboard being the worst page actually makes the fix order easier, not harder: it is the page every user lands on after login, so one broken pattern there repeats in front of everyone. I would fix the dashboard's shared components first β the low-contrast secondary text and the unnamed buttons you found everywhere β and re-run axe on that one page before touching the rest, because a component fix there propagates to every screen that reuses it. Did the dashboard's list come mostly from one repeated component, or was it spread across different widgets?
it was spread across different widgets on the dashboard! Not just one component causing everything. Looks like I'll need to check each widget and see what's going wrong!
Spread across widgets changes the fix order in a useful way: instead of hunting one shared component, I'd group the dashboard list by failure type first β all the low-contrast text in one pass, all the unnamed buttons in another β because the same widget pattern usually repeats with different labels. I once cleared a long axe list that way and found the real work was only three patterns wearing different names. Which failure type is longest on your dashboard list β contrast, names, or labels?
Labels are the longest on my dashboard list!
Labels being the longest tells me something useful, because the label bucket splits into two very different fixes and only one of them shows up cleanly in an axe list. A missing label announces itself as broken. A duplicated or misleading label β every button on different widgets just saying "Submit", or visible text that does not match the accessible name β passes the check while sending a user to the wrong control silently. When I cleared long axe lists of my own, the duplicated-name bucket was always bigger than it first looked, because the same component repeated across widgets with the same name. If I were ordering your fix, I would make every label unique per widget first, prefixing with the widget name, and only then sweep the missing ones β a wrong label misleads, a missing one just complains. Are your longest bucket labels missing entirely, or the same label repeated across widgets?
Ohh, in my case, I found both types of issues π Some elements need labels, while on others, adding a label doesn't seem to have any effect. I'll check them separately and see what's actually causing the issue.I fixed a few of them, but most of the issues are still left! Thanks for the detailed explanation, Odinaka!
Both types plus a third case is exactly the shape I'd expect, and the 'label has no effect' ones are the ones I'd chase first, because they're usually not label problems at all. In my own axe lists the label that changed nothing was almost always a mismatch somewhere else: a for attribute pointing at an id that doesn't exist, or the control wrapped in something that hides it from the accessibility tree, so the checker sees the label and the screen reader never does. Checking them separately is the right call β the missing ones are a sweep, the no-effect ones are individual diagnoses. Which pattern are the no-effect ones sharing, same component or scattered?
Haven't checked the pattern yet lol I fixed a few, but most are still left. I'll see if the no-effect ones are coming from the same component or different places. In next few updates mostly will be solved!
Also, what if I leave those no-effect labels as they are? If they aren't causing any actual issues, should I keep them or remove them? π€
I would not leave them as they are, even if they cause no visible issue, because the cost is not the label itself β it is what a permanently noisy axe report does to the next real finding. Once a dashboard always shows a pile of known-harmless warnings, the next developer learns to ignore the report entirely, and a genuine barrier slips through unnoticed. That is the failure mode I have actually seen. So I would split the pile before deciding: if the element is decorative, remove the label and hide it properly instead of leaving a label that does nothing; if it is interactive, a label that looks like it has no effect still matters to a screen reader user, so it is a real fix, not a cosmetic one. And since you suspect they may come from the same component, fixing the component once beats clearing them page by page. Are most of your remaining ones the decorative kind, or the interactive kind?
I'll remove those unnecessary labels! Most of the remaining issues seem to be with decorative elements, actually. I'll go through them and clean them up.
Good call removing those labels, and for the decorative ones I would go one step further: mark them aria-hidden and give them an empty alt where they are images, so a screen reader never announces a meaningless label at all. In my own UI work the trap was leaving a decorative icon with a name, so it sounded like a button. Are the remaining ones purely visual, or do any of them carry meaning a sighted user would miss?
Most of them seem to be purely visual, from what I've seen so far. I'll check if any of them actually carry meaning that might not be obvious to a screen reader and handle those separately. The decorative icon sounding like a button is a good example of what I need to watch out for
That distinction is exactly the right filter, and I would apply it the way you describe: if removing the icon changes nothing a user needs to do, hide it from the screen reader entirely. In my own work I have caught the same trap you name, a decorative icon that a screen reader announced as a button, which made a purely visual flourish sound like an action. The check I now run is simple: I tab through with the screen reader on and ask whether every announced element earns a decision. Which of your icons felt borderline to you, the ones that carry a status or a hint rather than just decoration?
That βevery announced element earns a decisionβ check is actually a really good way to think about it I haven't fully checked the borderline ones yet, but status icons and hint icons are definitely the ones I'll look at more carefully. I don't wanna hide something that actually helps users understand what's happening.
You are right not to hide those, and that is where my check needs a second step: a status icon should stay, but it should announce the state in words, not just as an icon name. In my own work a green dot that only said image meant nothing to a screen reader user, while the same dot with a short label like active or pending actually helped them understand what was happening. I would keep the status and hint icons, and give each one a plain-language label, rather than removing them. Which status in your UI would be hardest to describe in a few words?
That makes sense! A short label like βactiveβ or βpendingβ is way more useful than just announcing an icon I think the trickiest ones in Crowdwide would be status indicators around posts and community activity. I'll check how to make those clear without making the screen reader output too noisy.
That label point is right, but I would add the noun to it, because βactiveβ on its own still leaves a screen reader user guessing what is active. In my own UI work I found βPost: activeβ and βCommunity: pendingβ worked where a bare label did not, and I kept the icon itself silent so it was not announced twice. For the noise problem, I would put the status once in the card heading order rather than repeating it on every control. Which Crowdwide indicators are you finding trickiest, the post ones or the community activity ones?
Oohh, adding the noun makes sense! βPost: activeβ is much clearer than just βactiveβ π I think the post indicators might be trickier because I need to make the status clear without repeating it everywhere. I'll check both, though, especially how the status fits into the card heading order.
the card heading order is the right place to test that. post indicators do get noisy fast.
Yeahh, that makes sense! I'll work on the heading order and clean up the noisy post indicators. I think I can get these accessibility updates done by tomorrow π
nice, that cleanup should make the cards breathe a bit.
that's the goal! I wanna make the cards cleaner without losing the useful info for screen reader users too
cleaner cards that still carry the info for screen readers. that's the balance to keep.
Exactlyyy! π I don't wanna make the cards look cleaner visually but accidentally make them less useful for screen reader users. Still figuring out the right balance, but this is helping me think about it better π
yeah, cleaner that hides info isn't cleaner. keeping the useful bits in reach is the whole job.
Exactly! π Thatβs what I was trying to get right with Crowdwide β making the UI cleaner without hiding the things people actually need. Itβs done now! π You can check it out in my latest article, Day 23 of Building Crowdwide. π
cleaner without hiding things, and already on Day 23. fast loop.
yeah! Trying to keep the UI cleaner without hiding things. Day 23 already, and thereβs still a lot to improve. Appreciate you noticing the fast loop!
Day 23 and still a lot to improve sounds like the loop is doing its job.
true! Thereβs always something to improve, and every time I fix one thing, I notice something else. Just trying to make Crowdwide better step by step!
That's a good catch!
Thanks β the contrast-ratio habit feels slow the first week and then becomes automatic. Curious to hear what axe flags on your build; the first run always surprises.
Glad you think so! Really appreciate the feedback and support!
Huge congrats on hitting Day 21! Getting the core algorithm and mailing system done before touching the UI is a massive milestone. For the sidebar dilemmaβyou made the right call going with the overlay approach. Compressing the main feed causes heavy layout shifts and reflows, which feels really clunky, especially when users are just trying to peek at navigation. Keeping the main content fixed and just dimming the background is much smoother. Also, for that gradient stretching issue you mentioned, using background-attachment: fixed or a fixed pseudo-element behind the content usually fixes the weird top shading. Good luck with the remaining polishing, and thanks for the #HackerStage shoutout!
thank you so much!! π₯ And yeah, the overlay approach feels much smoother instead of compressing the whole page!
Also, thanks for the gradient fix suggestion, I'll keep that in mind! Really appreciate the detailed feedback and support, bro! π€
Anytime, bro! Glad the gradient fix makes sense. Keeping the overlay UI is definitely the right call for maintaining focus. You're absolutely crushing it on this build-in-public journey, keep going! π₯
appreciate you bro!! Thanks for the support, means a lot! More updates coming soon
You're very welcome! Looking forward to seeing what you build next. Take your time with the upcoming updates, and keep crushing it!
ππ
Can you add optimistic UI updates? Now when I try to like a post, it takes quite a while. If you want to make it accurate, then what about something to indicate that it's still loading?
YESSS!! I actually saw your post about this yesterday, added it to my to-do list, and started working on it a little yesterday! I'll get it done today!! Thanks for suggesting it, bro!! π€
Thanks for taking me seriously bro βΊοΈ
BROOO, OF COURSE!! π€ Your suggestions matter, and if something can make Crowdwide better, I'll definitely take it seriously!! We're building this together!!
Ok! π Thank you so much! βΊοΈ
π€ always bro, keep the suggestions coming
@effessdev Happy Birthday to you and your mom in advance! πβ€οΈ
The sidebar overlay sounds like a good choice, especially for avoiding layout shifts. One thing I'm curious about: when you're polishing a UI, how do you test what happens when the user interacts with it under less-than-ideal conditions?
For example, if someone clicks a button to create a post and the request fails, does the UI preserve their input and let them retry? Or if they click twice before the first request finishes, how do you prevent duplicate actions?
Do you have a small checklist for testing these edge cases, or do you usually discover them while using the app? I'd love to hear how you approach this as Crowdwide grows!
Iβve made my repo public and Iβm building Crowdwide in public so everyone can see what Iβm building and suggest changes I can make. Actually, most of the features and improvements in Crowdwide are suggested by users themselves!
For the optimistic UI, I built a system on the frontend and adjusted the backend accordingly. For example, when you click the button to post something, it disappears immediately without waiting for any server response, and you just see
Posting.... Thatβs how I prevented users from clicking the button multiple times while the first request was still processing.Iβve also made a basic checklist for testing the UI and other edge cases, but honestly, I discover the more serious issues when I test the app like a real user and use it normally. π
Thatβs also one of the reasons Iβm building in public β people can use Crowdwide, suggest improvements, and help me discover things I might have missed!
You changed the logo
which logo?
π i used AI to generate the banner of post... so tht ai added random logo there
logo of crowdwide is same lol
21 days in and core algorithms + UI polish done is a huge milestone! Congrats on staying consistent with #buildinpublic! π
Regarding your sidebar dilemma: making it an overlay with a backdrop blur/dimmer (rather than compressing the page content) is almost always the smoother choiceβespecially for mobile or responsive web layouts. It keeps layout shifts to a minimum.
Good luck with the final stretch and moderating! Rooting for Crowdwide! π
thank you so much!! and yesss, the overlay approach felt like the better choice to keep things smooth without shifting the whole page! Really appreciate the support, we're getting there!! π€
I have a built a cli based sql anamoly triage detector and I'm new to this kinda stuff . How do I find beta testers
Since you're just starting out, I'd suggest sharing your project in communities where developers and database engineers hang out. You can try Reddit, DEV Community, relevant Discord servers, and LinkedIn.
Make a short demo showing how your CLI tool detects SQL anomalies, explain who it can help, and ask people to try it out and share feedback. You don't need a huge audience even 5β10 relevant testers can help you learn a lot!
Also, make sure your GitHub README clearly explains how to set it up and try it. All the best!
UI polish is one of those stages where small decisions can make a big difference to how a product feels. I like that you focused on readability and the sidebar instead of just adding more features.
A platform built to give smaller creators a voice also needs to feel comfortable and intuitive enough that people actually want to stay and use it.
Exactly and i just realized that i made UI little bit complex...
Providing too many option in front just after login... I need to fix that too π
Some comments may only be visible to logged-in visitors. Sign in to view all comments.